Net Election

The Web’s Exit-Poll Strategy 

{{Industy Standard Gif#34651}} Slate and the Industry Standard join forces to examine the effect of the Internet on Campaign 2000.

{{Slate’s Political Roundup#73099}} 

You wouldn’t know it from watching television, but the Internet has already shaken up this year’s presidential race. It’s not fund raising (though that’s certainly been goosed by the Net), and it’s not Net voting (still in a testing phase). Rather, it’s the widespread availability of exit-poll data—which shows who’s winning and who’s losing—while voters are still voting. Already, a storm of heated debate and threatened lawsuits has erupted over the data’s release.

The storm reached a crescendo this week. Like Slate did in late February, National Review Online bowed to legal pressure and chose not to post the statistics before polls closed on Super Tuesday. Within hours, the Drudge Report leapt into the void, putting the numbers out there for all to see.

Exit polls are the Rosetta stone of modern political analysis: They provide the fullest picture of the motivations and demographic fault lines of the American electorate. They are also, when properly read, remarkably accurate barometers of who will win, which makes them of intense interest to political and media insiders on the afternoon of an important election.

Theoretically, any organization can take exit polls, but in recent years, a single company—Voter News Service—has provided the service for the American media. The company’s board of managers includes the three major networks, Fox News, CNN, and the Associated Press; its subscribers include major newspapers such as the New York Times and the Washington Post. (Published reports indicate that there are more than a hundred VNS subscribers. A VNS spokeswoman declined to say how many subscribers the organization has or what subscriptions cost.)

Typically, VNS makes data available under embargo to subscribers in several waves, starting in the early afternoon on an important election day. The early numbers allow media organizations to plan their next-day coverage (by interviewing members of key demographic groups, for example). But the numbers are a prized commodity among interested parties.

To say that VNS is protective of its information monopoly is an understatement. (It’s even protective of the phrase “exit poll.” VNS has reserved the domain name but does not produce a site there.) So are its members: Fox News President Roger Ailes essentially has threatened to fire any employee caught leaking, or even characterizing, VNS numbers to outsiders.

It was not terribly surprising, then, that right after the Feb. 1 New Hampshire primary, VNS attorneys began demanding that Slate Deputy Editor Jack Shafer stop publishing the exit-poll information that other journalists leaked to him. Slate continued to publish the information for the South Carolina and Michigan primaries, but by the Virginia primary Feb. 29, Slate’s editorial staff succumbed to the dictates of its attorneys and withheld the data. (See Shafer’s “Press Box” columns on exit polls, here, here, here, here, and here.)

At that point, National Review got into the game. The conservative magazine’s Web site published the Virginia data on the afternoon of Feb. 29 and let it be known that it planned to publish the numbers for 13 contests on Super Tuesday, March 7. On March 6, however, National Review received a letter from Robert Penchina, an attorney representing VNS, demanding that the site “immediately cease any and all conduct misappropriating VNS’ property or infringing VNS’ rights.”

After considerable internal discussion, National Review decided not to publish the numbers again. National Review Web editor Jonah Goldberg said he was “kind of bummed” about the magazine’s choice, which he described as “essentially a business decision.” The Review simply did not have the resources for a legal fight. “Journalistically, we think we’re in the right,” Goldberg says. “But we’re not an operation that can handle even a frivolous lawsuit.”

It’s disturbing that a threatened lawsuit can keep two established publications from publishing information—especially when a conglomerate of media companies is making the threats. But that’s where a very useful Matt Drudge came in. At about 1 p.m. PT March 7, the Drudge Report site posted a headline declaring Bush the winner in nine out of the 12 contests; it followed up a few minutes later with a state-by-state breakdown of numbers.

It’s not clear if VNS attorneys have tried to crack down on Drudge. Drudge did not respond to requests for an interview, and VNS attorney Penchina said his client will not allow him to discuss the question. Given Drudge’s past practice, he might well ignore a cease-and-desist letter.

But if the loss of power is new to the entrenched media institutions, the pinball-bouncing of exit-poll data from site to site is already a classic maneuver on the Internet. And it’s only likely to build. The more exit-poll data that is published, the more people who have access to the information seem to want to leak it. As Shafer memorably says of his sources, “When you start hanging around with alcoholics, it becomes easier and easier to find a drink.”

And, indeed, a number of other sites have made use of the material, thus far without consequence. Political Insider sent out e-mail to its subscribers at about 2:30 PT on Super Tuesday that didn’t give vote percentages but listed the states where Bush and McCain were expected to win. Taegan Goddard, co-publisher of Political Insider, said Wednesday that he has not been contacted by VNS lawyers, “which just proves how silly their strategy really is in the Internet age.” He estimates that there are hundreds of sites that would be interested in publishing the data.

Others have suggested a Web-chase strategy not unlike those used by gambling and pornography sites: An anonymous e-mail server could be used to tell interested parties where the exit poll data is published. As long as it’s never the same site twice, VNS attorneys might well be powerless.

Moreover, it’s far from certain that VNS has a solid legal case. VNS invokes the “hot news” doctrine, a legal concept dating back to a 1918 Supreme Court decision, which happens to be a shaky leg to stand on. The argument that a company owns facts—as opposed to, say, entire databases or entire stories—has a mixed legal history. In the mid-’90s, the National Basketball Association sued Motorola, which was offering a score-update service via beepers. Although the NBA’s “hot news” argument prevailed in the suit’s first legal round, the NBA’s argument that it owned the rights to the score of a basketball game in progress was ultimately, and explicitly, rejected.

The analogy is not precise: VNS creates its data in a more meaningful way than the NBA “creates” a score. And Penchina argues that the NBA case acknowledged a tailored protection for “hot news” in the information age. But the First Amendment protection for journalists is also stronger in this case. Without active information theft, many courts have been understandably reluctant to bar journalists from publishing data.

It’s true that VNS’s commercial interests are violated when a Web site publishes the exit-poll information. The company spends millions of dollars to gather electoral data, and it understandably wants to control the release timing. One could argue just as plausibly that journalists can’t publish unreleased data about, say, General Electric’s finances because the company has a proprietary interest in it. Alas: Good journalists disrupt the desired flow of information every day.