# How many people marched on Washington?

On 9/12/2009, several individuals gathered in Washington to protest recent governmental policies.  One of the things that has intrigued me about the “Tea Party march ” is the great disparity between the reports of how many people actually attended.  The mainstream media, for the most part, has used the phrase “tens of thousands” to describe the crowd.  Several organizations that would presumably support the march have reported figures of between 1.5 to 2 million.  I realize that if you add enough tens of thousands together, you could eventually reach millions, but the inference is that different sides of a debate are cooking the numbers to support their viewpoint.  (Imagine if McDonald’s claimed to have sold dozens of hamburgers.  Technically, that’s correct, but is hardly a good marketing device.)

So how many actually attended?  Typically, aeriel or satellite photos are used to determine the area covered by the crowd.  Samples are taken of the density of the crowd at various locations, and those values are used to calculate the estimate.  Problems with this method include the changing density of the crowd as people arrive and leave,  differences in crowd density at different locations, and the time at which the pictures are taken.

The problem with reporting crowd figures is that those reporting the figures often aren’t objective.  One side wants to say, “there weren’t that many in attendance, so people don’t really support the protest”.  The other side wants to say, “Look at the overwhelming response!”  Even police estimates might be a bit self-serving in order to justify budget items like overtime or extra personnel needed.

One solution would be to have two groups of researchers to evaluate the problem.  One group would simply take photos to pass along to the second group.  Along with the photos of the actual event, the second group would also receive photos from previous events taken at the same location.  When the second group is evaluating the pictures, they won’t know which photos are of the actual event.  Thus, they should be more objective when evaluating.

This method obviously has some weak points, but would seem to add a little objectivity to the problem.  Human nature being what it is, however, makes me think this is a problem that won’t soon go away.

The following is a time lapse video of the march.  How many do you think attended?  If you have an estimate or a better method of estimating, please let me know.