A statewide review of the Florida presidential ballots by two media organizations demonstrated just how subjective a hand recount can be, with the outcome hinging on which standards are applied and which ballots are reviewed.
The newspaper study said President Bush would have been the winner of a recount ordered by the Florida Supreme Court under the combinations of factors they considered most likely. But some combinations of circumstances the newspapers considered less likely could have given the election to former vice president Al Gore.
Even more uncertainty emerges, the study found, because their own reviewers at times came to different conclusions about the same ballot.
In addition, it was hard for county election officials even to locate undervotes -- the ballots on which no vote for president was discerned by the machine -- due to inconsistent results when ballots are re-run through balky voting machines. That meant that not all the undervoted ballots were reviewed and that other votes may have been counted twice, once by machine and again in the manual review.
"Some would say that [uncertainty] makes the whole effort futile," said Mark Seibel, managing editor/news for the Miami Herald, which published the story yesterday along with its parent company, Knight Ridder Newspapers, and USA Today. "I think it just tells you something about the kind of elections equipment we have. It's extremely imprecise."
The $550,000 study was intended to determine who would have won if the Dec. 8 statewide recount order by the Florida Supreme Court had not been halted the next day by the U.S. Supreme Court.
In trying to answer that question, the newspapers excluded counties that had already finished their recounts by the time the U.S. Supreme Court stepped in -- leaving the existing counts in place. Under that scenario, they said, Bush was the most likely winner -- whether or not counters used a permissive standard that included ballots that were merely dimpled or only counted as votes those with at least partially detached chads.
However, when the newspapers redid the counts themselves, reviewing ballots in counties such as Broward and Palm Beach that had already completed their own manual review, it found that Gore could have won the election if a loose standard was used, counting every dimpled ballot as a vote. Under a stricter test, Bush would have won in that situation as well.
The newspapers retained the BDO Seidman accounting firm to examine 64,248 "undervotes," ballots on which no vote for president could be read by machine. The accountants were not asked to decide whether any ballot markings constituted a vote, but used a classification scheme to describe ballots as having dimples or partially detached chads.
The political parties seized on the outcomes they preferred and disparaged the scenarios that were less generous to them.
"The truth is the election's been over for three months," said Mark Miner, a spokesman for the Republican National Committee. "While this may make for good commentary in Washington, the president is trying to move forward and focus on his agenda of tax relief and education."
A statement released by Democratic National Committee Chairman Terry McAuliffe dwelt on the Gore-victory view.
"The same study that Republicans tout as proving that Bush really won Florida, also shows that if all the ballots were counted on election night, Al Gore would have won," he said.
"Overall the study didn't conclude very much," said former Gore campaign spokesman Doug Hattaway. "Bush won under some scenarios, and Gore wins under others. It certainly confirms Florida's election system failed the voters."
"One of the most striking findings is Palm Beach and Broward counties threw out hundreds of ballots with valid votes under their own criteria," he added. "If that hadn't happened, Al Gore could well be in the White House right now."
The Herald, USA Today and some other partners are also examining overvote ballots from the presidential election, ballots that were invalidated because more than one presidential candidate was marked. The newspapers said that study would be done in about a month.
A separate ballot review study is underway by a group of news organizations including The Washington Post, the New York Times, the Wall Street Journal, Associated Press, CNN, the Los Angeles Times, the Chicago Tribune, the Fort Lauderdale Sun-Sentinel, the Orlando Sentinel, Newsday, Newsweek, the Palm Beach Post and the St. Petersburg Times. That study of both undervotes and overvotes is being conducted by the National Opinion Research Center at the University of Chicago. It may be complete next month.
The Herald/USA Today stories noted an obscure provision in Florida law that requires an election night manual inspection of "defective" ballots that the machines were unable to read.
If that inspection were done, and every dimple or other mark were counted as a vote, Gore could have won the election, the study said. But election officials noted that large counties have several thousand punch-card ballots without discernible votes and have never tried to review them on election night. Some smaller counties, such as Leon, routinely go through the undervotes and overvotes to try to find attempted votes on election night.
Phil Meyer, who holds the Knight chair of journalism at the University of North Carolina-Chapel Hill and advised the newspapers on the project, said declaring a winner is newsy but ignores the question of whether the uncertainties are too great to make that assertion.
"The really interesting question is whether it's ever really knowable," he said.
