Sanity Rally pt 2 - Attendance Size Doesn't Matter (But Mine's Bigger?)

Near the beginning of his October 31 Rally to Restore Sanity and/or Fear on the National Mall, Comedy Central host Jon Stewart made a few tongue-in-cheek remarks:

"As you all know, a Rally's success is judged only by two things: How Many people show up, and the diversity of those people."

He went on to explain that any gathering that was overwhelmingly white was, by assumption, racist; while any gathering that was overwhelmingly non-white must be demanding something that "we" didn't want to give them - voting rights, seats at a restaurant, etc.

He also jibed - "Early reports coming in place our attendance today at approximately 10 million... or was it 6 billion?"

The early points Stewart wanted to underline were:
1) The substance of a rally can actually be more important than the size of its audience.
2) 'Diversity' is a coded word for racial/ethnic mixing, when in fact other categorizations - age, profession, income - or even nuanced concerns like rational opinion - might be more important factors, even if they more difficult to immediately identify.

Fine and good. Point made, point taken. Stewart's Rally permit for the National Mall allowed for up to 60,000 people and no one is suggesting fewer than that showed up. Additionally, any 3 hour event broadcast live on national TV, with additional live coverage on multiple other channels (CNN, Fox, MSNBC, and others), as well as 4 millions on-line viewers doesn't need to make a big deal about how many supporters it has.

That hasn't stopped many - especially in the more left-leaning press organizations - to make a big deal out of how many people came to the Rally, nor has it stopped the Conservative opinionators at Fox News and elsewhere to either downplay the Rally's attendance, or conversely up-play the attendance at Glenn Beck's August 28 "Restoring Honor" rally that took place on the other end of the Mall.

Two rallies with two very different proponents, very different messages, and very different audiences. Beck, with support from half-term Governor and Fox co-pundit Sarah Palin, conducted a highly-charged, emotional, evangelical Christian call for the religious rebirth of Americans as a means to defeat the Woodrow Wilson-loving socialists who were, even now, plotting to destroy everything that once made America great. Stewart, along with farcical 'conservative' and real-life military supporter Stephen Colbert, lead a massive call to calmly ignore unreasonably charged partisan politics and return to an America of honest debate and practical compromise.

Does it even matter which one was better attended? Presumably, no one who went to see Beck also agrees with Stewart, and vice versa, so it's not a matter of conversion. And neither reached nearly the 156 million people required to technically possess a 'majority' of the American population.

Stewart says the numbers don't matter - so long as enough people willingly chose to participate to demonstrate that some respectable percentage of average, every day folks didn't appreciate being treated like idiots and constantly barraged by fear-based pseudo-journalism. Stewart's permit was for 60,000 and a showing of that size would have accomplished his task, especially in concert with the news coverage, TV ratings, and web viewership he enjoyed.

And I think he's right. The fact that more than 200,000 people came to his rally doesn't make his point any more valid than it would have been if only 60,000 came. And the same rule should be applied to Beck's rally. The fact that only 87,000 people showed up doesn't of itself make his points any less (or more) valid.

Unfortunately, that's not where the story ends. Liberals, and here we must especially point at Huffington Post, have taken the Sanity Rally attendance as not only a sign of the existence of the "Moderate Majority" that Stewart talks about, but specifically as a victory over Beck's assertion that there is a "Real America" majority of super-conservative fundamentalists. In the HuffPost's terminology, Sanity "Obliterated" Honor - in terms of attendance, and through inference, in substance.

I do not agree to this line of argument - again, Beck could be just as correct even if only 5 people showed up to his rally, and ditto for Stewart - but more than anything, I disagree with the factual warp that has come out - from both sides - following the rally.

Pro-Stewart (presumably pro-left, but such assuptions really should be stopped) point to a survey that said the Beck Rally drew 87,000 participants - a very impressive number up until October 29. But the same survey company and methods gave Stewart an attendance of 215,000 - which many have cited as being too low given the throngs of obvious Stewart supporters crowding downtown DC in failed efforts to even get close enough to be counted in the survey.

Firing back, the Pro-Beck camp insists that the 87,000 estimate is intentionally low, and that their own (undocumented, unscientific) census put their crowd at as many as 500,000.

National Park Rangers - Secretly Counting on Fingers and Toes?

Adding to this confusion are "unofficial" National Park Service estimates, leaked anonymously for each rally. According to Beck, he was told by someone 'on the inside' that his rally drew 300,000; while Viacom, the parent company of Comedy Central, announced that it was informally told its rally drew "more than 200,000."

But in both instances, the quotes were off the record and anonymous, so there is no way to fact-check them, check sources, publish data, or verify. In fact, the Park Service doesn't even keep counts any more - official or otherwise - because it has in the past been sued by people who felt its numbers were too low. So how would an individual Parks Service staff member arrive at either number? And how easy would it be - for either case - for a promoter to simply claim that they had been given a very high 'anonymous unofficial Park Service estimate' ?

In such instances - when a source is unverifiable or controversial - the solution is to simply check it against other sources.

Aerial Census: If It's Good Enough for Uncle Sam (and Uncle Moneybags)

Take, for example, AirphotosLive.com, a professional survey company that estimates crowd size using airborn photographs, manual counts and computer software to count people per sq ft and produce - what the Department of Homeland Security and several other corporate, government and non-government organizations deem - reliable crowd statistics. (other clients include the Associated Press, Boeing, Microsoft, National Geographic, T-Mobile, CNN, the Weather Channel, the Virginia Department of Transportation, and Verizon Wireless).

This is the firm that produced the 87,000 vs. 215,000 numbers, and while they admit to a relatively high 10% margin of error, it seems unlikely that they could have been off by the 70% margin Beck claims. Further, if they were off by such a wide margin, it would only be logical to assume they were equally short-counting the Stewart rally, which would then hypothetically reach 450,000 people.

Count not the Feet but the Footsteps?

Another nice, quantifiable number to consider is Metro ridership. Public transportation is not a highly-accurate barometer of rally attendance. The Metro goes to many non-rally destinations, and is used by many non-rally participants. Further, rally attendees could easily use other transportation - walk, bike, drive, taxi - that would not show up in such a count. The fact that both events took place on the Mall, certainly the most-visited spot for tourists on any weekend, only further confuses the numbers.

But let's use it as a relative scale. On August 28, 2010, Metro recorded 510,000 trips taken on its bus and rail systems. That's 180,000 more trips than the average Saturday traffic of 330,000; or we might think of it as 90,000 extra people (making round trips). True, many Beck supporters arrived on charter buses and were dropped off directly at the rally; and many more stayed in hotel rooms within walking distance of the rally itself. Still, this number is approximately in the ballpark of the AirLivePhoto estimate.

On October 30, 2010, Metro broke a 17-year old Saturday record with 825,000 trips recorded, or about 495,000 than the average. If we again assume round-trips, that puts rally attendance at about 247,000 - again, roughly in the ballpark of the AirLivePhoto numbers.

I attended both rallies, and rode Metro in both instances. In the case of August 28, Metro was certainly busy. My Greenline train - which Beck supporters told to avoid - was less than 1/4 full, but when I transferred to the Red line, my car was standing-room only, but still with some room to move around. Coming back from the Rally - again on the Red line - my car was even more full, with little room to move. But I was able to board the first train that came, and when I transferred back to the Green line, it was again mostly empty.

For the Stewart rally, I almost didn't fit into the first Greenline train to show up, and that was arriving at my stop at 9:40 am, more than 2 hours before the rally began. I was the last person to press into the toe-to-toe cars, leaving more than 10 people behind just at my door site. Each station we came to had more people who wouldn't fit in our train. Finally, we gave up, exited 2 stops early, and walked the remaining distance.

After the rally, we couldn't even fight out way into the first, second, and then third metro stations we passed. Lines clogged the escalators, ticket machines, and entry booths, and I can only imagine (sorry) how clogged the platforms and trains must have been. Ultimately, we decided to walk all the way home - 4 miles - because there were just too many people for Metro to handle.

As a comparative study, then, there is no question that the impact of the Stewart rally on DC Metro was overwhelmingly greater than the Beck rally, and both personal observation and hard numbers back this up. If Glenn Beck really had 300,000 people at his event - and Stewart "only" 200,000 - I cannot imagine where all of Beck's people parked their cars/ buses. Even at 50 seats/bus, that is over 30,000 buses at a minimum.

Area, Footprint, Density and - gasp - Math!

Let's consider one other measure that Beck supporters have intentionally warped to prove their side - footprint. The rallies took place in different locations, making direct 1:1 comparisons slightly tricky. As Pro-Beck bloggers have pointed out, the footprint of Beck's rally - stretching from the footsteps of the Lincoln Memorial down the Reflecting pool and into the WWII Memorial - is quite a bit longer (and they assert, larger) than that of the Stewart Rally, which went from 4th street back to the Smithsonian Castle, approximately at 10th street.

Using simple ovals, and photographs taken from the Washington Monument (which is about twice as close to Lincoln as it is to the Capital) they suggest that the Beck rally had a bigger footprint, and so must therefore have had more people.

But any good chemist can tell you that Density is an important factor when considering the mass of an object; and any high school geometry teacher will tell you that the area of a 2-dimensional object requires consideration of both its length and width. So let's consider those.

Here is a graphic showing the approximate footprint of both events, made on a single map of the National Mall to demonstrate the shared scale, and outlined based on my personal observations from 3 hours spent moving around each event.


What becomes clear from this picture is that, in fact, the total area of the Stewart rally was larger. Using Google Maps (specific tool I used is here), we can see that the Beck rally took up approximately 32 acres (the giant reflecting pool, and the large fountains in the middle of the WWII memorial have to be excluded, as no one was standing in them). However the Stewart rally - which didn't have any water hazards or monuments in the way - includes 48 acres of flat, accessible land.

But again, footprint area - even when calculated properly - only tells part of the story. What struck me most at the Beck rally was how sparse it was, especially on the large lawn immediately south of the pool. True - the area directly in front of the stage, and the first 10-20 yards all around the pool remained 'very dense' - that is, people standing shoulder-to-shoulder. But behind them were crowds laid out on picnic blankets and lawn chairs, able to see the jumbo screens even while seated on the ground. In these, 'moderately' dense zones, it was easy to move between blankets, and people had enough room to setup large umbrellas, ice coolers, camp chairs, etc. Further back, it was "very light," by which I mean the same type of setup as in the moderate zones - lawn chairs, blankets, etc. - but now they were spaced several feet, often as much as 10 ft, away from each other. This was more like a cool day at central park, than an activist rally.

Again, using rough math, I would say about 10% of the Beck rally was Dense, 60% was moderate, and another 30% was lightly populated. 'Moderate' was certainly the biggest category.

A similiar gradation occurred at the Stewart rally, except the proportions were radically different. In a crush of humanity, maybe as much as 75% was densely populated, with a very small moderate zone (10%) around the close fringe, and a 15% light zone at the very far rear (where, I am told, the rally couldn't even be heard).

For the sake of adding some ridiculous numbers to my bad science, let's take a cue from Professor G. Keith Still, PhD. Let's assume that a 'shoulder-to-shoulder' crowd of only 3 people per square meter, a relatively modest assertion on Dr. Still's range of up to 7. Using his measures, we would assign a value of 1 person per sq meter to a 'moderate' density, and some smaller number - 0.5 seems to high, but let's use it be safe - for a 'light' density.

Applying this formula to the total areas I block out, and the rough percentages I assign, we get the following (see table 1).

First, note that the area calculations themselves are decently accurate, while the % of relative density categories are much less so. Applying my horrible observation estimates equally to both, I put the Beck rally at 132,000 people, and the Stewart rally at 441,000 people (table 2).

I do not believe this is how many people were at either event, but it again underscores the incredible discrepancy between the two. I next tried to correct for my errors by using the density and area figures (the two stats I do trust) to re-make my % estimates. I intentionally skewed from my observations to see what % of each category would be necessary to match the posted attendance figures (table 3)

In order to get the Stewart rally to approx 215,000 people, I had to assume it had only 15% dense, 50% moderate, and 35% light. Pro-Beck supporters would say this shows how far off my left-wing bias has driven my observations. But wait! In order to get the Glenn Beck figures to match that statistical 87,000 number in the area of the rally, I had to assume 5% dense, 10% moderate, and a whopping 85% light density of participants.

Now, what about that mythical 300,000+ person attendance that Glenn Beck continues to claim he had? In order to get that many people into the actual footprint of his rally - and assuming the same density figures I'm using for Stewart - Beck would have needed at least 70% of his rally to fit my definition of 'very dense' and only 5% as 'light.'

Unless I was temporarily blinded on both August 28, and October 31, such wide divergences from observation are not possible.

(note: you can download my tables and re-scale them as you see fit. I'd be interested to see your results in the comments)

Summary

Again, I will allow that my methods are on weak scientific ground, but even the vaguest personal observation combined with any statistical framework confirms two things:

1. There is no way the Glenn Beck rally included 300,000 people. In fact, all evidence points to it being in the range of 100,000.
2. Jon Stewart's rally was absolutely larger, and in fact all evidence suggests it was considerably larger, probably in the realm of 200,000+

Weber

::(lame)Texpatriot