Close call

Don't blame competition-crazed network execs for the double-clutch in Florida. Blame those botched ballots instead.

by Dan Kennedy

MONDAY, NOVEMBER 13 - In the immediate aftermath of Election Night's media meltdown in Florida -- Gore wins! No, it's too close to call! Bush wins! Whoops, it's too close to call again! -- the conventional wisdom was that the networks' exit pollsters had really screwed up. Certainly that was my first reaction.

But in light of what we know now, I wonder. In fact, given what happened in Palm Beach County, it's possible that the networks' first call -- that Al Gore had won -- was actually right on the money, but was undone by the infamous butterfly ballots. Taken to its logical conclusion, it also seems possible that the second, erroneous call in George W. Bush's favor stemmed from the same set of circumstances.

MORE ELECTION COVERAGE

The Phoenix's Election 2000

It's not hard to see how the exit pollsters could have been led astray, especially in the case of the first, pro-Gore call. The cause could have been direct (Palm Beach County voters telling exit pollsters they had voted for Gore when, in reality, they had either mistakenly voted for Pat Buchanan or marked their ballots for both Buchanan and Gore); indirect (voters elsewhere in Florida giving pollsters information that was consistent with a solid Gore win, assuming Palm Beach County voted as expected); or some combination of the two. When the anticipated numbers for Gore coming out of Palm Beach County failed to materialize, all bets were off.

So how could the networks then have erroneously awarded Florida to Bush? Very simple. With Palm Beach County failing to give Gore the margin that had been anticipated, the networks assumed a significant Bush trend was under way. But there was no trend -- just a heaping pile of botched ballots.

According to news reports, 3407 votes were cast for Buchanan in Palm Beach County, despite a large population of Jewish retirees who presumably are not sympathetic to Buchanan's defense of Nazi war criminals. In addition, 19,000 ballots were tossed out because people had voted for two presidential candidates, mainly Gore and Buchanan. Now, let's award, say, 14,000 of the discarded 19,000 ballots to Gore, and 1000 of the Buchanan ballots to Gore as well. Am I being too generous? I'm probably being too conservative.

Under this scenario, Gore would have won 2,924,907 votes in Florida, to 2,910,195 for Bush. (At least based on numbers reported by WashingtonPost.com on Sunday evening. They continue to change by the hour.) That's a pretty narrow margin -- 49.07 percent to 48.83 percent -- but a landslide compared to the current 288-vote difference.

Now, I realize I'm making some fairly broad assumptions here. But if I'm right, were the networks actually correct in calling the election for Gore on Tuesday night? And do these numbers explain why they mistakenly called it for Bush on Wednesday morning? I don't know. I do know that election results are announced on the basis of exit polls all the time, and those calls are almost always right -- even when the numbers are incomplete and the final tally is close.

Yes, all the finger-wagging lessons of Election Night remain true. The networks shouldn't make a call unless they're absolutely sure. Exit-poll results shouldn't be announced when they could affect the outcome of the race. Blah, blah, blah.

Still, when this is all over and the post-mortems get under way, the media ought to consider the possibility that maybe -- just maybe -- they weren't as wrong in Florida as everybody thought.

Dan Kennedy can be reached at dkennedy[a]phx.com