About 250 million people showed up, wrote the author, Prem Panicker, in an opinion piece in The Wire. He compared it to a “full-fledged total revolution, a phrase last heard in 1974 when the students of Bihar took to the streets and triggered a chain reaction that ended in the declaration of Emergency. " The estimate was of the number of people nationwide who had joined the protests on 8 January this year against the government’s “anti-people" policies.
How do you count 250 million people who are scattered all over the country, happen to be marching and came together as a mob without much notice? Who counted them? Where did that number come from?
In this case, it appears to have come from another article, this one by The Wire staff four days earlier, reporting that “an estimated 250 million people were participating in the nationwide strike called by ten central trade unions on Wednesday against the “anti-people" policies of the Central government".
I instinctively mistrust numbers, especially those ending with several zeroes, and so should you. Crowd-counting has never been a science as much as a best-guess aggregated from several methods and perspectives. However, its importance has been spiking sharply of late and not just because US President Donald Trump falsely claimed his inauguration day crowds were larger than his predecessor Barack Obama’s. As teeming populations of the homeless flee from war zones, seek sanctuary from droughts and drug lords or march against authoritarian regimes, the need for better estimates has become urgent.
There are several algorithms for estimating crowd size, some newer ones even drawing data from drones and tethered balloons with 360 degree cameras, but I would rather ask why we estimate crowds than how. There is assumed to be a direct correlation between the number of people who show up for a shared cause and the importance of that cause. It is therefore a given that the organizers of a rally stand to benefit by overstating the turnouts, while their targets would gain by trivializing it. In other words, there is no reason why any number you read is likely to be true.
I have been trying to find out how many attended the sustained pro-democracy rallies in Hong Kong, knowing that it isn’t a single number but a range and it varies daily. The New York Times estimated that the protest on Human Rights Day stretched “several miles". As many as 800,000 people attended the march, the newspaper reported, but its source happened to be the Civil Human Rights Front, the organizers of the march. The Chinese government might have lopped off a couple of zeroes.
India’s largest contiguous crowd would probably be during Republic Day celebrations in the gardens along Raj Path. No estimate is available from the government or on the internet, though this is a recurring and stationary crowd. However, estimates are available of the anti-rape protesters all over India after the Nirbhaya rape case. It varied from hundreds in Bengaluru and other cities to thousands on Delhi’s Raj Path. As “thousands" could range from two to 999, this is a profoundly useless estimate.
Between us, I would not be able to tell a crowd of 25 million apart from 250 million, and neither could you. When there are too many people, we lean on words like giant, massive, and humongous.
Crowd science says that area and density are two characteristics of crowds. The tightest packing, called mosh pit density after rowdy crowds in heavy metal concerts, is 2.5 sq. ft per person. In such a crowd, you would not fall down even if you could somehow lift both feet off the ground.
Even assuming that the alleged 250 million protesters of 8 January were packed at mosh pit density, they would have covered barely 2.25 sq. miles. Suddenly, the protests feel more like a congregation than a revolution.
Terms like revolution, civil society movement, leaderless protests, and uprisings can evoke quick but misleading images of upheaval and traumatic change. The 8 January protests were not a spontaneous groundswell of public anger, but a string of massive rallies organized by ten trade unions. I’d be astonished if the crowd estimate was based on anything but wishful thinking.
My checklist of cross-questions when people share crowd numbers are—who counted, using what method? What was the other side’s estimate? Is the source cited? Was a VIP the source?
A newspaper headline after the 1978 cyclone in Andhra Pradesh read “Lakhs feared dead in AP cyclone". I was on the ground and the police station put total deaths at 59, as they only reported bodies they had seen. The person who feared lakhs dead was the chief minister, looking down from a helicopter.
Crowds do not change societies, individuals do. How many people would you say followed Mahatma Gandhi in his historic Salt March that sparked the freedom movement? The disappointing but perhaps inspiring, answer is… 79.
C.Y. Gopinath is a journalist, author, designer and cook who lives in Bangkok