The anti-government protests in the capital over the last two weeks have been large, that’s for sure. But when it comes down to exact numbers, some say 100,000 and others half a million. What’s the science behind estimating crowd sizes, and how accurate can you be? Bennett Murray and Khouth Sophak Chakrya found out.

IUf Counting blades of grass is the pinnacle of a difficult and tedious task, try counting protesters at a political rally. For journalists, police and public officials in cities around the globe, from Cairo to Bangkok, determining the sizes of large public gatherings is a regular, if painstaking endeavour. Over the past three weeks, Cambodia has experienced a fair few.

Sunday, December 21, marked the beginning of the Cambodia National Rescue Party’s “tsunami” marches that have eclipsed all other Cambodian demonstrations in the post-war era in size. But while everyone agrees that the protests have been massive, the estimates vary drastically. Opposition leader Sam Rainsy announced that about 500,000 protestors took to the streets in the “tsunami”, while the Phnom Penh municipality estimated that no more than 100,000 showed up, and the Post reported “more than 100,000”.

A number of theories and factors come into play when guessing crowd numbers, and not least the bias of different parties. Both the government and CNRP have their own vested interests when they guess protest sizes, said independent political analyst Kem Ley in an interview this week.

“I found that the [Cambodia People’s Party] reports less and less because they don’t want other people to attend, and the CNRP wants to report more and more than the reality because they want to encourage others to attend who did not attend,” Ley said.

Cambodia is not the first country where politics has provoked contention in protest size estimation. In 1994, Louis Farrakhan of the Nation of Islam black nationalist group threatened to sue the US National Park Service for libel after it reported that only 400,000 people attended Farrakhan’s ‘Million Man March’ on Washington (the National Park Service stopped crowd counting altogether afterward).

But even independent observers have their work cut out when protesters take to the streets in large numbers. In the midst of the sometimes chaotic atmosphere of major protests, getting a scientific head count requires swift thinking and potentially compromising guesses.

Herbert Jacobs, a late Berkeley journalism professor, pioneered crowd counting in the 1960s to determine the size of the campus’ Vietnam War protests from his office window. The general rule he came up with is that the loosest crowds have about 10 square feet per protester, a tighter crowd has 4.5 square feet per person, while the densest crowds have about 2.5 square feet. Converted to the metric system, that equates to 3.05, 1.37 and 0.76 square metres respectively.

In Phnom Penh on December 21, where the protest march was about five kilometres long on boulevards about 20 metres wide, Jacobs’ formula for the densest situations suggests that some 131,500 marchers took part. This is well below the CNRP’s estimate of 500,000 and somewhat higher than the city’s guess of no more than 100,000.

The CNRP, however, assumes a much denser crowd at 0.25 square metres per protester, said party spokesman Yim Sovann. This would make for about 400,000 marchers in Phnom Penh’s streets on December 21. The other 100,000 protesters, said Sovann, included non-marching demonstrators at Freedom Park, Wat Phnom and elsewhere in the city.

Cambodian analyst Ley, who emphasised that his crowd counting was not an exact science, estimated that an average of three to four people filled each square metre of the march. His total estimate was 300,000 to 500,000 demonstrators.

The intermingling of demonstrators and bystanders can complicate any scientific count. Long Dimanche, Phnom Penh municipality spokesman, said those stuck in the traffic jam were “hostages” of the protest.

“In fact, they were the victims of the CNRP’s demonstration. They were not the supporters of the CNRP,” he added.

Although he stuck by his original statement that no more than 100,000 marchers were present, he acknowledged that it was an unscientific guess.

For many analysts, however, the implications of the marches are more important than nailing down a specific number.

“The number is not important right now, because this number is enough for the Cambodian government to respond to the needs of the people,” said Ley.