Stampede

As if the deaths from hurricane Katrina were not enough, yesterday brought the news that up to one thousand people, predominately women and children and the elderly, had died in a stampede while taking part in a Shi'ite religious pilgrimage in Iraq. Another horror of nearly unimaginable proportions.

It became clear almost immediately that, although terrorists were not directly to blame, they were an indirect cause, since the stampede was apparently sparked by a rumor of a bomber in the crowd. Such a rumor was given credence by the crowd's knowledge of previous suicide attacks on Shi'ites at similar religious pilgrimages. So, ironically, a rumor of terrorism has caused more innocent deaths than any single terrorist attack has done so far in Iraq. The terrorists must be extremely pleased to have achieved such a goal with so little effort on their part.

It's not surprising that the smaller and weaker--women and children and the elderly--would be the most likely to be overwhelmed by the force of the crowd and crushed, and were therefore overrepresented among the dead. I would guess (though I haven't been able to find information to document it) that most stampedes involve a similar grim statistic, if they occur in crowds that feature any women and children or elderly people.

I hold a vivid memory of being caught in a rush hour crush in a New York subway some years back. Although, fortunately, the situation never reached stampede level, it was terrifying to me because I was holding the hand of my 3-year-old son, who promptly disappeared, thoroughly engulfed in the crowd. I still recall the feel of his small and delicate fingers in mine, and the panic that engulfed me as the force of the crowd started to carry him away. Fortunately, all was well, but ever afterwards it has taken very little imagination for me to realize the dangers small children and mothers face in adult crowds.

Yesterday's stampede was very big news mostly because it occurred in Iraq. But it turns out that stampede deaths are a regular occurrence around the world, and not just in third world countries, either, although in general they seem to cause higher death tolls there. This website has an exhaustive list of such stampede events, which can be found by clicking on "disasters" in the left-hand sidebar.

On analysis, it turns out there are three main categories of venues that would appear to favor stampedes: the soccer stadium (or other large sporting event); the crowded nightclub in which a fire breaks out; and the religious pilgrimage. They all share the characteristics of having very large and moving groups of people packed into a restricted space. But panic, such as apparently occurred on the Iraqi bridge, is, surprisingly, not a required element to start such stampedes, although it inevitably happens as the stampede begins to take shape, and makes them that much worse. Stampedes can sometimes be sparked in the absence of any panic, when chance events block the flow of traffic in an overcrowded and spacially restricted situation.

The situation, as far as I can determine, is a bit analogous to the elements that go into a tsunami, strangely enough. That is, a huge and extremely powerful force (in the case of crowds, the moving people; in the case of tsunamis, the moving water) is initially spread out horizontally. Then, some sort of blockage impedes that horizontal movement and converts it, at least partially, into a vertical one. I haven't found a website that explains this too clearly, so I'm not linking to any source for it, but it appears that, in the case of a stampede, people become stacked up and those on the bottom are the ones who are crushed by the force of those above. Of course, in yesterday's Iraq tragedy, some also went off the bridge and were drowned.

It's no accident, either, that the Iraqi stampede occurred on a bridge. Any sort of bottleneck or narrow passage through which the crowd must funnel itself represents a grave danger, because it potentially impedes that flow of horizontal movement.

That's not all that contributed, in this case. Here is a particularly telling description of the Iraqi stampede, from the previously linked Globe article:

General Rawad Rumediam, a military commander at the bridge, said that 3-foot-high concrete barriers put in place to prevent car bombs from entering probably contributed to the crush. Saddoun Dulaymi, Iraq's defense minister, said the checkpoints at the bridge meant to search pedestrians for explosive devices may have slowed the flow of the crowd across the bridge and contributed to the disaster.

So it seems that terrorists helped the disaster to occur in two ways: by giving credence to the rumor of a suicide bomber, and by causing security considerations to override the implementation of basic crowd control safety measures.

We don't usually think much about it, but any time there is a large public gathering, the science of crowd control comes into play. The police study crowd control, and there are also other scientists who try to improve on current knowledge and apply it to future situations. This website, on which I found the list of stampedes, is an example of a firm that specializes in such research, consulting with groups around the world to prevent similar disasters. For example, they were hired by the Saudis to supervise this year's Haj (in particular, they redesigned a certain bridge over which the crowd needed to pass). They seem to have been successful because, unlike some earlier Hajs, this one apparently went off without a hitch.

Crowd control is designed to minimize the possibilities that things will get out of hand. But, sadly, there are no guarantees that they won't, despite the best of precautions. And often, especially in third-world countries, the expertise and the money to implement the best of precautions are not in place (here, for example, is a description of how a lack of preparedness and resources helped create the conditions for a stampede in India during a religious pilgrimage).

What follows is a description of a stampede which, as far as I can tell, could not have been prevented by any method, other than total redesign of the area. Occurring in an urban setting in a developed country, it was started by a series of chance happenings. It's another example--as if we needed reminding--of the limits of human control over events. (This passage appears at the same Crowd Dynamics website, but it has no separate URL. It can found there by clicking on "disasters"):

1999 (May 31) Minsk, Belarus. 53 dead, 150 injured, 78 hospitalized when a crowd of 2,500 rushed to get out of the rain at the railway station. From The Daily Telegraph, June 1, 1999. An unprecedented tragedy happened on May 30 in the centre of the Belarusian capital. Over 50 people died and some 300 were wounded in a crush at the entrance to the underground station...The tragedy was caused by heavy rain that started at about 8 p.m....A few thousand Minsk residents, mostly young people, had gathered ...The first thunders and rain drops made people rush to find shelter in the underground crossing...Somebody fell down on the concrete floor and the first blood was shed. People were slipping over and trampling those lying on the floor...People were falling at the feet of the crowd. Over two thousand people poured into the 10-metre wide underground crossing thus creating a dense moving jam...there were people literally smeared against the walls, pressed into the floor, ...Meanwhile, screams of those who were unable to escape on their own, kept echoing from this hellish meat grinder..."We are soccer fans, so we know what to do in a crowd--cover your head with hands and make your way to the exit."..."People kept arriving until there was almost no space and then the whole mess started. There was no escape. The people surging in from behind just left the others lying and walked over them," one of the survivors told Russian television.."About 300 people were lying here, one layer on top of another," a policeman said "We were carrying out the top layer of people and they were still alive. Those in the bottom layer were either dead or injured." Two policemen were caught in the crush and also died as they tried to rescue those who had fallen...More than 150 people were taken to 10 hospitals in Minsk as doctors battled through the night to save the lives of the victims in the tragedy. In his speech president Lukashenko said "There is nobody to blame, there is no one to make a claim to, it happened because it happened, even if there was anybody responsible it was the rain that caused the disaster."