The Ballot Botch

If ever we do choose a governor, his or her first priority ought to be systemic reform of state elections.

It's hard to disagree with Democratic King County Council Chair Larry Phillips when he says: "More work has to be done to raise the level of confidence that the system is working." A glitch in the county's voting system that Phillips personally discovered this week could turn him into a political kingmaker, reversing the tentative outcome of a historic and excruciatingly narrow gubernatorial race by putting fellow Democrat Christine Gregoire in the Olympia mansion. On the other hand, who can argue with Mary Lane, the state Republican Party's communications director, when she describes some King County vote-counting results as "bizarre . . . and it always seems to happen in Christine Gregoire's favor." The stunning discovery this week of 573 "wrongly discarded" county absentee ballots—likely to favor Gregoire in Democrat-happy King County—is the reason everyone from Republican Gov.-elect Dino Rossi (he's won twice—and counting) to Joe Sixpack had to be thinking this voting thing flat sucks.

We're not Florida, thank God, and nobody's asking for too much perfection in an openly democratic process that for the most part works surprisingly well, given its limits. King and the state's 38 other counties tally their votes using sometimes balky, unaligned, and myriad tabulating methods. We have touch screens, image readers, punch cards, and what now seems to be emerging as the most reliable new technology: slow, deliberate human hands. But the King County Department of Elections' new absentee screwup came just when the state Supreme Court brought some order and confidence to the postelection chaos this week, effectively deciding not to get involved (it agreed there should be no statewide recount of other previously rejected ballots, as is historically the case). King County elections director Dean Logan said the 573 absentees in question were put aside for lack of reliable signatures. But it turned out election workers failed to retrieve on-file signatures for comparison and validation—for the most part, a mental error. "We take full responsibility," Logan adds. Indeed. But is there something to be done about this?

Is it time for widespread systemic reform of the county and state vote-tallying procedures? As Michael Burton, an Ohio University expert on ballot counting, recently told Seattle Weekly: "My fear is less a stolen election than good-faith human error causing a crisis" (see "When It Doesn't Add Up," Nov. 17). Jim Adler of Bellevue, CEO of VoteHere, says his company's new e-vote technology can provide the solution, giving each voter a receipt at the polls that can also be used to verify online that the vote was counted.

That wouldn't necessarily prevent the error discovered by Phillips, and it's clear from the numerous complications of vote-counting here and in other states (for a national roundup of headaches, check out that there are systemic issues as well as technological ones that need to be addressed—whether by enacting better procedures for the evaluation and handling of questionable ballots or ensuring that the procedures we have are uniformly followed statewide.

At the state Supreme Court in Olympia, protesters held a vigil outside on Monday, Dec. 13, while Democrats argued that previously disqualified ballots ought to be evaluated anew for the manual recount that is under way. In their ruling the next day, in which they declined to intervene, the justices noted that the Democrats failed to make a case. Neither the law nor the circumstances of the manual recount called for anything more than a third tally of the same ballots approved for the original Nov. 2 tabulation and the subsequent machine recount. That's a reasonable reading of this particular argument, but it's no endorsement of state election law or procedures. That there even is litigation should tell us something.

Whoever wins the governorship, one of the first things he or she should do is convene a task force of citizens, preferably including not many lawyers or professional politicians, to design—perhaps from scratch—an election system that is fair, consistent, transparent, and trustworthy. While we're at it, how about accurate? Statisticians tell us there will always be a margin of error when dealing with millions of ballots. Maybe so, but how great must that margin of error be? We can do better, and this state has the engineering talent to do it. In the past three years, there have been no crashes of commercial jetliners in the U.S., and that's no accident. The government and the aerospace industry have painstakingly refined the air-transport system to the point where risk in commercial air travel is almost too small to measure. The system is amazingly complicated, with humans interacting with technology in every conceivable way. It works because of standardized good practices and careful technological refinement. Improving the election system is not a political problem, it's an engineering challenge.

Or maybe it is a political problem. "I don't know that we've gotten to the point of needing major reform from Olympia," says County Council member Phillips. He has an avid new interest in vote security. "There were almost 900,000 votes cast in King County alone, and it's a deliberative process to count them, while, at the same time, citizens and the media want to know the outcome immediately. The [573] ballots weren't counted because of both human and technological failure—bad enough in an election year but especially so with the governor's race being so marginal. I think there are corrections we can make to be able to trust the system better and for us to verify that it's actually counting our votes—in this system, you have to wait to be contacted. I know I don't want the system replaced with something where there's no paper involved. Thankfully, that made the difference in my case."

It was only by chance that Phillips discovered his absentee ballot had been rejected. He was preparing to help his party contact rejected voters on Sunday, Dec. 12, when he spotted his own name on the list of the rejected. "I was pretty irritated," he says. He knew his vote was valid and that his signature was on file in the county elections office. He eventually reached Logan, the elections official, setting off a ballot review to see if others had been wrongly rejected. Ah-ha! There were 572 others. Phillips is now being accused by the Republicans, he says, of trying to help the Dems steal the election. With Rossi leading Gregoire by about 100 votes out of 2.8 million cast statewide, the new ballots in a county that heavily backed Gregoire could swing the election her way.

"It might very well happen and I'd be very glad if it does," Phillips says without hesitation. "I say hallelujah! that we took the time to get it right, regardless of accusations. [State GOP chair Chris] Vance and [GOP Attorney General-elect Rob] McKenna are trying to cast doubt on my vote, cast doubt on these . . . votes, going so far to accuse me and King County of fraud. That's going to stop right now! I voted on time, and voted correctly, my vote is going to count. If that doesn't happen, well, then I have been defrauded."

He sounds like any other indignant Washington voter, regardless of party. Even Gov. Gary Locke, a Democrat, was calling the ongoing recount saga a "mess" this week. Secretary of State Sam Reed, a Republican, was hoping to reassure the public, announcing that two members of the Election Assistance Commission, a federal elections authority created after the debacle in Florida in 2000, have arrived to observe the manual recount. And King County's Logan was asking for calm while allowing that we were on "a roller-coaster ride." Somehow, referring to it as a carnival event seemed, for a change, an accurate account.

Oddball Election Outcomes

In 1989, a Lansing, Mich., school-revenue proposition failed when the final recount produced a 5,147 to 5,147 tie. In the original count, votes against the proposition were 10 more than those in favor. The result meant that the school district had to reduce its budget by $2.5 million.

In 1994, Republican Randall Luthi and independent Larry Call tied in a race for the Wyoming House of Representatives, with 1,941 votes each. A recount produced the same result. Luthi was declared the winner when, in a drawing before the State Canvassing Board, a Ping-Pong ball bearing his name was pulled from the cowboy hat of Democratic Gov. Mike Sullivan.

In 1997, Vermont state representative Sydney Nixon was seated as an apparent one-vote winner, 570 to 569. But Nixon resigned when the state House determined after a recount that he had actually lost to opponent Robert Emond, 572 to 571.

In 1997, South Dakota Democrat John McIntyre led Republican Hal Wick 4,195 to 4,191 for a legislative seat. A recount showed Wick the winner, 4,192 to 4,191. The State Supreme Court, however, ruled that one ballot counted for Wick was invalid due to an over-vote. This left the race in a tie. After hearing arguments from both sides, the Legislature voted 46 to 20 to seat Wick.

Source: U.S. Election Assistance Commission

comments powered by Disqus

Friends to Follow