Recounting the Media’s Florida Recount

The economics of newspapering dictate that any expensive investigative project must lead to a firm conclusion. So the media consortium (the New York Times, the Washington Post, the Associated Press, the Los Angeles Times, the Tribune Co. newspapers, the St. Petersburg Times, the Wall Street Journal, the Palm Beach Post, and CNN) that recounted the Florida presidential recount couldn’t very well undersell the statistical findings of the $900,000, nine-month study. It had to inflate them. And inflate them all the newspapers did, at least in their Page One headlines and ledes yesterday, assertively giving the election to George W. Bush under the most restrictive recount standards but to Al Gore under the less rigorous ones. All, that is, except for the Tribune Co., whose flagship newspaper, the Chicago Tribune, undersold the whole project with its wishy-washy headline, “Still Too Close To Call,” and wishy-washier still subhead, “Conclusion Not Clear Even If Recount Allowed.”Regrettably, the Tribune was unable to maintain the courage of its headline’s convictions and proceeded to game out the various recount outcomes, as did the other newspapers, alternately crowning Bush and Gore president. But underlying the Tribune’s coverage was its belief that it is probably “impossible to design a study that would determine who should have won the Florida balloting. That is particularly true given the degree to which the Florida election was tainted: Thousands of felons voted, people not registered were allowed to vote, others voted twice and even the dead voted in small numbers. Other voters were erroneously turned away from the polls.” But even if you assume that all ballots cast were legal and nobody was unjustly turned away from the polls and that the 6 million plus Florida ballots that weren’t contested were cast accurately, the Florida winner is still unknowable. This wisdom comes from Kirk Wolter of the National Opinion Research Center at the University of Chicago, the point person in assembling the data for the project, who is quoted by the Tribune and the AP (and also, only deeper, in stories by the Wall Street Journal, Washington Post, and the Los Angeles Times). The consortium diligently collected 175,010 overvotes and undervotes, but it concedes that it can’t be sure if it collected every overvote and undervote. If the margin of victory is just a couple of hundred votes, as it was in nearly all the media recounts, the contest as “too close to call,” Wolter says, because the variability of the vote counts—the possible error—would be larger than the margin of victory.”One could never know from this study alone who won the election,” he told the Tribune and other publications.That’s a fairly devastating critique, but to find it, readers had to be investigative reporters themselves or subscribers to the Chicago Tribune.