Bottom line: Gore lost, fair and square.
`OVERVOTES' LEANED TO GORE
But to win, he needed help of dimpled ballotsBY MARTIN MERZER
mmerzer@herald.com
Democrat Al Gore might be president today if Florida's ``overvotes'' had been examined and counted -- but only if dimples on ``undervote'' ballots were accepted as valid votes, the first statewide review of overvotes shows.
Republican George W. Bush still would have prevailed -- even with the overvotes tabulated -- if undervotes had been counted under more restrictive standards, the review indicates.
The review, conducted by The Herald, its parent company, Knight Ridder, USA Today and several other newspapers, also shows that Gore's name was marked on overvotes far more often than Bush's name -- fodder for Democrats who insist that most Floridians intended to vote for Gore.
Overvotes are ballots rejected by counting machines because they show more than one vote for president. Undervotes are ballots without presidential votes detected by counting machines.
The findings produce an ambiguous conclusion to an unprecedented effort to examine more than 176,000 untabulated ballots in Florida's disputed presidential election.
VIRTUAL TIE
The bottom line: After study and analysis of 111,261 overvotes and 64,826 undervotes, the agonizingly tight 2000 presidential election still ends in a virtual tie.
And the outcome still depends on the standard used to gauge undervotes.
Gore wins narrowly under two undervote standards, by margins of 332 and 242 votes; Bush wins narrowly under two other undervote standards, by 407 and 152 votes. All are closer than Bush's official 537-vote margin.
And so, the new analysis of the election intrigued or irritated political partisans Thursday, depending on their point of view.
``These numbers certainly back up our feeling that more people turned out to vote for Gore than for Bush,'' said Doug Hattaway, a former Gore campaign spokesman. ``It's just a shame the system didn't count those votes properly. It's hard not to cringe when you think about the possibilities.''
Speaking for the Republicans, former Montana Gov. Marc Racicot said he never feared ballot reviews by the media.
``So what has changed?'' Racicot said. ``I always believed we would come to the same conclusion at the end as at the beginning. The election was not flawless, but it never is. The only count that really counts is the one conducted under the rule of law.''
The overvote study is the third examination of how the acrimonious 2000 presidential election might have ended if manual recounts had gone forward without court challenges and intervention.
In February, The Herald reported that Bush still would have won the presidency if undervotes in Miami-Dade County had been counted by hand, as Gore's campaign had asked.
The Miami-Dade canvassing board halted that recount after concluding that it could not be completed by a state deadline. The Herald's review found that Gore would have picked up, at most, 49 net votes in Miami-Dade, using the most liberal undervote standard.
Last month, The Herald reported that Bush's victory almost certainly would have endured even if a statewide recount of undervotes ordered by the Florida Supreme Court had not been terminated by the U.S. Supreme Court. By the most liberal standard, Bush's lead would have widened to 1,665 votes had that recount been completed.
FULL RECOUNT
Now, the study of overvote ballots indicates that Gore might have won the election -- if his campaign had requested and received a hand recount of every uncounted ballot in the state.
The Herald found that an astonishing 108,115 overvotes -- 97 percent of the total -- are lost forever. When a voter selects two or more candidates, the true preference cannot be deduced or assumed, so those votes are unsalvageable.
But the rest -- 3,146 -- bore markings that made it clear who the voter preferred, The Herald found. Generally, this occurred when voters chose a candidate and then cast a write-in vote for that same candidate.
``It's amazing to me how some people misinterpreted how to vote, but it's clear what they intended to do,'' said Ronald Legendre, chairman of Osceola County's canvassing board.
Most of those recoverable overvotes -- 1,871 -- were for Gore. Bush received 1,189 such votes. Other candidates received 86.
That net gain for Gore of 682, when coupled with the undervotes, would have been enough to carry the Democrat to the White House under two standards for gauging undervotes. Gore's margin would have been 332 votes if all dimpled ballots were counted. It would have been 242 votes if dimpled ballots were counted only when dimples also appeared in other races.
But if the undervotes were declared valid only when chads were detached by at least two corners, Bush would have come out on top by 407 votes. And if only cleanly punched ballots were counted, Bush would win by 152 votes.
Even so, there is ample room for argument that a significant plurality of Floridians intended to vote for Gore -- but more of them ruined their ballots by overvoting or undervoting.
One measure of that: Three of every four overvotes -- a total of 84,204 -- contained a mark for Gore; only one of three overvotes -- a total of 37,738 -- contained a mark for Bush. (Some overvotes show marks for both).
``Gore would likely have won if all overpunched ballots had been properly marked, based on measures of voter intent,'' said Anthony Salvanto, a political scientist at the University of California-Irvine who analyzed the overvote results.
This also seems clear:
At an early, crucial stage of the electoral impasse, Gore made a momentous strategic mistake by not asking for an official statewide review of every rejected ballot -- undervotes and overvotes. Instead, he requested a recount only in four counties.
His decision might have been a historic blunder. A statewide recount of undervotes and overvotes might have given Gore the vote lead and the political high ground -- before the legal brawl ever reached the Florida Supreme Court or the U.S. Supreme Court.
Hattaway, the former Gore aide, said the Democrats did not request a recount of that magnitude because it seemed too sweeping, too desperate and too unwieldly.
`NO PRECEDENT'
``There was no precedent for it -- until now,'' he said. ``It was discussed but the consensus was that we couldn't get it. There was a feeling the courts wouldn't give it to us.''
Interestingly, if Florida's new election law had existed last November, Gore would have been entitled to just such a review -- automatically.
Signed Wednesday by Gov. Jeb Bush, the law requires a recount of every undervote and overvote in the state when the victory margin is less than a quarter of one percent of the total votes cast.
The 2000 presidential election ended in Florida with a margin much slimmer than that -- 0.006 percent after the last complete machine recount.
DIFFERENT REVIEW
The Herald's review of the state's overvotes differed in several ways from the newspaper's previous undervote review. The entire project cost The Herald, Knight-Ridder and its partners about $825,000.
The undervote review was undertaken by a public accounting firm, BDO Seidman, LLP, under the sponsorship of The Herald, Knight Ridder and USA Today. An accountant reviewed each ballot and noted its characteristics.
All tabulations of undervotes were provided by BDO Seidman, based on those observations. Separately, a reporter also reviewed each ballot, but the results of that review were used only as a statistical check for variation.
Each overvote, however, was reviewed by one person only -- a reporter from The Herald, USA Today or six other newspapers: The Tallahassee Democrat, The Bradenton Herald, Florida Today, The Tampa Tribune, the Fort Myers News-Press and The Pensacola News Journal.
In 59 counties, reporters physically examined every overvote ballot and recorded what they found. In eight heavily populated counties, the newspapers used official computer tapes to identify combinations of candidates marked on overvoted ballots. In those counties, reporters then examined all ballots that contained handwritten marks that would not have appeared on computer records.
The results of the overvote review were assembled into a database, and the various combinations were examined.
The Herald then attempted to recreate the conditions of the election prior to any court action and before hand recounts in Volusia, Broward and Palm Beach had been added to the official count.
Machine counts certified by each county by Nov. 14 plus subsequently counted overseas absentee ballots and some recently identified undervotes from Orange County showed Bush leading Gore by 1,133 votes in Florida.
Gore's 682-vote net gain from overvotes would have sharply reduced that lead -- to 455. And that lead would have been overtaken had undervote ballots with dimples been included.
But no such statewide recount of undervotes and overvotes was mandated by law last November, so county election supervisors were not required to search for recoverable votes or even accept any they found.
In Lake County, for instance, canvassers happened upon some obviously valid overvotes -- ballots that contained double votes for Gore or for Bush.
The canvassing board voted 2-1, however, not to rehabilitate those ballots because the board planned no systematic search of all rejected ballots.
``I was whooping and hollering the whole night,'' recalled Donna Miller, the canvassing board chairwoman, who lost that battle. ``I was so frustrated. There was no question in my mind who they wanted. I could hardly sleep that night, I was so upset.''
OTHER FINDINGS
Among The Herald's other findings:
- Statewide, the combination of votes for Gore and Pat Buchanan, the Reform Party candidate, represented the most common overvote, accounting for 10,234 botched ballots or nine percent of the total. Other common combinations: Gore and Libertarian Harry Browne, Bush and Gore, Bush and Buchanan.
- A mark for Buchanan appeared on 36,757 overvotes, nearly as frequently as the marks for Bush. That probably was caused by Buchanan's proximity on ballots to both Bush and Gore, especially on Palm Beach County's now-notorious ``butterfly'' ballot.
- Nearly 1,000 people voted for every candidate, disenfranchising themselves.
- More than 3,600 voted for every candidate except Bush. More than 700 voted for every candidate except Gore.
``We could call these `anybody but him' voters,'' Salvanto said. ``One theory is that these are people who simply do not understand the voting process and are marking everyone of whom they approve. Just as long as the guy they don't like doesn't get in.''
Most of the overvotes in Florida came from punch-card counties, show clearly punched ballots for multiple presidential candidates and thus can never be rehabilitated.
In Palm Beach County, for instance, 5,237 people punched holes for both Gore and Buchanan, but no one can say precisely how many of them wanted to vote for Gore. Exactly 13,511 people cast other types of overvotes in Palm Beach County.
BUTTERFLY BALLOT
A recent statistical study by six political scientists from Harvard University, Cornell University, Northwestern University and the University of California-Berkeley found that Palm Beach County's butterfly ballot probably cost Gore at least 3,400 votes because of double punches and up to another 2,400 votes that were mistakenly cast for Buchanan.
Duval County, which includes Jacksonville, experienced an even larger, more astounding number of punch-card overvotes -- 21,888. Later, many voters there also complained about a confusing ballot, one that distributed presidential candidates over two pages.
Eighty-four percent of overvoters in Duval punched presidential candidates on both pages.
Regardless of ballot design, regardless of voting system, every county in the state reported overvotes.
VOTING DIRECTIONS
``I have directions printed on the ballot, directions with pictures, sample ballots with graphics sent to every household in the county, and the same thing goes in the [news]paper,'' said Pat Hollarn, supervisor of elections in Okaloosa County, which used the more modern and reliable optical-scan system but had 677 overvotes anyway.
``When we give them that much instruction, if they sent an absentee ballot that's marked double, then I can't tell what somebody meant by that. Nobody can. You can't guess what somebody meant.''
Still, The Herald's study confirms that many overvotes would have been accepted as valid -- if election officials had looked at them in time.
And in an election as close as the 2000 presidential race between George W. Bush and Al Gore, those ballots might have made a difference.
Herald staff writers Lila Arzua, Tyler Bridges, Tim Henderson and Jay Weaver contributed to this report.