Part three of a series evaluating the media’s performance during the 2004 campaign.
A month ago, we ran a satirical piece about the political implications of a giant meteor crashing into Kansas. Some of our readers didn’t get the joke — one e-mailer demanded this reporter be fired in part for his “unprofessional contempt for … the Asteroid Belt” — but most saw in the piece an effort to mock the phenomenon that the “-ism”-happy staff here at Campaign Desk likes to call “horseracism.” Horseracism is the practice of reporters and their editors obsessing over polls and process instead of substance, and it’s propagated by reporters whose hunger for inside information causes them to focus on the ceaseless torrent of minutia and meaningless numbers that pop up during a campaign.
Horseracism gave us a five-column banner headline in USA Today screaming that “Bush Leads By 8 Points!”; it led political junkies who hadn’t bothered to become versed in the candidates’ health care proposals to refresh regularly for the latest returns at NowChannel and Electorial-Vote.com; it fueled endless speculation about the existence of post-convention and post-debate bounces; and it allowed the carnival barkers who inhabit cable television’s daily shoutfests to fill airtime debating the likely impact of John Kerry’s mention of the sexual orientation of Dick Cheney’s daughter on “Sex in the City” Voters. With horseracism a media obsession in 2004, it wasn’t much of a stretch for us to imagine pundits butting heads over the potential bump in Kerry’s approval rating resulting from the obliteration by meteor of the sunflower state.
Matt Tiabbi nailed the problem in a pre-election rumination on the coverage of that most-obsessed-over swing state of Ohio:
In the thousands of “key battleground of Ohio” stories that have appeared in the press in the last few weeks, this is almost always the key angle. Who’s winning? And why? Did John Kerry’s goose-hunting gambit work? Is Teresa Heinz Kerry’s big f—king mouth a liability in Cuyahoga territory? Who, goshdarnit, is going to win Ohio?
Very occasionally in these stories, you will see references to enormous job losses in Ohio’s manufacturing sector, but these are always placed within a certain context: If the job-loss numbers are up, this is momentum for Kerry; if they’re not so bad, that’s momentum for Bush.
What is so amazing is that because of Ohio’s significance in the electoral college, we now have the whole country staring right at the very face of our vanishing manufacturing economy, and yet what we’re talking about is goose hunting, the widow of Chris Reeve and the “likability gap.”
Editors and news directors, of course, might simply reply with the old standby, We’re Just Giving The People What They Want. And there’s some truth to that. People like contests, after all — that’s why some newspapers devote as much coverage to sports as they do to politics. But that doesn’t excuse the manufactured horseracism that dominated political coverage during the campaign, which led to the continual hyping of numbers that were unreliable, misinterpreted, or just plain not that interesting. On July 12, for example, CNN’s Carlos Watson cited a Newsweek poll showing the Kerry/Edwards ticket ahead of Bush/Cheney 51 percent to 45 percent. Edwards had been named to the ticket a week before, and Watson argued that the poll showed “a real Edwards bounce” outside the margin of error, since there was a six-point gap. The poll’s stated margin of error was just four points.
The problem? Margins of error play off both the high number and the low number, so Watson had misinterpreted the poll; subtract four points from the Kerry number and add four to the Bush number, and you had a two-point Bush lead. (For more details — and a rundown of other culprits — see Tom Lang’s excellent post from August.)
It’s not just mathematical errors that lead horserace-happy reporters and editors to deceive readers. Many media outlets focus solely on the polls that they themselves commission, structuring analyses based almost exclusively on the information returned by their own pollsters. The defense for this practice is that consistent methodology produces more reliable returns. But while it’s true that all polls aren’t created equal, it’s absurd, with the avalanche of polls that came down daily during a campaign, for a prestige-hungry media outlet to operate as if its own poll exists in a vacuum.