Toss out felon vote, Gregoire still wins
The court battle
Republican gubernatorial candidate Dino Rossi sued in Chelan County Superior Court in January, seeking to overturn Gov. Christine Gregoire's election. He said illegal votes and errors by election workers made her victory illegitimate.
Rossi's case is set to go to trial and is expected to last nine days. Judge John Bridges will hold a hearing early in the trial to determine the validity of statistical evidence Republicans want to use. That evidence, if approved, could be used to subtract alleged illegal votes from Gregoire's total. Democrats have countered with votes they say should be taken from Rossi.
Bridges could nullify Gregoire's election and declare Rossi the winner. Rossi has said he wouldn't accept victory by court judgment, so the ruling could create a vacancy in the governor's office. State law provides for a special election to fill a vacancy. Temporarily, the office would be filled by Lt. Gov. Brad Owen.
Both sides agree that whatever happens in Bridges' court, the outcome will be appealed to the state Supreme Court.
How we did it
The Times spent nearly three months examining the Republicans' list of alleged felon voters. The Democrats released their list only recently and without court-case numbers, making it impossible to do a similar investigation.
The Times obtained from the GOP a list of 1,135 names and thousands of pages of supporting court documents. The size of the list fluctuated as the party amended it for its court case.
Reporters compared the list with three state databases: a list of convicted felons from the Washington State Patrol, and two Department of Corrections lists — one of convictions and the other of felons whose voting rights had been restored.
Because those databases were incomplete, the Times used a computer to randomly select 300 names from the Republican list for more thorough research. That sample later was trimmed to 289 as the Republicans dropped names from their original list. Their list as of Friday morning contained 946 names.
For the sample, reporters then:
• Visited 13 counties and reviewed court files to verify felony convictions and ensure that voting rights had not been restored.
• Used three national databases, court files and other public records to search for phone numbers, addresses, signatures, birth dates or Social Security numbers of the people on the list to determine if the alleged felon was the same person as the voter the Republicans had named. This step revealed some cases of mistaken identity.
• Got copies of polling-place record books and absentee-ballot envelopes to compare the signatures of the voters against those of the felons.
• Interviewed 49 felon voters, by telephone or in person, and verified that they belonged on the list. Four others who were interviewed had had their voting rights restored.
This story was reported by Cheryl Phillips, Justin Mayo, Emily Heffter, Jonathan Martin, Mike Carter and Nick Perry, and written by Heffter. Researcher Miyoko Wolf and reporter David Heath also contributed.
Copyright 2005, The Seattle Times Co.
For months, state Republicans have insisted that hundreds of votes cast illegally by felons put Democrat Christine Gregoire in the governor's mansion.
But a Seattle Times analysis finds that even if those votes were disqualified, Gregoire would still prevail over GOP challenger Dino Rossi.
That finding undercuts what has been the most prominent element of the Republicans' case to overturn the election. The case moves to trial in Wenatchee tomorrow.
Gregoire won the election by a scant 129 votes — and only after Rossi finished ahead in the first two of three vote counts.
Republicans filed suit in Chelan County in January, contesting a variety of votes they say were illegal. Those include ballots cast by people the GOP says voted twice and those cast in the names of dead people. But felon votes are the GOP's biggest target.
The party submitted a list of 946 people it says were felons who voted illegally in the election. Assuming those felons voted proportionately the same as others in their precincts, the GOP argued that removing their votes from the tally would hand the election back to Rossi.But the Democrats followed with a list of their own, citing 794 illegal felon voters. And they concluded the opposite: Gregoire still wins.
Using the "proportional deduction" method proposed by the Republicans, The Times' initially analyzed both lists, presuming they were entirely accurate. That approach showed Gregoire ahead by 112 votes.
So how could that be? The key is not necessarily in the number of felon votes, but where those votes were cast: Almost all of the felon voters identified by Democrats were cast in precincts won by Rossi. Using the proportional deduction method, that means more Rossi votes would be tossed out.
The Times went further in its analysis. Its investigation found that both parties made mistakes in compiling their lists: Some people on the lists were not felons — nor had they voted.
An analysis of a random sample of names on the GOP list submitted in March found that about 1 in every 9 names was wrongly included — an error rate of 11 percent.
The Democrats' list, completed just a little over a week ago, has not been thoroughly analyzed by The Times. But assuming any errors would be distributed evenly across the list, its error rate would have to be astronomical — about 74 percent — for Rossi to emerge on top. A preliminary check indicates an error rate only slightly higher than that of the Republicans.
So even adjusting for errors, tossing out felon votes alone won't make Rossi governor.
State Republican Party Chairman Chris Vance responded to The Times' analysis by emphasizing the other elements of Rossi's case.
"I'm not disputing the factual things that you've found, but you've got to place it in the context of the case," Vance said. "The felons are just one aspect of the case."
Republicans in the past week have emphasized other problems with the election. They called a press conference to allege fraud in King County, then submitted a court filing late Friday that both underscored election errors and shrunk the number of alleged felon votes to 883.
Facing high hurdlesThe math is only one of the Republicans' challenges. Proving the accuracy of their lists will be extremely difficult for both parties.
Chelan County Superior Court Judge John Bridges, who is presiding over the election lawsuit, has laid out six standards for proving a felon voted illegally: that the voter was convicted as an adult; was convicted of a felony and not a lesser charge; did not receive a deferred sentence; had not had his or her voting rights restored; cast a ballot in the November election; and voted for a gubernatorial candidate.
Meeting Bridges' standards relies on:
• Superior Court files that are sometimes incomplete. Even if some felons had their voting rights restored, proof might not have been placed in the felon's court file. In some of those cases, other government agencies, or even the felons themselves, provided proof to The Times.
• County voting records that officials admit are inaccurate, making it harder to find out if someone voted. True confirmation requires turning to poll books, which voters sign on Election Day, and digging out absentee-ballot envelopes.
• Statements by the felons themselves, many of whom are hard to find or may be unreliable.
Many had unlisted phone numbers. Two had moved abruptly, leaving landlords with unpaid rent and no forwarding addresses. Some refused to talk about their conviction or their vote. One confirmed his identity — then said he couldn't remember whom he voted for.
Consider the case of Tracey McCullough, who was on the Republicans' list of felon voters. McCullough has an unlisted phone number, but reporters tracked him down using the address on his voter registration. The former poll worker from West Seattle never had his voting rights restored after his robbery conviction 18 years ago. He said he voted in November — but not in the governor's race.
Bridges has acknowledged meeting all of his standards would be difficult. But he told both sides to "come up with all you have." Attorneys, he said, should meet his list of standards "to the extent these elements can be established."
That's an out the Republicans are counting on.
"If we get the poll-book information and we get all the information we can, obviously that's as much as you can do," said Rossi spokeswoman Mary Lane.
State Democrats proposed the standards hoping to show the absurdity of using felon voters to call the election.
"The point of it is, you're really guessing whether people voted, you're really guessing how they voted, if you use these methods," party Chairman Paul Berendt said.
Plenty of pitfallsBased on The Times' examination, each of Bridges' tests reveals different opportunities for error.
Among the 289 names Times reporters checked from the Republican list, 33 definitely were not felon voters. Fourteen had received a Certificate and Order of Discharge signed by a judge, which restored their right to vote. The Times found most of the discharges in the felons' court files. A couple of other discharges were never filed with the court, but the felons themselves provided copies.
Some of the 33 mistakes occurred because the people had pleaded to misdemeanors instead of felonies, a fact the Republicans had missed.
Election workers improperly recorded some of the people as having voted, and in a few cases, the felon and voter were two different people with the same or similar names.
The Times attempted to reach 134 of the people on the Republican list to ask which gubernatorial candidate they voted for. More than half could not be contacted. Of the 53 reached, 49 were felon voters; the others had voted legally. Among the 49, 25 voted for Gregoire, 18 for Rossi and one for Libertarian candidate Ruth Bennett. Two did not cast votes in the governor's race, and the three others either wouldn't say whom they voted for or couldn't remember.
Nearly all the felons reached said they didn't know they were voting illegally.
More than a decade after his last conviction for forgery, Michael Matthews was trying to contribute to civic life by voting, he said. He said he has long since kicked his cocaine habit and serves as the leader of street ministry in his Tacoma church. He and his wife went together to the Pierce County elections office to register to vote last year.
Learning his vote for Gov. Christine Gregoire was illegal "blew me away," he said.
"I really didn't know that. I thought everything was legit."
Savy Hem, a Cambodian immigrant who voted for Rossi, said King County kept sending him ballots in the mail after his 1994 extortion conviction. "They don't want me to vote, I won't vote. I won't say nothing. No problem."
Chuck Nelson, a Gregoire voter, is a convicted felon who has paid his "debt to society," he said, for his 1998 drug-possession charge. Reached through his mother, he said if felons can pay taxes, they should be able to vote.
Getting people like Hem and Nelson to discuss their votes in court could require them to testify to a new crime: voting illegally.
"The person has a Fifth Amendment right," said Bob Boruchowitz, executive director of The Defender Association, a King County public-defense group. "They would need a lawyer to advise them."
Selective listsAs the parties have gathered names for their lists, each side has accused the other of purposely looking for felons who would help its cause — in other words, those who likely voted for the opposing party's candidate.
In fact, both lists are heavily partisan.
On the Republicans' list, 75 percent of the alleged felon voters come from precincts where Gregoire won.
On the Democrats' list, names are more scattered geographically — they come from 36 counties compared to the Republicans' 13 — but they're even more heavily skewed. Ninety-nine percent come from precincts Rossi won.
Democrats say they started their own list only after it became clear to them that the Republicans were "cherry-picking," a charge the Republicans deny.
Berendt, the Democratic chairman, said it's no surprise that an analysis of felon votes now shows the "proportional deduction" method would leave Gregoire ahead.
"We intend to use the Republicans' own methodology to prove that," Berendt said.
"I think that we have discredited their theory that felon voters changed the outcome of the election, and that is why there has been an increasingly shrill voice in just the last few days, using the fraud word, raising fraud fraud fraud."
Vance, the Republican chairman, said it's wrong to analyze the felon votes without considering other illegal and improperly counted ballots.
Republicans have identified 68 other illegal votes they have included in their statistical analysis. In addition, they are looking at nearly 2,000 other ballots that are in question, including provisional and absentee ballots that couldn't be accounted for. They expect to find more illegal votes there.
"We think we win on either illegal votes or neglect, fraud, mistakes, errors, whatever you want to call it," Vance said. "We think we win either way."
Emily Heffter: 425-783-0624 or email@example.com
Mike Carter: 206-464-3706 or firstname.lastname@example.org
Copyright © 2005 The Seattle Times Company