04/04/2001 - Updated 09:38 AM ET
Newspapers' recount shows Bush prevailed
By Dennis Cauchon, USA TODAY
George W. Bush would have won a hand count of Florida's disputed ballots if the standard advocated by Al Gore had been used, the first full study of the ballots reveals. Bush would have won by 1,665 votes — more than triple his official 537-vote margin — if every dimple, hanging chad and mark on the ballots had been counted as votes, a USA TODAY/Miami Herald/Knight Ridder study shows. The study is the first comprehensive review of the 61,195 "undervote" ballots that were at the center of Florida's disputed presidential election.
The Florida Supreme Court ordered Dec. 8 that each of these ballots, which registered no presidential vote when run through counting machines, be examined by hand to determine whether a voter's intent could be discerned. On Dec. 9, the U.S. Supreme Court stopped the hand count before it was completed. That gave Bush Florida's 25 electoral votes, one more than he needed to win the presidency.
USA TODAY, The Miami Herald and Knight Ridder newspapers hired the national accounting firm BDO Seidman to examine undervote ballots in Florida's 67 counties. The accountants provided a report on what they found on each of the ballots.
The newspapers then applied the accounting firm's findings to four standards used in Florida and elsewhere to determine when an undervote ballot becomes a legal vote. By three of the standards, Bush holds the lead. The fourth standard gives Gore a razor-thin win.
The results reveal a stunning irony. The way Gore wanted the ballots recounted helped Bush, and the standard that Gore felt offered him the least hope may have given him an extremely narrow victory. The vote totals vary depending on the standard used:
* Lenient standard. This standard, which was advocated by Gore, would count any alteration in a chad — the small perforated box that is punched to cast a vote — as evidence of a voter's intent. The alteration can range from a mere dimple, or indentation, in a chad to its removal. Contrary to Gore's hopes, the USA TODAY study reveals that this standard favors Bush and gives the Republican his biggest margin: 1,665 votes.
* Palm Beach standard. Palm Beach County election officials considered dimples as votes only if dimples also were found in other races on the same ballot. They reasoned that a voter would demonstrate similar voting patterns on the ballot. This standard — attacked by Republicans as arbitrary — also gives Bush a win, by 884 votes, according to the USA TODAY review.
* Two-corner standard. Most states with well-defined rules say that a chad with two or more corners removed is a legal vote. Under this standard, Bush wins by 363.
* Strict standard. This "clean punch" standard would only count fully removed chads as legal votes. The USA TODAY study shows that Gore would have won Florida by 3 votes if this standard were applied to undervotes.
Because of the possibility of mistakes in the study, a three-vote margin is too small to conclude that Gore might have prevailed in an official count using this standard. But the overall results show that both campaigns had a misperception of what the ballots would show. The prevailing view of both was that minority or less-educated Democratic voters were more likely to undervote because of confusion.
Gore's main strategy throughout the post-election dispute was to secure a recount of any kind in the hope of reversing the certified result. Bush's strategy was to stop the recount while he was ahead. But his views on how recounts should be done, in the counties where they were underway, would have been potentially disastrous for him if used statewide.
Bush and Gore were informed Tuesday of the new study's results. Both declined comment. But White House spokesman Ari Fleischer said, "The President believes, just as the American people do, that this election was settled months ago. The voters spoke, and George W. Bush won."
The newspapers' study took three months to complete and cost more than $500,000. It involved 27 accountants who examined and categorized ballots as they were held up by county election officials.
The study has limitations. There is variability in what different observers see on ballots. Election officials, who sorted the undervotes for examination and then handled them for the accountants' inspection, often did not provide exactly the same number of undervotes recorded on election night.
Even so, the outcome shows a consistent and decisive pattern: the more lenient the standard, the better Bush does. Because Gore fought for the lenient standard, it may be more difficult now for Democrats to argue that the election was lost in the chambers of the U.S. Supreme Court rather than the voting booths of Florida.
The study helps answer the question: What would have happened if the U.S. Supreme Court had not stopped the hand count of undervotes?
However, it does not answer all the questions surrounding another set of Florida ballots: the 110,000 "overvotes," which machines recorded as having more than one presidential vote. These ballots were rejected by the machines and were considered invalid. Some Democrats say if all of Florida's overvote ballots were examined by hand to learn voters' intent, their candidate would have prevailed.
USA TODAY, The Miami Herald and Gannett and Knight Ridder newspapers also are examining Florida's overvotes for a study to be published later this spring. Overvotes contain some valid votes, mostly instances when a voter marked the oval next to a candidate's name and then wrote in the name of the same candidate.
No candidate requested a hand count of overvotes, and no court — federal or state — ordered one. The U.S. Supreme Court cited the state court's failure to include the overvotes in its recount order as an example of arbitrariness.
Immediately after Gore conceded the election to Bush, The Miami Herald began to evaluate what might have happened if the U.S. Supreme Court had not stopped the recount of undervotes.
Florida is one of the few states that permit members of the public to examine ballots after they've been cast. The Miami Herald and the BDO Seidman accounting firm began examining ballots on Dec. 18. USA TODAY joined the project in January. The last undervote ballot was examined March 13.
Florida law requires that political parties be notified of ballot inspections. The Republican and Democratic parties took different approaches to the three months of ballot inspections.
The Democrats took a hands-off approach. They rarely showed up at election offices during the evaluation. "We want to see what you find. It's not our role to be at the table with you," Tony Welch, spokesman for the Florida Democratic Party, said during the newspapers' study. "If we're spinning and the Republicans are spinning, people won't believe the result."
He said at the time that the party expected the outcome would show that Gore received more votes than Bush.
By contrast, the Republicans attended every ballot inspection. They devoted hundreds of days of staff and volunteer time. The party delayed cutting its post-election staff of field directors from 12 to 6 so it could staff the ballot inspections. Some Republicans took meticulous notes on the contents of the ballots. Others just watched. The Republican Party of Florida published a daily internal memo called "Reality Check," which critiqued the media efforts to examine ballots.
In an interview before the results were released, Mark Wallace, a Republican lawyer assigned to critique the media inspections, said, "The media appear ready to offer unprecedented liberal standards for judging what is a vote. The appropriate legal standard is what was in place on Election Day: cleanly punched cards only."
Before this election, almost nothing was known by the public and by political parties about what types of marks appear on undervotes and overvotes, which make up about 2% of ballots cast nationally. The newspapers' study shows how both parties predicted incorrectly which of these ballots would help them.
Democrats and Republicans noted that voter errors on punch-card voting machines were most frequent in low-income and predominantly minority precincts. Because these voters tend to vote Democratic, the disputed votes were assumed to be a rich trove of support for Gore.
Likewise, both parties noted that the 41 Florida counties that used optical-scan ballots, a system similar to standardized school tests, tended to vote Republican.
Bush supporters attacked Gore for asking for hand counts in three Democratic-leaning counties. If any hand count occurred, it should include the Republican-leaning optical-scan counties, too, the Bush supporters said.
The USA TODAY/Miami Herald/Knight Ridder study shows that the Democratic and Republican assumptions were largely wrong. The undervote ballots actually break down into two distinct categories:
* Undervotes in punch-card counties. In the 22 punch-card counties in which BDO Seidman examined undervotes, 56% of the 35,761 ballots had some kind of mark on them.
The study found that punch-card undervotes correlated less to race or party affiliation than to machine maintenance and election management. Counties that maintain machines poorly — not cleaning out chads frequently, for example — have plentiful undervotes. The study shows that when undervotes are hand counted, they produce new votes for the candidates in proportions similar to the county's official vote.
For example, in Duval County, where Jacksonville is the county seat, Bush defeated Gore 58%-41%. Among the undervotes, Bush defeated Gore 60%-32% under the lenient standard and by similarly comfortable numbers under all standards. Bush picked up a net of 930 votes, including 602 dimples.
Likewise, in Miami-Dade, where Gore hoped to score big gains, he received 51% of the marked undervotes, about the same as the 52% that he got in the official count.
* Undervotes in optical-scan counties. In the 37 optical-scan counties in which BDO Seidman examined undervotes, one third of 5,623 ballots had discernible votes.
The most common was when a voter made an X or check mark, rather than filling in the oval properly. Other common errors included circling the candidate's name or using a personal pencil or pen that couldn't be read by the machine. Black ink that contains even a trace of red will not register on many vote-counting machines, even when the mark appears pure black to the human eye.
The study shows that these errors were disproportionately common among Democratic voters. For example, in Orange County, home of Orlando, Gore edged Bush 50%-48% in the election. But Gore won the undervotes by 64%-33%, giving him a net gain of 137 votes. That accounted for half of the 261 votes Gore gained in optical-scan counties, which Bush won overall by 53%-44%.
The study found that optical-scan counties are the only places where Gore actually picked up more votes than Bush: 1,036 to 775 for Bush.
In the punch-card counties, where Gore had placed his hopes, his chances of winning a hand count were washed away. On dimples alone, Bush gained 1,188 votes. When all the possibilities are combined — dimples, hanging chads, clean punches — Bush outdid Gore by 8,302 to 6,559.
USA TODAY's analysis is based on accepting Bush's official 537-vote margin. This figure includes hand counts completed in Broward and Volusia counties before the U.S. Supreme Court intervened.
The newspaper also accepted hand counts completed in Palm Beach, Manatee, Escambia, Hamilton and Madison counties, plus 139 precincts in Miami-Dade.
These hand counts, which were never certified, reduced Bush's lead to 188 — the starting point for USA TODAY's analysis.
The newspaper excluded these counties from its analysis. However, BDO Seidman collected data in these counties, and they are available on USATODAY.com.
In the end, Florida's presidential election remains remarkably close by any standard: 2,912,790 to 2,912,253 in the official count.
In an election this close, the winner often depends on the rules and how they are enforced.
Contributing: Gannett News Service database editor Robert Benincasa
CAN YOU STOP THE DAMN WHINING NOW?????????
February 26th thread on recount results
-- Ain't Gonna Happen (Not Here Not@ever.com), April 04, 2001
MIAMI, April 3 — George W. Bush almost certainly
would have held his lead in Florida, the state that
was key to clinching the presidential election, even
if the U.S. Supreme Court had not stopped a vote
recount, results from an analysis of ballots
released by the Miami Herald on Tuesday showed
