3 months ago
19
At some point during Sunday’s pro-Palestine march across the Sydney Harbour Bridge, NSW Police decided to send geo-targeted text messages asking participants to stop walking north. The concern, they said, was a potential bottleneck at the northern end of the bridge, where dispersal options and train access were limited.
Was it an overreaction? Was there a genuine crowd safety risk?
I believe it’s not fair to judge the police response simply because no disaster occurred. In real-time risk management, you rarely have complete information. You act based on estimated probabilities and the potential severity of consequences if things go wrong. And when the potential consequence is serious harm or death, even a low likelihood can justify taking action.
Pro-Palestine protesters march across the Sydney Harbour Bridge on Sunday.Credit: Janie Barrett
In this case, having considered the risk factors and the level of uncertainty involved, I believe police acted in the interest of public safety.
The Sydney Harbour Bridge is not a narrow structure; it’s almost 50 metres wide. But any fixed piece of infrastructure can become overwhelmed if enough people are on it. Once a crowd is committed to the bridge, there are no lateral escape routes or soft dispersal zones. That makes it harder to manage crowd pressure or respond quickly if something goes wrong.
The Supreme Court ruling that gave the go-ahead for the event came less than 24 hours before it took place. That left very little time for anyone to carry out formal crowd-flow analysis or a proper risk assessment. As a result, safety on the day depended almost entirely on how well things were managed in real time. Decisions had to be made on the spot.
There was little time to estimate how many people would turn up. Even after the event, the numbers are still being debated. On the day, it’s likely police had to rely on quick estimates, often using aerial views to judge how crowded the bridge was. But the rain would have made that more difficult. Many people were using umbrellas, which would make crowd density more difficult to assess.
More technology, like camera-based crowd analytics, may have helped. But those systems need more lead time to set up and rely on clear visibility to work well.
Another risk factor was the make-up of the crowd. This wasn’t a uniform group of adults. It included families, children, older people. In a crowd crush, it’s often smaller or less physically capable people who are affected the most. A diverse crowd requires more conservative safety assumptions.
Further to that, rain changes how people move, just like it changes how traffic flows. On the road, even a light shower can slow things down and cause bottlenecks. The same applies to crowds on foot. Especially, when people are holding umbrellas, they take up more space and move more cautiously. Basically, their physical footprint gets bigger. Studies show that pedestrian flow can drop by 20 to 40 per cent in one direction, and by up to 50 per cent when people are moving in both directions.
The police intervention was far from perfect, though. The first NSW Police geo-targeted text message, sent at 3pm, read: “Message from NSW Police: In consultation with the organisers, the march needs to stop due to public safety and await further instructions.”
The intention was to slow down the upstream crowd due to concerns about a potential choke point forming at the northern end of the bridge. Whether the message was clear enough to achieve that outcome is debatable. While a later follow-up message provided a clearer directive, protesters may have interpreted the message differently. Compliance is never 100 per cent. Some may have ignored it, thought it was spam, or didn’t believe it was serious. Others may have stopped, while some turned around. That meant creating a two-way pedestrian flow which adds its own risks and complications.
That said, if there were serious concerns that the northern end couldn’t be cleared quickly enough, then holding back the flow upstream was likely necessary.
The fact that no one was injured doesn’t mean there was no risk. Crowd crushes have happened during political protests in other countries. I can point to nearly a dozen examples.
An estimated 90,000 people attended the rally in Sydney.Credit: Getty
Would a crowd crush have definitely occurred without police intervention? The likelihood, in my estimate, was low. The more likely outcome would have been discomfort, pressure, and a degree of crowding chaos — similar to what we sometimes see during peak moments at Vivid, where people grow anxious and complaints follow.
But here’s the thing: if it had reached that point — a near-crush situation even with no injuries — we’d likely be having a very different conversation today. We would be asking why police didn’t step in earlier.
That’s the reality of risk management. These decisions are not made based on hindsight. They’re made in real time, with limited information, under pressure, and with one rule in mind: when the potential consequence is serious (injury or death) even a small chance may warrant action.
Australia has traditionally had a low tolerance for public safety risk. But this is a big part of why major crowd disasters are rare in this country, but common elsewhere. Sunday’s police response was consistent with that safety-first culture.
Milad Haghani is an Associate Professor of Urban Risk, Resilience & Mobility at The University of Melbourne.
Most Viewed in National
Loading
