Clash over exit polls pits new vs. old media
This column appeared March 6, 2000, in the Online Journalism Review. Here’s the version on the OJR site.
By J.D. Lasica
Slate raised a ruckus early in the primary season by publishing the results of exit polls in New Hampshire, South Carolina and Michigan hours before the ballot boxes closed.
When Slate stopped the practice last week under threat of legal action, the National Review stepped into the breach, publishing exit poll data from Virginia while voters were still casting ballots. They plan to do the same during tomorrow’s Super Tuesday primaries in New York, Ohio and California. So does Matt Drudge.
Good for them.
Once again we have a culture clash between old and new media, with both sides talking past each other and the lawyers now getting into the act. It’s as if the traditions of both mediums have become so entrenched that neither party can understand the needs or prerogatives of the other.
What this really comes down to — and what nobody has said so far — is that online journalism and broadcast journalism serve different masters and are, at bottom, two very different mediums. (More about that in a moment.) While I’ve worked extensively in both old and new media, my sympathies in this case rest with the Net crowd. Let me tell you how I got here.
In one corner we find Jack Shafer, managing editor of Slate, who has single-handedly tried to quash what has over the years become a private love fest between journalists and political insiders. “We all know this exit poll data zips around on telephones and in e-mails between journalists, politicians and other members of the elite,” he says. “Why not throw the doors open and invite the public to the party as well?”
The group organizing the party, Voter News Service, has tried to keep it a small, private affair. The VNS consortium, which comprises the news divisions of ABC, CBS, CNN, Fox, NBC and the Associated Press, hires polling firms to sample the electorate on each election day, dissecting voters by age, race, income level, political bent and so on. Slicing and dicing the electorate like a sous chef, the pollsters and pundits serve up a dish of election analysis that seeks to answer not only who and what but why — why a candidate is succeeding among certain segments of the public.
But there’s a second aspect of exit polling, one that involves not issues or demographics but the horse race itself — who’s ahead and who’s trailing. More complex and problematic than straightforward exit poll numbers, these “decision records” are the figures that journalists often leak to outsiders, Shafer says, in a “titillating game of vanity, status and looky-look.” To stamp out the practice, Shafer asked people to e-mail Slate the figures and then posted the totals online.
Shafer harbors little patience for news organizations that “play nanny” by shielding the public from real-time election information. “I think the voters aren’t that fragile. There’s a real elitism and paternalism here that’s aggravating to me,” he says. “It’s the job of the political parties, the candidates and the League of Women Voters to get out the vote; it’s not the job of journalists. It’s our job to find the truth and publish it, not to promote democracy.”
Slate will no longer run projected election returns because VNS has threatened to sue. “We’re declaring victory and moving on,” Shafer says. “We’ve made our point about the ludicrousness of these major news organizations trying to censor us.”
For its part, VNS refers all press queries to its member organizations. Evans Witt, who analyzes exit poll data as a consultant to NBC News, seems at once mystified and incensed by Slate’s actions. “Journalists often have information they don’t share with public for embargo reasons,” he says. “The current system was designed to allow journalists to prepare a report after taking some time to digest and thoughtfully consider the results of an election. It’s terrible journalism to release raw, partial data that may be wrong or misunderstood — especially if it could discourage people from voting.”
Witt, who served as executive director of VNS in 1997 and is now president of Princeton Survey Research, calls Shafer’s actions “a mindless ego trip.” He concedes that VNS rules prohibit journalists from sharing early exit poll data with outsiders, but had no suggestion on how to staunch the torrent of leaks that result from both VNS members and some 150 other news organizations, including most major newspapers, that also have access to election data. “I just hope that good sense would prevail,” he says, none too persuasively.
At CNN Interactive, election director Carin Dessauer calls exit polling “an indispensable tool that helps us pursue certain angles of the story and provide better coverage. Primarily, it provides valuable information to users on our site when the polls close.” CNN maintains a strict policy of not calling a race until the polls close in that state, she says.
Dessauer says “a very small group of individuals” at CNN see exit polling data during the day. Decision numbers that show partial vote projections “are even more closely held. We make it clear to our teams that this information be closely held, and we take great precautions within the broader team of CNN so those numbers don’t get out.”
An online editor at a news organization affiliated with VNS says this about those pesky leaks: “Numbers can always get out, but this is the first time anyone has ever tried to report and publish them. Just because the numbers are out there doesn’t mean a wider audience should have access to them. It has nothing to do with elitism. It has everything to do with our being able to do our jobs.”
A contrary view
Nonsense, says Rich Lowry, editor of the National Review. “This information is all over Washington and the campaign staffs. Our attitude is, we’re not just going to tell our friends and confidants, we’re going to share this news with our readers, too.”
And so they did, starting at 3 p.m. Feb. 29 when the conservative magazine’s Web site showed George W. Bush with a 12-point lead over John McCain in the Virginia primary (he eventually won by 9 percent). The online publication, which got a “huge bump in traffic” during the day, causing its servers to crash, plans to do the same for the major primaries March 7, Lowry says. (Media sycophant Drudge linked to both Slate and the National Review rather than relying on his own sources.)
“I think it’s legitimate news,” Lowry says. “We’re reporting on things that people tell us — that’s what journalists do. I think it’s ridiculous that these goo-goo good government types don’t trust the public with certain information. If knowing who’s ahead deters some people from voting, I’m not sure they should be voting in the first place.”
So there you have it. Slate and the National Review suggest that journalists have no stake in democracy. The establishment media suggest that they know what’s best for the public.
Both sides deserve a good whuppin’.
Online: The proactive medium
I admire Shafer and Lowry’s forthrightness, but their public-be-damned arguments smack of the arrogance has long been a hallmark of old media — and a perennial curse on the house of journalism. Over the years there have been countless accounts of citizens fuming at the news media for declaring an election over before they’ve had a chance to vote. It’s the height of hubris to pretend that these voters and their concerns should be brushed aside.
On the flip side, the establishment media and other guardians of the status quo still don’t get it. Thousands of news executives, journalists, pollsters, politicians, campaign aides and others all have access to timely, newsworthy information embargoed from the public. That’s wrong.
I propose a middle course: Keep the early numbers off the airwaves, but let Web sites publish them on their politics pages. (Large sites that have the reach of a broadcast or cable network, like Yahoo or AOL, could link to an interior story.) That’s largely the Slate-National Review approach — without the dissin’-democracy attitude.
One might choose to argue this issue on narrow terms: Are these Web publications violating a contract? Wrong question. Let’s leave that question to the lawyers. (We’ll note for the record, however, that journalism would be in a sorry state if we were precluded from reporting information that a corporation or government claimed was proprietary, confidential or otherwise secret.)
A better question is: What serves the public’s interest? Clearly, it’s not in the public interest when broadcasters reveal real-time results while an election is in progress. But the Web is not a broadcast medium. Online users don’t passively receive information — they actively seek it out, and elections are especially well-suited to a medium that lets users tease out information that’s specific and immediate. The Internet’s philosophical girdings and architectural framework both play to our culture’s want-it-all, want-it-now mindset. Love it or hate it, that’s the way of the Web.
Yes, newspapers and TV newscasts ought to provide thoughtful analysis and synthesis rather than giving in to the online medium’s appetite for immediacy at any cost. But they also need to recognize that they no longer serve as choke points controlling the flow of information. Anyone with a Web site can be a news provider.
That’s a healthy thing for journalism — and for democracy.