Amid all the partisan sniping, talking-head screeching and judicial decisions, there are two indisputable facts that go far toward explaining the true tragedy of the Florida recount.

Fact one: In this election, punch-card voting machines recorded five times as many ballots with no presidential vote as did the more modern optical-scanning systems. A New York Times analysis of forty-eight of the state's sixty-eight counties found that 1.5 percent of the ballots tallied under the punch-card method showed no vote at the top of the ticket, while only 0.3 percent of the ballots counted by the newer machines registered no vote for the President. An Orlando Sun-Sentinel examination concluded that counties using the best optical-scanning method recorded presidential votes on more than 99 percent of the ballots, and counties using the old punch-card devices counted presidential votes on only 96.1 percent of the ballots.

Fact two: Punch-card machines were more widely used in areas where low-income and African-American citizens vote. Two-thirds of the state's black voters reside in counties using punch cards, while 56 percent of white voters do.

Put these two undeniable facts together and the conclusion is inescapable: A statistically significant slice of the Florida electorate was disfranchised by voting technology. That is, a disproportionate number of voters done in by the error-prone punch-card machines were low-income and black Floridians, who generally favored Al Gore over George W. Bush. Presumably, some no-vote ballots actually did not include a vote for President. But given the closeness of the election--decided by .008 percent--it is likely that presidential votes missed by the punch-card machines would have decisively affected the contest. Bush "won"--among other reasons--because of voting-machine discrimination.

This crucial part of the tale has been overwhelmed by dimple-mania and the usual campaign back-and-forth. But ten days after the election, the Sun-Sentinel reported that "Florida's different vote-counting machines resulted in more GOP votes." For example, Brevard County, the home of space-shuttle launches, spent $1 million on more advanced machines in 1999, moving from punch-card tabulators to optical scanning machines that read pen-marked ballots (and that immediately return to the voter a ballot with a problem). Under the new system, the voting machines in this Bush-leaning county found presidential votes on 99.7 percent of the ballots. In 1996 the county's punch-card machines read presidential votes on 97.2 percent. Which means Bush, thanks to the upgrade, likely banked an additional 453 votes for his statewide total--practically his post-recount victory margin. The paper noted that the twenty-five counties that used the punch-card machines went for Gore over Bush 51.8 percent to 46 percent and produced 144,985 ballots with unrecorded presidential votes. Had the people who cast these ballots entered voting booths equipped with the more efficient machines, Gore no doubt would have collected hundreds--if not thousands--more votes than Bush.

There have been allegations that black Floridians encountered racial intimidation at voting sites. (The Justice Department has initiated an informal assessment, not an investigation.) And Bush benefited from the all-too-routine bias by which minority areas receive poorer government services. Unfortunately not just for Gore but for the victims of this quiet bias in Florida, this inequity was unaddressed by the Florida circuit court and the US Supreme Court, partly because the Gore campaign didn't raise it.

The Gore legal challenge focused on 14,000 or so supposedly no-vote punch-card ballots in Miami-Dade and Palm Beach counties, not the statewide problem, and called for a manual review only of those ballots. The Veep's lawyer did not argue that the county-by-county patchwork voting system operated less effectively for blacks, a constituency that Democrats rely on to win elections. In his ruling against Gore, Circuit Judge N. Sanders Sauls noted that the record "shows error and/or less than total accuracy in regard to the punch-card voting devices utilized in Dade and Palm Beach Counties." But Sauls declared that Gore's legal team had not established "a reasonable probability" that the statewide results would turn out differently if those ballots were counted in a better fashion. Either Gore's attorneys screwed up big by not making this point more obvious--which they might have done had they filed contests based on the wider issue--or Sauls misread the math. As for the US Supreme Court, it displayed no eagerness to adjudicate such a touchy and fundamental voting-rights matter as systematic disfranchisement through technology. Its decision--in which it told the state Supreme Court to try again--indicates that the Court wanted to approach the Florida case narrowly, at least in the first go-round.

If a system is decisively skewed to one group's advantage, does that amount to theft? Or is that just the way it is? Clearly, a more equitable vote-counting system in the state--punch-cards for all or optical-scanners for all--would have yielded a different final count. This is an injustice that no court has confronted, on which Bush may well ride into the White House, and that should not be forgotten.

• Get The Nation at home (and online!) for 75 cents a week!
• If you like this article, consider making a donation to The Nation.