September 21, 2014

Once More on Counting Crowds at Demos

[UPDATE: This caused some controversy when it was first posted, so I wrote a shorter--and crankier--follow-up piece a day later, which has a few additional thoughts.]

What a splendid march!

Props first of all to the 100,000 plus people who came to NYC from around the US (hello, South Dakota Quakers!) and around the world to stand up against the carbon-burning—and not coincidentally, capitalist--economy that is destroying the habitability of the planet for an awful lot of the present biosphere, including humans. You tended a tad toward the white end of the spectrum to be sure and were perhaps a bit naive, but you were young, you were jazzed and you were mighty imaginative in your posters and costumes and slogans.

Props too to the organizers who turned out all these folks on a very tight time-line, who made excellent use of the Internet and social media to build the protest, and who organized a very smoothly run march.

But let's face facts, nobody is much interested my review of the People's Climate March. What you want to know from me is how many people were there. I will give you two answers:

1. There were well over 100,000 people, likely a bit upwards of 120,000 in the march.

2. No way in hell were there 310,000 people on that march.


How did we arrive at our first figure which was actually, before a fudge factor described below, 106,500?

Okay, let's start with we. We is me and labor journalist Peter Hogness. Both of us are experienced demo counters who think having an accurate estimate of the strength of our own forces is invaluable for organizers. We set up shop on a marble bench on a raised space in front of some ugly-ass skyscraper on 53rd Street and 6th Avenue, which put out heads well above the level of the march.

Because the Climate March was going to be too big to count directly (anything over 15 thousand, maybe 20 on the outside, becomes too hard to do that way), we sampled. That is, we counted marchers passing a chosen point for three minutes. (We don't count 1,2,3,4, of course, but in clumps, 5s, 10s, even 20s, losing exactness. but maintaining a grasp on large numbers). This let us figure the rate per minute, and note the time or start a stopwatch.. When the density of the crowd or the speed at which it was moving visibly changed, it was time to stop the watch, note the time elapsed and take another three-minute sample. If the march stopped in front of us, the clock was stopped.

The march started at 11:30 and reached our vantage point at about 12;02. During the first hour, it stopped and started repeatedly, leaving about 42 minutes of moving time. Repeated sampling gave results as high as 850/minute and as low as 600. Overall the average was, generously, 750/minute.

42 X 750 = 31500

For another 20 minutes the march ran at about 500/minute, but never stopped again.

20 X 500 = 10000

For the rest of the march, which ran without interruption for another two hours and forty minutes, samples bounced within a range of 300/minute and 500, with most around 400.. It was remarkable. We'd start a three minute sample when it seemed the march had thinned out to almost nothing and a big contingent of high school kids would roar through, chanting and bopping and knock the number up again. And vice versa. We figure that 400/minute was a good shot at a constantly changing target.

160 X 400 = 64000

There you have it: a total of 106,500. With smaller demos I generally allow a fudge factor of 10% to cover counting errors, people who left the march before the counting point or took another route to the end point, etc. Brother Hogness, I think, usually gives 15%. For a march of this magnitude we considered 20% a generous but not unreasonable figure. That would take the total to about 125,000.

It would not take it to 310,000.


So where does this number I keep harping on come from? That's a mighty good question. I first saw it in an email from Doug Henwood, attributing it to "the organizers."  Then it appeared at as official.

Now, I think that Bill McKibben has done amazing work with 350 org in forcing this crucial issue to the fore in this country. I've known Leslie Cagan, and many of her crew, and have worked with them for decades, and I rather doubt that anyone else could have pulled this off. The Avaaz people I know almost nothing about.

But I think that whoever is pumping this number is doing the movement a real disservice. I have written before about why I think inflated demo counts are such a bad thing, including this rather passionate short piece here at FotM. You may think that it's harmless, a "little white lie," but consider that this whole movement bases itself on the accuracy of numbers, numbers about global warming. It behooves us to be extra-careful about this.

So how do we know that the official figure is not correct? Well, for starters, we have confidence in our methodology and so do a lot of people who have tracked our earlier counts. But it is not hard to show that the 310,000 is impossible. I hope no one will challenge our statement that the march took 4 hours and a hair to get past our checkpoint. If that number had marched past us, they would have had to do so at a rate of 1300 a minute, non-stop for four hours. That's almost twice as fast as the fastest contingents we saw. 6th Avenue can comfortably accommodate 25 adults standing abreast, in a squeeze maybe 35. (We checked before the march came through). That would mean 30 rows a minute. The only people who can do that are troops, drill teams and maybe joggers in a race, not senior citizens, folks pushing strollers, marching bands (a score of them, and only one specializing in drumming on five gallon plastic compound buckets!), kids showing off the signs they made, giant puppets and the rest of the panoply which made the march so wonderful.

Friends asking about our count have raised the fact that many people (including them) who waited a long time to step off finally fell out and did not make it to our vantage point. That's what our 20% fudge factor is for, but maybe it's too low. Let's say one person bailed for every one who took more than two hours to reach our counting station. That would double the number of those who actually passed after 2:00, from 48,000 to 96,000. and that, that would only bring the total to 154,500, still less than half of what the organizers are evidently claiming.

Somebody is going to have to have to show me in some detail where that 310,000 came from, before I'll buy it. Meanwhile, I'm calling bullshit.

Whenever I write about this topic, I find myself quoting the great Guinean agronomist and martyred revolutionary, Amilcar Cabral.
Tell no lies. Claim no easy victories.


Unknown said...

How funny to read this & see my name. I think I tweeted the 310,000 figure from a text I got from officialdom. The NYT said a Carnegie-Mellon mathematician vetted.

Mark B. said...

The author seems to think that the 300,000 number is some kind of mistake, trivial but misdirected. It is not - it is fundamental to the entire effort. The lie is understood to be a lie by all involved, and is supported not as an unfortunate necessity, but as an Orwellian Truth. That's the kind of people we're dealing with here.

Anonymous said...

"tended toward the white end of the spectrum..."

How's that relevant?


Anonymous said...

Note that Polar Bears International says "nearly 400,000 people participated in the People’s Climate March in New York City"

Is that a deliberate exaggeration or do they really think that any number over 300,000 is "nearly 400,000"?

At this rate, by tomorrow the estimate will be 500,000.

ilma630 said...

You say "this whole movement bases itself on the accuracy of numbers, numbers about global warming". If it (the movement) were that worried about accuracy, then perhaps it should stop fudging/tampering/adjusting the temperature records and use the actual observed data. Perhaps it should also stop saying that the outputs of models are accurate or 'evidence' considering the input data is 'parametrization', i.e. guesses, or the output from other models, and the model code reflects the modeller's bias. Perhaps the marchers would actually look at the (untampered-with) observed data to form an accurate judgement rather than the political rhetoric of the organisers and vested interests (e.g. renewables industry). Perhaps the marchers should receive some history education so they can accurately compare recent with past climate, so they would realise that nothing of recent is unprecedented. Accuracy of numbers in climate?! Your march counting is probably the most accurate number in the whole of the climate arena. At least you call the organisers' BS out on it, and so we have that to be thankful for.

Nik said...

Climate “science” has now gone fully brazenly Enron level corrupt, as seen by exposure of a plot of all the input data of the latest Orwellian revision of climate history, a hockey stick media sensation shown to have no blade in any of the actual input data:

AzCountry said...

Jimmy Higgins:

An excellent presentation of your counting methodology. It seems very intuitive, relatively accurate with reasonable precision.

The important observation that the participants 'tended toward the white end the spectrum' addresses a number of issues. In the U.S., typically persons in the middle-to-lower social-economic levels can not afford to attend distant protests with increased probability that they are non-white. Second, at the world level, the vast majority of the population is non-white (i.e., 54% Asian, 15% black, 15% white, etc.) so the majority of the affected world is not present.

Lastly, I do not understand how 'climate change deniers' can in good faith allege that the vast majority of all scientists are either being hood-winked by a few, all powerful climate change advocates, or the vast majority of scientists are so dumb that they are being lead around like sheep. For those not trained in science, science is a method of open communications where your theory or data change be tested. Over time, your position gets stronger or weaker.

All the sciences, including climatology, ichthyology, geology, paleontology dendrochronology, ice core analysis, etc, clearly have a consensus that humanity is seriously affecting the world's climate via increased CO2 levels at its corresponding temperature variability impacts. Lastly, excluding pure math, science only increases the probability of what think is right (correct) is right (correct) - - science never proves anything!!!

Brad V