The first of two major reviews of uncounted ballots in the 2000 presidential election in Florida announced its final conclusions earlier this month. A total of 171,908 ballots—60,647 undervotes (ballots that registered no presidential vote in machine tabulation) and 111,261 overvotes (those disqualified because they were marked for more than one presidential candidate)—were examined in a review organized by a consortium made up of USA Today, the Miami Herald and the Knight-Ridder newspaper chain.
The report, like partial results announced in February and April, was generally presented in the news media as legitimizing the installation of George W. Bush as president on the basis of a margin of several hundred votes in Florida, despite his having lost the popular vote by a significant margin nationwide. The widespread voter disenfranchisement that took place in Florida continues to be obscured beneath an avalanche of minutiae about various methods of ballot-counting.
Reporting on the undervotes, USA Today discussed four possible standards for judging whether punch-card votes (used in 25 of Florida's 67 counties in the 2000 election) were valid. Bush would have won under the two standards most widely used, the paper said. If at least two corners of a partially attached bit of paper, or chad, must be detached to validate a vote, Bush's lead would have dropped from the official 537 to 407, still enough to award him Florida's 25 electoral votes and the election. By the strictest standard, requiring a completely clean punch for the vote to be counted, Bush's lead would have been even smaller, 152. Under two looser standards, in which “dimpled” but not detached chads are counted, Gore would have won the state by a margin of between 242 and 332 votes.
These figures were used to manufacture headlines suggesting that Bush's occupancy of the presidency was valid and reasonable. “Bush Would Win Recount of Disputed Ballots,” said Reuters. “Bush Still Wins Florida,” reported CNN.com. “Vote Analysis: Bush Wins, Again,” ABCnews.com declared.
Some reports acknowledged continuing doubt. “No Clear Florida Winner,” said the Associated Press. The New York Times, burying the story on its inside pages under a one-column headline, reported: “Second Review of Florida Presidential Vote is Inconclusive.”
None of the headlines, however, declared what was fairly obvious before the recount and was in fact confirmed by the review: more votes were cast for Gore than for Bush in Florida. They covered this up by excluding the overvotes from their analysis. Most of the stories grudgingly acknowledged, well into the text, that if the overvotes had been counted, Gore would have won between 15,000 and 25,000 additional votes, leading to a substantial margin of victory in the state.
In most reports this fact was added almost as an afterthought, along with the suggestion that this is of only academic concern, since the Gore campaign never requested a review of these ballots. The recount ordered by the Florida Supreme Court and halted by the 5-4 decision of the US Supreme Court on December 12 called only for a review of undervotes.
To the extent that the media was forced to acknowledge the huge number of uncounted Gore votes, this was reported in such a way as to blame the individual voters. “Florida Voter Errors Cost Gore the Election,” was the USA Today headline. Once again, only deep inside the article did the newspaper acknowledge that these errors were largely the result, not of voter indifference or negligence, but of faulty ballot design and instructions that, intentionally or not, had the effect of disenfranchising less experienced and sophisticated voters.
A study of voting patterns for 56,000 of the overvotes conducted by Anthony Salvanto of the University of California showed that far more Democratic voters were led to make errors that invalidated their votes. Voters who marked Bush or Gore on overvote ballots (i.e., who marked another candidate as well as Bush or Gore) usually voted for the same party's candidates in other races. Some 83 percent of overvoters who voted for both Gore and a third-party candidate voted Democratic in the US Senate race, for example. Only 6 percent of those who overvoted in the presidential race made the same mistake in the Senate election, which was next on the ballot. A total of 84,197 of the overvotes included Gore, while only 37,731 named Bush. (The numbers add up to more than the total of 111,261 overvote ballots because some ballots included the names of both Bush and Gore).
The reason for the disqualified Gore votes was design and instruction. In Duval County, for instance, including the major city of Jacksonville, voters were shown the first five presidential candidates on one page and another five candidates on a second page. After the first page they were instructed in writing to “turn page to continue voting.” At the same time, a sample ballot distributed by election officials instructed them to “vote every page.”
As a result, there were 21,188 overvotes in Duval County. This one county had more than one-fifth of the state total of overvotes. Some 55 percent of the Duval County overvotes included just two candidates, one from the first ballot page and one from the second, indicating that the confusing instructions were the cause. Most of these invalidated ballots were for Gore. As Salvanto commented, “The Duval County ballot alone likely cost Gore the election.”
The Palm Beach County “butterfly ballot” has already been much discussed. The names of the presidential candidates appeared on two facing pages, and the ballot was designed in such a way that Gore, who was the second candidate listed, was the third hole to punch. The second hole was assigned to extreme right-wing spokesman Patrick Buchanan, which resulted in 5,237 overvotes for Gore and Buchanan.
Thousands of elderly Jewish voters in Palm Beach County unwittingly cast ballots for a candidate notorious for his apologies for anti-Semitism. Once again, a fair vote in Palm Beach County would by itself have given the state's electoral votes to Gore. A recent study indicated that the butterfly ballot cost Gore at least 3,400 votes because of double punches, and another 2,400 votes that were mistakenly cast for Buchanan.
Even if the ballot review were confined to undervotes, the number of undervote ballots produced by officials in most counties did not match the totals reported by these same counties immediately after election day. The numbers matched in only 8 of 67 counties. In Orange County, for instance, carried narrowly by the Democrats, the “ballot slippage” resulted in 966 reported undervotes evaporating to only 639 when the Miami Herald consortium recounted the votes.
The issues relating to undervotes and overvotes are only half of the story of Florida and the 2000 election. The vote recounts do not consider the variety of other ways in which citizens were denied the right to vote. Black voters made numerous complaints of intimidation and harassment on election day itself. Tens of thousands of qualified voters, many of them black, were unlawfully purged from the rolls on the false grounds that they were felons. (Twenty-four percent of Florida's black men of voting age are, under state law, permanently denied the right to vote because they are felons, even though many have completed their sentences and are no longer on parole).
The network exit polls on election day were not mysteriously wrong in Florida, while correct everywhere else. They accurately forecast that Gore would carry the state, because they did not—and could not—take into account the enormous number of ballots that would be invalidated. A combination of legal chicanery, Republican intimidation, media pressure and the intervention of the US Supreme Court handed the election to Bush.