Were the Exit Polls Juiced?

Commentary by Greg Lewis / OpinionEditorials.com
November 10, 2004

It occurred to me very early on election night — as Brit Hume and the Fox News Channel analytical team expressed befuddlement at the fact that afternoon exit polls (overwhelmingly reliable in the past, as Dick Morris would attest) seemed to point to a stunning Kerry victory, while actual election returns, which began to be released shortly after 7 p.m., made it clear that Bush was at least holding his own and very probably doing quite well, thank you — that those exit polls might have been "juiced," to use the term that has become current to describe athletes who cheat by using substances such as steroids to enhance their performance.

Indeed, the day after the election former-Clinton-operative-turned-adversary Dick Morris asserted unequivocally that the exit polls were juiced. More specifically, Morris said that in all his years in politics, he'd never seen exit polls that were so wrong as this year's crop. He claimed that exit polls are by far the most accurate types of polls and that in order for dozens of them to be wrong, someone had to be intentionally skewing the numbers, either by juicing the samples (the early afternoon polls were weighted, for example, about 60-40 in favor of women, which favored Kerry and which was either a stupid mistake or a deliberate attempt by pollsters to distort results) or by the Democrats' flooding the polls with voters in locations where and at times when they knew the exit polls were being taken, to cite two ways the alleged juicing might have happened. Far from ruling out network collusion (the exit polling data was gathered on behalf of participating news organizations), Morris mentioned the need for a congressional investigation.

It is not without the realm of possibility that the skewing of early exit poll results was done intentionally to convince people, particularly those who had not yet voted, that Kerry was on his way to a landslide victory, and thus to dampen Republican enthusiasm and turnout in the mountain west and far western states. While I don't think that intentional misrepresentation can be conclusively proven, I do think (or rather I did think for about a day and a half) that there was the distinct possibility that the exit polls which produced such dramatically inaccurate results were in some way juiced.

The argument against intentional juicing of exit polls actually rests on the amateurish and unsophisticated behavior of many internet bloggers. Upon fairly close examination, their (the bloggers') role in the matter of the perpetuation of phony exit poll numbers is pretty deplorable, especially that of one "Wonkette" — whoever the hell she is — who was chanting a kind of perverse mantra, "run free, information," as if it were her duty to make sure that nothing, even the most patently false info, was prevented from getting put out there on the web for everyone to see. The fact that many bloggers — without even bothering to take into consideration such Statistics-101 principles as the margin of error implicit in a set of several dozen small local samples which added up to no more than 12,000 voters nationally — should, like high school sophomores breathlessly passing along the latest gossip, "inform" their readers (read "groupies") that Kerry was possibly amassing landslide victory numbers would hardly be worth comment if it hadn't caused such a misguided stir among so many people.

The blogs' role in the (happily) temporary perpetuation of such blatantly inaccurate information was puerile, self-involved romanticism at its absolute worst. In fact, the post-election drivel I listened to on Air America was no worse than this horsecrap. Even Glen Reynolds (of Instapundit.com) sounded a lot more like a drunken college underclassman at 3 a.m. than he did a lawyer. On balance, I guess that's not that much of a stretch, is it?
In the face of this irresponsible reportage, Bush-Cheney campaign manager Ken Mehlman has testified that the Republican assessment of the bogus information at the time was along the lines of, "You know, what we're seeing in these exit polls is just so far from what we know has to be the case among voters, that we simply can't give it any credence." Mehlman and his associates were quick to recognize that the early exit poll results were heavily skewed toward women voters and that they came from predominantly Democrat precincts, both of which would tend to argue that Kerry's position was dramatically over-represented.

At any rate, in this case it was actually the networks who put the brake on the exit poll data. For once it was the big guys who acted responsibly. The bloggers' role in this whole thing ultimately argues against a widespread conspiracy to influence voters to believe that Kerry was winning. The exit-pollers were, at best, utterly incompetent. In this case, however, it was the blogosphere in which the results of their incompetence got broadcast, and not the mainstream media.