The problem with crowd estimates

Digtriad — “Dozens of people have gathered in downtown Greensboro to honor Trayvon Martin and his family.”

MyFox8 — “Dozens gathered in Greensboro to protest the acquittal of George Zimmerman on Sunday night, one of several such protests across the country.”

News & Record — “She joined more than 200 people — mostly black but many white, and all ages — in a protest for Trayvon Martin held Sunday night in response to the acquittal of George Zimmerman, the man who shot him.”

Was it dozens or hundreds? Most journalists I know hate crowd estimates. How do you actually count a moving crowd? When do you start counting? Crowds are fluid; people come and go all the time. And don’t you have reporting to do? Why are you spending your time counting?

The answer to the last question is that you don’t really count. You estimate and you go to police. Police, of course don’t have time to count either. They are experts at many things but estimating attendance isn’t usually one of them. So they eyeball and guess. Just like you. Except you can quote police.

Editors demand crowd estimates because readers and viewers want to know how many people were there. They lend perspective to how newsworthy such a gathering is. Naturally, large numbers build credibility (unless you’re Republican legislators in Raleigh.)

Kelly Poe, who covered the rally for the newspaper, explained on Twitter how she got “more than 200.” (I’ve cleaned up some of the Twitter shorthand she used.)

“The protest split within the first 1/2 hour, most going to the jail. Probably 150 in the march, 50 left chanting (at the) original site. I put them together…. Notably, the protest grew in size and picked up many (people) on the march. I went (with the end number). TV might gone (with the) beginning or just what was left @ original site. I saw them there post-march.”

Makes sense to me. And Kelly is gracious enough to suggest that both estimates were correct.

But how is the reader or viewer to know? My guess is that the TV stations were called by people at the protest who objected to the almost dismissive term “dozens.” Whenever crowd estimates are given, the news media invariably hears from boosters saying the estimate was low, and often from opponents who say the number was inflated.

A case in point is today’s story in the News & Observer about yesterday’s Moral Monday rally. “Barber told one of the largest crowds yet – police estimated 2,000, organizers put it closer to 5,000 (italics mine) – that he came back for another reason, too.”

Big difference between 2,000 and 5,000.

Best solution: Provide photos that let people decide for themselves. Here is the N&R’s gallery of the Sunday rally. Depending on the photo selected, the rally was either sparsely attended or huge.

6 thoughts on “The problem with crowd estimates

  1. When I was wire editor at the N&O and pitched a protest story for the front page, the first question from other editors was always: “How many people were there?”

  2. I’m semi-recalling a journalism-review debunking of parade crowd estimates that actually counted heads from aerial photo grids….

  3. And then there’s the old “evangelistically speaking” where the estimate is rounded up to the nearest thousand, ten thousand, one hundred thousand or million, depending on usefulness to the evangelist. 501 becomes 1,000, and so forth. You can always count on preachers for accuracy on estimating the crowd.

  4. John, it’s ridiculous that journalists don’t see crowd estimation as part of their job of collecting facts. There are relatively reliable techniques for estimating crowd size that go back decades and reporters should be using them; relying on estimates from partisans is nonsense. Even early on during the Moral Monday protests, when it was incredibly easy to count the crowd as it entered the legislative building 2×2 (I know because I did it when the crowds were ~400), no reporters bothered to get an easy count. Use interns if you have to, but training reporters to estimate crowd size should be part of the job.

  5. A couple of links I’m adding here so the previous comment doesn’t get blocked as spam; estimating crowds isn’t simple, and always involves margins of error, but it can and should be done by journalists.
    The Curious Science of Counting a Crowd:
    “But some fairly simple math can be used to make defensible estimates of crowd sizes”:
    Tips on estimating crowd sizes for journalists:

Comments are closed.