- Peter McGuire
Why Do Humans Stampede?
Part 1 of a series on human stampedes and crushes
Every year people die from being squeezed or trampled by other people. What causes these human stampedes and how do we prevent them?
Muhammad Hamed / Reuters
You’re at a football game with some friends. There’s standing room only. You got there fairly early so you’re somewhere near the middle of the crowd. At first, it’s not much different from any other big event. As more people arrive, it gets crowded. For a while, you’re annoyed that people keep filling in the space in front of you. You were there first! But soon it doesn’t matter because every space around you is now full. You can feel a slight but perceptible force pushing you into the person in front of you. Then the calls from the front start: Move back! We’re squished against the fence.
You dig your feet into the ground to try and resist the forward motion but it’s too strong and you have no leverage. You’re packed in on all sides so your feet can do nothing but keep you up. Your friends are gone now. If you’re not tall, you can’t see anything but heads and shoulders. You are no longer in control of your fate.
Panic starts to set in. Not just for you, for everyone. You can hear it in their voices. Let me out! Move back! Someone help me! I can’t breathe!
The screaming starts. People aren’t just scared. They’re in agony. Breathing has become a zero-sum game. You try to use your arms to protect your lungs but no matter how hard you push, they stay pinned to your sides. Every now and then a wave surges through the crowd, blasting the air out of you as you slam into the person in front of you. A few lucky ones around you are climbing. But as they get above the crowd, others subsume. The air is hot and wet like a sauna, so the few breaths you manage to take still feel suffocating. You look at the man next to you. His skin is chalk white and his lips are blue. Now you know for sure. You’re dying.
Breathing has become a zero-sum game.
The only way out is up but going up involves a 7-foot climb up a fence, the last foot projecting back at you. These fences were specifically designed to prevent people from getting over or through them. Despite the enormous pressure, it’s the human bodies that are bending, not the steel. Someone small manages to scramble onto your shoulders, a kid. You hold him up long enough for someone on the fence to grab him but then your knees buckle. You simply don’t have the strength to stay standing. The crowd jostles around you. Someone steps on you. Several people step on you or kick you, stumbling over your body. The little space you took up has now been completely filled. A solid mass of human weight pins you to the ground. You already took your last breath a long time ago. Too long ago. You pass out. In a minute you’ve suffered irreparable brain damage. In five minutes, you’re dead.
This describes the experience of 94 people who died in Sheffield on the 15th of April, 1989. Despite the characterization of this event and others like it as “stampedes,” panic is rarely a contributing factor to the deaths. Crowds suffer mass casualties like the Liverpool fans who died in the Hillsborough disaster, are notable for their calm in the face of imminent death.
The Science of Crushing
Perhaps the most frightening thing about crushing incidents is how little control the people experiencing them have to stop it. Once you realize a crush is coming and you are in it, there’s little that you can do to change your outcome. Crush victims quickly find themselves carried along by the flow of people.
A recent review of human stampede literature published in Prehospital and Disaster Medicine describes how stampedes happen in clinical but vivid terms:
...in normal situations, the crowd moves in a free or laminar flow. When density increases, the flow changes to “stop-and-go waves.” This can lead to loss of movement control by the crowd and individuals are pushed randomly – “crowd turbulence.” When a person loses balance and stumbles, or is pushed down, the people around them fall due to sudden forces imbalance. To avoid falling, trampling, which is the act of standing or walking on someone else, may occur. People at the bottom of the crowd eventually die of traumatic asphyxia.
"People don't die because they panic. They panic because they are dying."
Injury from trampling is not the main cause of death. It’s actually the pressure that builds up in the crowd that kills, and often kills people while they’re still standing. They only fall when the crowd disperses, long dead.
In up to 95% of cases, victims have at least one of these three signs: conjunctival petechiae (burst blood vessels in the eye or eyelid), facial petechiae (burst blood vessels in the face), and congestion of the face and neck caused by backpressure of the venous flow. This indicates that victims become so compressed that blood can’t drain back down from their heads, causing blood vessels to burst from the pressure. Oxygenated blood, if any is available, can’t reach the brain. All three of these signs point to the same cause of death: traumatic asphyxia.
Asphyxia is oxygen deprivation. In this case, it’s the brain being deprived of oxygen. It takes less than 5 minutes to die this way.
As such, throughout this series, I prefer the term “human crush” to “stampede.” Stampede conjures visions of people running around trampling others. In fact, crushes are caused by people exceeding a safe local density. Rarely are people able to get anywhere close to running. They quickly lose all agency over their body and the trampling the occurs is the effort to keep from falling down themselves. The term “stampede” creates entirely the wrong image.
This is not to say that crowd rushing doesn’t contribute to tragedies. Many deadly crushes happen when a crowd presses in towards something of interest or away from something of fear or danger. But the key detail is that this isn’t an every-man-for-himself pile-on of panicked sheep. To quote an authority on crowd safety research, GK Stills, "People don't die because they panic. They panic because they are dying." The flow of people that creates the pressure that kills can only be managed on a macro scale. The individual decisions of people in the crush might contribute to whether an individual lives or dies but has little impact on the overall magnitude of the disaster.
If crush victims get rapid first aid and are triaged correctly, many lives can be saved.
Many of the deadliest stampedes were also the deadliest fires. As such, many fire safety measures that were enacted over the 20th century were also human stampede countermeasures. Virtually every commercial building and workplace has a maximum occupancy determined by local fire code to prevent the crushes that happen at evacuation bottlenecks. Bottlenecks at exits can quickly block an exit entirely, dooming everyone still trying to use that exit. The majority of deaths in large-scale fire disasters are caused by exits being blocked by a crush. As recently as 2003, 100 people died in a nightclub fire trying to use a single exit when another was available.
Music venues are especially susceptible because crushes can be very localized events. Even in large, outdoor, festival-style events, small-scale crushes can happen when there is a surge in local traffic, like a crowd pushing towards a stage.
In most cases, the victims of compression can be saved if treated quickly. If crush victims get rapid first aid and are triaged correctly, many lives can be saved. Unfortunately, many of the systemic failures that lead to crushes also lead to a slow and haphazard response. We saw that most recently and viscerally with the Travis Scott Astroworld disaster.
Ultimately, the best remedy for these events is planning the space correctly and reacting quickly to developing problems. Every large venue needs to train staff in crush procedures just like any other disaster that can occur.
Let’s take a closer look at some of the deadliest and most memorable stampedes and crushes in human history to see what they share and how we can avoid them.
Human Stampedes: An Updated Review of Current Literature
Maria Moitinho de Almeida, MD, MPH; Johan von Schreeb, MD, PhD