Atlantic Cities

Why Crowds Can Turn Deadly

Why Crowds Can Turn Deadly
Wikimedia

A crowd in motion is a pretty remarkable thing. Pedestrians walking down a sidewalk in opposite directions usually manage not to bump into each other. Same for people sharing a crosswalk from opposite corners of an intersection. At a larger event, people will often form what look like neat traffic lanes moving through a static crowd when, say, the port-o-potties are on one end of a festival and the beer stand is on the other. Or think of a time when you’ve tried to cross the concourse to get to the bathroom at a baseball game when everyone else is streaming for the exits. Frogger-like, most of will us move diagonally through the flow of fans such that no one has to stop and no one collides.

All of these patterns are subtle signs that pedestrians in motion behave an awful lot like fluid.

"It happens like magic," says Dirk Helbing, a professor in Zurich, Switzerland, who studies sociology and crowd modeling. "People don’t have to think about it, you don’t need to have legal regulations or policemen to organize the crowd. It just happens, like this invisible hand like what Adam Smith described."

Pedestrians naturally self-organize, leading to macroscopic patterns of motion that allow us to seamlessly move among each other without experiencing friction. Or, at least, this is what happens under normal conditions, on a sidewalk or in a spacious park. Pack people into a really dense crowd – at a concert or a political rally – and the physics become quite different.

Helbing has been studying pedestrian flows for about 20 years, but his most recent research examines this thornier scenario: Why do deadly crowd disasters happen? In the news, we know these events by their more salacious vocabulary – stampedes, panic, trampling, crushing – loaded language that further muddies the question of why such events occur in the first place. Do soccer matches and rock concerts periodically turn deadly because individuals in crowds act like jerks? Does this happen because crowds collectively go "mad"? Something about the idea of a "stampede" seems so prehistoric. How could people in the 21st century still crush one another at a music festival?

Helbing’s research suggests, as with other pedestrian modeling, that the answer lies more in physics than psychology.

"At very low density, when everybody can move freely, [crowd dynamics are] like a gas," he says. "When the density goes up, then eventually peoples' movements are constrained, and it becomes more like a fluid. And then at very high densities, when people are squeezed in between other bodies, it’s more like a granular material."

Like sand, or rice, or small pebbles.

Helbing and co-author Pratik Mukerji brought this perspective to an in-depth study of the 2010 Love Parade techno music festival in Duisburg, Germany. The festival was held in a fenced-in freight yard designed to hold about 250,000 people, with only one dedicated entrance and exit. By some accounts, more than a million music fans turned up. And, in the end, 21 of them died and more than 500 were injured in what most news reports called a "stampede."

The Love Parade followed a series of high-profile deadly crowd disasters over the past several decades, from soccer stadiums in England to rock concerts in the U.S. This catastrophe, however, was widely documented by the public on social media (and, amazingly, has its own YouTube channel).

The Love Parade suffered from familiar missteps: walkie-talkies and speaker systems didn’t work, event organizers and police miscommunicated, the logistics of the festival itself were poorly planned relative to the size of the crowd that eventually turned up. But all of these cascading effects culminated in a phenomenon best described by physics: "crowd turbulence."

A crowd of people trying to get into such an event will queue up until they've moved closer and closer together, creating particularly tight compression at the front of the queue. As a crowd grows denser, people wind up accidentally touching one another – and this happens even without overt pushing or shoving.

"As the density increases even further, the forces would be transmitted from one body to the next, and this is the moment where forces start to add up," Helbing says. "That causes these turbulent waves in the crowd."

That "crowd turbulence" actually looks like wavelike patterns you can see on aerial video footage of the event (skip to the 1:01 mark on the below video, and be warned the scene is disturbing):

The small movements of so many people aggregate into a powerful force – one that security officials are often helpless to halt – that has the capacity to knock over bodies, shove them together and, ultimately, asphyxiate them.

This sounds impossible, but 21 people died at Love Parade inside a crowd that had essentially been standing still. There was no real crowd rush or dramatic "stampede." And this is the heart of the mystery to non-scientists as to how such a thing could happen.

"Why do people think it's panic that causes crowd disasters?" Helbing asks. "They just cannot understand how it can happen that people can die although nobody is behaving in a ruthless way."

Traditionally, we've explained such disasters as resulting from the forces of a panicked crowd (or, worse, an angry or violent one). "But these are just the forces that are transferred from one body to the other," Helbing says.

Inevitably society looks to blame someone for such disasters (and that happened in Germany after the Love Parade). It’s much harder to blame physics, or to understand the circumstances at a mass event that would allow these laws of physics to unleash.

"There is a change in perspective in the scientific community," Helbing says. "However, that’s not enough. Because in the end people who organize mass events – the security, and crowd managers and police – need to know about it."

The above photo, via Wikimedia Commons, shows the crowd at the entrance ramp to the Love Parade in Duisburg, Germany in 2010.

Emily Badger is a former staff writer at The Atlantic Cities based in Washington, D.C. She now writes for The Washington Post. All posts »

Join the Discussion