There were lots of people at the Jeremy Corbyn rally in Liverpool earlier this week, but not as many as a few of his fans made out. Some tweeted photos of the city’s huge Champions League victory parade in 2005 claiming that they showed the Labour leader’s rally. “Not that you’ll hear about it in the mainstream media,” they cried.
There wasn’t much need for faking. Photos of the real rally, published widely in the mainstream media, still show an impressive turnout. But how impressive? And how do you count a crowd anyway? Carefully, is the short answer, because when estimates can make or break campaigns – or even trigger legal threats – size matters.
The Guardian reported 5,000 people at the rally after police officers there gave estimates between 3,500 and 5,000. Organisers suggested 10,000. One method of calculating crowd numbers is the Jacobs method, named after its inventor, the American journalist Herbert Jacobs. While teaching at Berkeley University, Jacobs watched anti-Vietnam war protests on the square below his office. He applied a grid to pictures of the crowd, counted the people in one square, determined crowd density, and made a calculation.
According to the Jacobs Method, in a light crowd, in which people can reach out their arms, each human occupies 10 sq ft (1 sq m). In a medium crowd: 4.5 sq ft; or 2.5 sq ft at heavy, mosh-pit levels. Police forces and other agencies have used the same principle, adding more complex measures to account for moving crowds. But now, with Google Earth and an app, anyone can calculate crowd sizes.
“In some countries it takes a lot of courage to demonstrate, and I just don’t like the idea of that kind of courage being discounted or exaggerated,” says Gordon Arnold, a retired architect from Dallas, Texas, and creator of crowd-counting app, CrowdSize. Using the app, I zoom the satellite view into Liverpool until the square on Lime Street fills the screen and then finger-tap an outline of the crowd . Select a density and hit done. For Corbyn’s rally, CrowdSize estimates 1,500 people at medium density, or 5,700 at high density – and photos suggest somewhere between the two, so around 4,000 people seems likely.
But people get angry when you downsize them. When the National Parks Service in Washington said 400,000 people had joined the 1995 Million Man March, its organiser, the Nation of Islam leader Louis Farrakhan, threatened to sue. Eventually, the Center for Remote Sensing at Boston University used tree-counting methods to analyse photos and settled on 837,000. Disputes like this, and the varying counts at the 2003 Iraq war protests in London (between 700,000 and 2 million) are why the Metropolitan Police say they no longer issues estimates.
Soon we may not need them. A team from Warwick university working in Milan has used mobile phone and Twitter data to reliably estimate numbers at Linate airport and the San Siro football stadium, where they used ticket numbers as a reference. No photos, grids, density estimates or subjectivity required; at the rally of the future don’t count us, count our phones.
