Should Voter Data Be Released?-- May/June 2000

May/June 2000 | Contents

...Should Voter Data Be Released? New Media, Old Media Disagree
By James Ledbetter

The 2000 presidential primary season was short, but not without its sparks. Political observers marveled at sometimes nasty exchanges between the four front-running candidates, but an equally contentious battle erupted between Old Media and New Media over the topic of exit polls.

It began with one Web site, Slate, posting exit poll results of the New Hampshire primary several hours before voting booths closed in that state. By March, several Web sites were involved -- at least two of which were threatened with lawsuits from Voter News Service (VNS), the media consortium that conducts exit poll research. And while the storm is temporarily over, November's general election is almost certain to revive it.

The exit poll controversy represents the most important journalistic impact that the Internet has had so far this year, and one of the most significant since the Internet came to mass prominence. It is, for starters, a battle that only the Internet could make possible. Yes, radio and television both have an immediacy of delivery like the Internet's. But no single radio station has nationwide reach, and most of the national broadcast and cable television news outlets are signatories to VNS.

Moreover, the battle neatly splits participants along technological, rather than ideological, lines. Slate is a generally liberal venue, but its chief allies in putting out the exit poll data were the National Review's Web site and the Drudge Report -- both of which are usually considered right of center. Arrayed against them are the old-line media that make up VNS: ABC, CBS, CNN, Fox, NBC, and the larger newspapers.

The assault on proprietary exit poll numbers is actually a longstanding mini-crusade for Slate editor Michael Kinsley. Back in the early '90s, Kinsley began attacking the major networks for what he considered fraud in their election night coverage: the anchors and reporters all know a set of reasonably predictive data, but won't share them with viewers. In 1994, when Kinsley co-hosted CNN's Crossfire, he accused the major television networks of "self-censorship" when they sat on exit poll leanings for that year's crucial congressional elections.

The legal issues surrounding the publication of VNS's data are complex and highly contested, but essentially VNS and its attorney have used several variations on the argument that the exit poll data are its exclusive property. Those who published them early disagree passionately. Jack Shafer, the Slate columnist who posted the VNS figures in three primaries this year, says: "They tried to assert copyright protection, but these are facts, and you can't copyright facts."

Absent a lawsuit, the legal issues will not be resolved. But more than a few journalists believe that there are ethical and civic obligations that might be trampled by the practice of publishing exit poll numbers early. As Lawrence Grossman discusses elsewhere in this issue of cjr, network executives and members of Congress have for years expressed concern that if television disclosed exit survey numbers before West Coast polls closed, it could make millions of votes irrelevant and discourage westerners from voting at all.

Keeping people from voting is not something that journalists generally want to do. So how do those who published the VNS numbers respond to the charge that they're harming democracy?

"I haven't seen any proof that it deters voter turnout," says Jonah Goldberg, who edits National Review Online, which published the results of the Virginia primary in late February, and was promptly warned by VNS lawyers not to do it again. "I don't dismiss the ethical argument. I just don't think much of the ethics of most journalists, so I need more than their assurances. I need facts."

Goldberg has a powerful point: elections don't have control groups, so there's no definitive way to prove that releasing the numbers reduces voter turnout. And then there's the result from this year: in almost every single important primary in 2000, the results of the exit polls were available on the Internet hours before the polls closed. And yet turnout in nearly every contest was up from previous years.

The provocative Shafer takes the argument a step further. "It's not my job as a journalist to help or hinder the electoral process," he argues. "My job's to convey information when I have it. Besides, if that principle is important, what about tracking polls? Every network broadcasts tracking polls right up until the day before the election. Studies show that tracking polls deter turnout, so do we toss those out, too?"

There's another ethical question: VNS spends millions of dollars to collect exit poll data for its paying clients. (Exactly how much is a bit of a mystery; for a consortium of media players, VNS displays surprisingly little interest in talking to the media. The spokeswoman for the New York-based group refuses to release even fundamental information about VNS's work, such as the number of paying clients or the cost of polling.) The exit information is private property, and publishing it against the wishes of its owners amounts to a kind of theft.

Shafer has zero patience for this argument, pointing out that the very people who leak exit polls to him and his Web counterparts are employees of VNS partners. "There's an old Chicago newspaper tradition: if you don't want your competitors to know what you're doing, then keep your mouth shut," he says. "I'm not responsible for keeping those embargoes, they are."

And that's the power of Web journalism: it can take what is known among a small number of insiders and publish it almost instantaneously before anyone has a chance to react. Experienced political journalists point out that there is ultimately no way for VNS to enforce silence on all its participants. Some have argued that delaying the release of the numbers until late afternoon, or even a few minutes before poll-closing, would help contain leaks. But that would alienate VNS's print clients, who use the data to plan the next day's coverage.

The most porous elements of the whole exit-poll process are the candidates themselves. As one veteran political writer puts it: "The networks are always going to leak the numbers to campaign officials; TV people trade those figures for access. And the campaigns are always going to leak them to other reporters to get an early spin on the story."

And so, despite the best efforts of VNS attorneys, it's almost certain that one or more Web site or e-mail newsletter will get the VNS numbers for the general election and circulate them. "Politex," the pseudonymous editor who runs the site Bushwatch.com, says he considered publishing the numbers himself, and heard of others who intend to. And the National Review's Goldberg even has a prediction: "I would be amazed if Drudge didn't."

James Ledbetter is the new-media columnist for cjr. He is editor-in-chief of the European edition of The Industry Standard (JimL@thestandard.com), a magazine that covers the Internet industry. He formerly was media critic for The Village Voice and is the author of Made Possible By -- The Death of Public Broadcasting in the United States.