Crowd Counts

For better or for worse, numbers evoke as well as inform. Beware.

Discussing the Obama campaign’s proclivity for announcing the crowd counts at its rallies, The Washington Post’s Dana Milbank today says that, for the campaign, size matters:

At virtually every stop the candidate makes in these closing days of the election, the campaign sends out an announcement with a boast about how really, really big Obama’s audience is.

Sunrise, Fla., Oct. 29: “A capacity crowd of 20,000.”

Norfolk, Va., Oct. 28: “22K: 11 in the stands/11 on the field.”

Fort Collins, Colo., Oct. 26: “45,000-50,000, with thousands still flooding in.”

Not only does Obama want you to know how huge his crowds are, but he also wants you to know his opponent, John McCain, has itty-bitty crowds. “University of New Mexico fire marshal Vince Leonard quotes approx 35,000 inside Senator Obama’s event tonight and at least another 10k-15k outside,” the campaign boasted Saturday night. “Senator McCain reportedly had less than 1,000 this morning.”

Milbank critiques the campaign for touting numbers that are sometimes on the sketchier side of an educated guess, and takes issue, in particular, with its use of crowd-count “validators”:

Of course, crowd estimating is a rather inexact science… To add legitimacy to the crowd boasts, therefore, the Obama campaign accompanies each measurement with the name of a “validator.”

“Total is 35k,” Obama campaign aide Ben Finkenbinder wrote in an e-mail to reporters on Wednesday night. “Validated by Danny McAvoy, Osceola County Fire Marshal.” He then provided Marshal McAvoy’s phone number.

Some validators are more valid than others. From Las Vegas came this questionable boast by the Obama campaign: “Bonanza HS Assistant Principal Zane Gray … says: ‘18k ppl.’ ” The “verifier” for Obama’s Norfolk event on Tuesday was the unofficial-sounding “Megan Mensick, Event Coordinator, Seven Venues.”

Granted, in his somewhat condescending assessment of the “validators,” Milbank is being just as inexact as the people he’s profiling. Who’s to say, after all, that the assistant principal at Bonanza High School isn’t an excellent crowd counter? Yet Milbank’s point reminds us that, while the idea of inflating crowd counts isn’t anything new, there’ve nonetheless been tales of fumbled or obviously exaggerated numbers during this election cycle. And rally accounts from today’s newspapers remind us that the figures included in such stories are as evocative as they are informative:

The Miami Herald states that Obama “drew standing-room-only crowds in heavily Republican Sarasota and in Orlando and Sunrise over the past two days,” including a “crowd of about 12,500 people” in Sarasota County.

The Richmond Times-Dispatch reports that Obama “drew more than 20,000 people to a rally Tuesday in Norfolk” and “spoke to roughly 10,000 supporters yesterday at the Verizon Wireless Virginia Beach Amphitheater.”

And the Pittsburgh Post-Gazette states that at a campaign stop this morning in Latrobe, Pennsylvania, Palin was “greeted by several thousand fans who started lining up around 6 a.m. outside a freezing hangar at Arnold Palmer Regional Airport.”

Sometimes, the evoke-and-inform mandate can be slightly more problematic. The LAT, for instance, makes blanket mention today of the record-setting “100,000 people” who “in a sign of their enthusiasm for Obama…showed up at his Oct. 18 rally at the riverfront arch in St. Louis.” (The lump-sum number of 100,000 made it into all the headlines two weeks ago: the AP report confirmed the number with the Obama campaign, while MSNBC’s First Read confirmed it with the St. Louis police. An LAT blog account cheekily but responsibly introduced it as “a figure a police official signed off on.”) In today’s LAT article, the loss of the word “estimated” and the suggestion that all 100,000 attendees were there “in a sign of their enthusiasm for Obama” evokes rather than informs.

Sometimes, when reporters run with campaign validators and their number estimates, the story suffers, and not just because of errant implications:

Covering a Biden rally in Florida, the Ocala Star-Banner described “a wildly cheering crowd estimated at 3,500 people at the Dancing Horses Farm, just west of Ocala,” adding that it was “Bob Walla, owner of Dancing Horses Farm” who made the estimate.

But Milbank deconstructs that estimate:

“Owner of Dancing Horses Farm, Bob Walla, is estimating crowd count at 4,000,” the campaign announced at a Biden stop in Ocala, Fla., on Tuesday. But Walla’s crowd inflation was easily exposed: A Washington Post reporter stood on a riser and counted 1,200.

It is, of course, unreasonable to demand crowd-count fact checks for all of these innumerable rallies (and there’s no guarantee that the Post reporter was wholly accurate). But the Obama campaign has made a big deal of its ability to draw crowds; it has, in fact, been a bit of an emotional selling point that Obama can so consistently congregate ridiculously large numbers of people. And press accounts that uncritically accept campaign-provided crowd numbers ultimately abet the campaign in selling that point. There’s a difference, when reading, between “a wildly cheering crowd” of “3,500 people” vs. one of 1,200. In other words, let’s use caution in employing unverifiable numbers in the service of colorful crowd descriptions.

Plus, there are always other ways that numbers can be evocative—in more verifiable ways. The Dallas Morning News, for instance, has a whimsically quantitative story about the availability (or lack thereof) of Obama-Biden and McCain-Palin signs in the North Texas area:

Staff members at Dallas County Democratic headquarters are fielding calls from disappointed people still seeking signs. The party went through more than 6,000 signs, even after it started asking people for donations to offset costs.

It uses quantitative details to illustrate a community’s excitement to visually show support for the candidates: “A voicemail on the Denton County Democrats’ phone says they are out of signs for this election,” while “in Collin County, Republican county chairman Fred Moses said the final order of 1,000 yard signs just arrived and should last through Election Day.” And it’s actually a pretty good way of accurately using numbers to tell a story.

Has America ever needed a media watchdog more than now? Help us by joining CJR today.

Jane Kim is a writer in New York.