Field of Science

Death by human stampede

Over the past 30 years, stampedes have killed at least 7,000 people and injured another 14,000. That's the conclusion that Edbert Hsu (Johns Hopkins Medical Institutions) and colleagues reached after a painstaking trawl of news reports in the world's English-language media.

The real toll is probably even higher, of course, but the data were enough to allow Hsu to work out the characteristics of the most lethal stampedes. They found reports on 215 stampedes, of which 49 occurred at sporting events, 25 at musical events, 38 were political and 41 were religious. The rest (totalling 60) were due to a mixed bag of causes and were mostly spontaneous.

And the award for the most lethal type of stampede goes to... religious ones! In simple terms of the number of fatalities per stampede, religious events come out over double that of their closest rival.

The simple comparison is not a very fair, however. Religious stampedes take place in different parts of the world (often in the Middle East, which is the most dangerous place to be in a stampede), often in low income nations (also very dangerous), and often outdoors (slightly more dangerous than indoor stampedes).

But even when you take all this into account, religious stampedes still come out on top of the lethality stakes - but sporting stampedes are so close as to make it a photo finish.

There's one other factor that contributes to the lethality of a stampede, and that's the size of the crowd. Unfortunately, Hsu was only able to determine the size of the crowd in 130 cases.

Even taking into account crowd size, religious stampedes are still pretty dangerous. When you look at fatality rate (i.e. deaths per crowd member), they're 6 times riskier than stampedes at sporting events.

But with with crowd size taken into account, religious stampedes drop into third place. The riskiest kind of stampede by a long way are the spontaneous ones (because of the lack of crowd control), followed by political ones.

The explanation for all this is fairly simple. Religion is the one event that brings together truly massive crowds, often in settings that are poorly controlled.

One of the most lethal stampedes in recent history occurred in Iraq in 2005, when nearly 1000 people died when fears of a suicide attack sparked panic. In the same year, over 250 died (out of a crowd of 400,000) when Hindu worshippers set fire to shops.

But the biggest contributor is the annual Hajj, which these days draws crowds in excess of 2 million. Five of the biggest stampedes in the past100 years occurred in Mina Valley, Saudi Arabia, during the Hajj.

Over the past 3 decades, nearly 3,000 people have been killed in stampedes during the Hajj - the last big disaster being in 2006. With crowds that big, I suppose the surprise is that there are so few casualties!

ResearchBlogging.orgHsieh, Y., Ngai, K., Burkle, F., & Hsu, E. (2009). Epidemiological Characteristics of Human Stampedes Disaster Medicine and Public Health Preparedness, 3 (4), 217-223 DOI: 10.1097/DMP.0b013e3181c5b4ba

Creative Commons License This article by Tom Rees was first published on Epiphenom. It is licensed under Creative Commons.


  1. This post has been linked for the HOT5 Daily 3/19/2010, at The Unreligious Right

  2. My first thought was to wonder what happens to the comparison if you subtract out the Hajj. The Hajj, and its inevitable accompanying stampedes, are somewhat of an aberration.

  3. Having lived in Asia for 12 years, I grew amazed at how close people stand in line. Where I live in the USA, there is a much broader area considered personal space. It irritated me unendingly in Asia (though I slowly acclimatized).

    I've always wondered if this "smaller personal space" is one of the contributing factors to deadly human stampedes. Religion does not seem important in this study except for the number who gather at an unstructured event.

  4. Is there a difference in the health profile of people caught in a religious stampede vs other types (sporting, musical for example.) Off-hand, I'd think they might tend to be older and more frail, less able to withstand the traumas involved in a stampede

  5. It's an interesting point. They don't make any mention of age or frailty, either in the paper or in the accompanying review (although they do mention that most people die while still standing - they're not knocked to the ground and trampled).

    Arguably, a stampede of fit, young men would be more dangerous than a mixed crowd - because of the forces involved!


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS