In what may be the (belated) final act of the Florida miselection saga, a consortium of major media outlets anointed George W. Bush the beneficiary–though not necessarily the winner–of their ten-month, $900,000 joint review of 175,000 ballots that were not included in Katherine Harris's official count. Bush gained because most of the consortium's participants emphasized one slice of their review–the finding that had the US Supreme Court not infamously halted the recount ordered by the state Supreme Court, Bush still would have squeaked past Al Gore by more than 400 votes, instead of 537. The front-page headline in the Wall Street Journal, a consortium member, was typical: "In Election Review, Bush Wins Without Supreme Court Help." Yet the header on the inside page was a better summation of the ballot project: "Final Tally in Florida Yields Mixed Results."

The true story is that the Florida mess cannot be definitively sorted out. That's no news flash, since the Florida election exposed numerous problems in vote tallying and demonstrated that the system was not up to collating accurately 6 million ballots in a 50-50 race. As the consortium–which included the New York Times, the Los Angeles Times, the Washington Post, AP, CNN and others–reported, there were several ballot-counting scenarios that placed Gore in the narrowest of leads (between 42 and 171 votes). Each of these was based on reviews of untallied ballots statewide, including undervotes (ballots not registering a presidential preference when tabulated) and overvotes (those on which counting machines detected more than one vote, as when someone selected a candidate and also wrote in the same candidate). But the Gore campaign, hoping to cherry-pick enough votes to slip past Bush, requested recounts only of undervotes in four heavily Democratic counties. It ignored the undervotes elsewhere and did not ask for hand counts of overvote ballots (which, it now seems, contained more uncounted Gore votes than did the undervote ballots). Gore and his crew waved the banner, "Let every vote count," but their legal strategy did not match the rhetoric. If Gore had pushed principle rather than relying on tactics, the headlines might be different today, for what that's worth. The consortium review also suggests that Bush's legal plan–an antidemocratic strategy of opposing all recounts–was excessive, leading him to an uglier victory than was necessary.

It's important to keep in mind that the numbers tossed out by the consortium are wisps. They don't provide the basis for sweeping declarations (like "the US Supreme Court had nothing to do with the outcome," or "Gore was robbed of victory"). They're based on various assumptions. For instance, the what-if-the-Supremes-had-not-butted-in scenario–which produces a slender Bush lead–assumes that four counties that had declared their refusal to recount ballots would have maintained that position. But suppose the state Supreme Court had forced them to conduct a recount. The consortium doesn't say what the results would have been. And, as the consortium acknowledges, it was a challenge for it to identify, months after the fact, the undervote and overvote ballots not tallied on Election Day. Conceivably, its 153 ballot inspectors–fielded by the National Opinion Research Center–may have missed a handful of ballots. (On only one out of seven ballots examined did the consortium discern an actual vote, under its most inclusive standards.) With the margin of victory in the consortium review swinging between .003 percent for Gore and .008 percent for Bush, any one error could have flipped the findings.

The consortium produced the perfect split decision. Bush (probably) would have won by the rules Gore advocated; Gore (probably) was preferred by several dozen more voters. But on each side, the point spreads are too slim to mean much or settle anything. The consortium effort does offer other valuable information. The review found that precincts with African-American majorities rejected ballots at much higher rates than did white precincts. Hispanic precincts and those with large elderly populations also had high discard rates. The data do not identify the cause–likely a combination of older machines, poorly trained election workers and a higher number of less-educated voters–but they suggest that vote-counting, intentionally or not, had a racial and age bias. The consortium's report strengthens the arguments for electoral reform. Florida did pass a measure, which was criticized by voting-rights advocates, and Congress and other states have been lollygagging.

The inconclusive conclusions of the consortium do not disperse the fog shrouding the election. Ascertaining the hard-and-fast results (or what should have been the results) of the Florida election remains virtually impossible. Still, it's not difficult to demonstrate that the wrong guy triumphed. The consortium's scenario calculators did not take into account the butterfly ballot of Palm Beach County or the similarly confusing ballot of Duval County. In each instance, thousands of voters errantly double-marked ballots for Gore and a minor candidate or Bush and a minor candidate. No official recount would have been able to include these self-invalidating ballots, so the consortium left them out of its equations. Which was reasonable.

Yet it did confirm what had been previously reported–that these ballot screw-ups cost Gore thousands of votes. The Washington Post says Gore may have lost up to 8,000 votes in Palm Beach and 7,000 in Duval County. The Los Angeles Times reports that Gore was shortchanged by up to 9,000 votes in Palm Beach and 3,000 in Duval County. The Palm Beach Post maintains that he might have missed out on 6,000 votes in its county. All these numbers overwhelm the two- or three-digit margins of every consortium scenario. With or without the consortium, the picture is clear: Gore was favored by several thousand more Florida voters than Bush. That is nearly undeniable–yet that reality could not be reflected in the numbers of the official count, the recounts or the unofficial reviews. It has also not been acknowledged by the fellow who, without shame, accepted the benefits conveyed by a faulty system. In other words, you don't need a consortium to know which way the wind blew–or to know that the man in the White House is there, legitimately or not, by mistake.