The Jon Stewart Rally to Restore Sanity was big by almost any measure. So big that some experts calculate it drew a bigger crowd than did Glenn Beck’s Restoring Honor rally in August. Is that true? How accurate are these attendance estimates, anyway?
The answer to the first question is yes, there are some indications that comedian Mr. Stewart drew more people to the National Mall than did talk-show host Beck. For instance, Washington’s Metro moved a record 825,437 people on Saturday, compared with an average Saturday ridership of about 350,000. That’s more than they moved on the day of Mr. Beck’s rally.
But such indications are just indirect evidence. And the answer to the second question above is that the science of crowd size estimation is far from exact, as we noted after Beck’s event.
Some “estimates” of Mall crowds may be just stone cold guesses. A number of media organizations, for instance, cited unnamed National Park Service officials who estimated the crowd at more than 200,000. That’s interesting – but the National Park Service does not do official estimates anymore, due to the controversy that always follows. We have no idea how this unnamed official came up with the 200,000 number. Were they flying overhead in a helicopter? Standing next to the podium, looking at the audience and guesstimating?
Canadian television, in contrast, broadcast a larger estimate of 250,000 for the Stewart crowd. But as far as we can tell, they have not said how they came up with this number. In the absence of some explanation of methodology, be suspicious. Even political pollsters usually provide their raw data and questions for study alongside their conclusions.
There are rigorous attempts to judge how many people are present in a space. But even these seem to be close cousins of “shedology.” That’s the old method in which Central Intelligence Agency analysts estimated Soviet weapons numbers by looking at overhead photos of military sheds and calculating how many tanks they might hold.
The most scientific such stab at judging the Stewart crowd appears to have been carried out by CBS News, which hired Digital Design at AirPhotosLive.com to do the work. They came up with a figure of 215,000.
Digital Design uses cameras on tethered balloons to take high-resolution photos of a crowd. Then they enlarge the photos and begin counting. They try to figure out the density of sample crowd areas. Then they extrapolate from there based on how far the crowd spread. You can read about how one of their estimators goes about the work here.
CBS used the same company to estimate the Beck crowd. Their figure then was 87,000. The organizers of the Restore Honor event criticized that number as too low, and even some mainstream media figures who covered it felt that crowd was closer to 200,000.
What’s the bottom line? As you see, the estimates for the Stewart rally are all over the place. The difference between the low end and the high end is at least 25 percent. Beck rally estimates had an even wider variance.
So we’d say that comparing the size of two numbers, which were both subject to large margins of error to begin with, will probably yield a result commensurate with what the comparer hoped would occur. In other words, you’ll see what you want to see.