Spin Control: Election results do add up, if you dig into the details

Most people only pay attention to one bit of math in elections: Who got the most votes?

But a few careful readers who look at the results that we publish, and are quicker at math, call each year with another question:

Why don’t the numbers add up?

For example, on the secretary of state’s election results web page on Friday morning for the U.S. Senate seat, Maria Cantwell had 1,765,880 votes and Susan Hutchison had 1,251,920 votes. That equals 3,017,800 votes statewide, not including write-ins, the tally sheet says.

Initiative 1631, for the proposed carbon fee, had 1,311,361 yes votes and 1,705,248 no votes, for a total of 3,016,609 votes statewide – more than 1,000 votes less than the tally for senator, and you can’t write in a choice on a ballot measure.

So even without write-ins, there were more votes showing in the Senate race than in I-1631. This could lead the suspicious to wonder who stole votes for the initiative or who is stuffing the ballot box for Cantwell.

The answer is no one. The write-ins are only counted in rare circumstances, so your brother-in-law or drinking buddies will never know you voted for them for some office. But also missing from basic counts are the undervotes and overvotes.

Overvotes happen when a person fills in more than one oval in a race, and that’s pretty rare. Undervotes happen when someone doesn’t mark the ballot in a particular race and are very common. Voters can’t decide between two good candidates or pick the lesser evil between two bad ones. They know nothing about either and don’t want to pick wrong. They are most common in races with only one candidate, particularly judicial contests, when many voters say, “Why bother?”

Undervotes can become important in races so close that they require a hand recount, because sometimes a voter did pick a candidate but the machine didn’t register it if the oval wasn’t properly filled in. For those undervotes, a human eye can be a better judge than an optical scanner.

There is no race in Spokane County close enough to require a hand recount, but several legislative races on the West Side are heading that way. If so, election workers will review each ballot under the watchful eyes of at least one Democrat and one Republican, and ballots that registered as undervotes but were actually marked can turn up. They can also be controversial in cases where only a part of the oval is marked but the rest of the ovals on the ballot are solidly filled in, leading to a question of whether the voter intended to vote for that candidate or changed his mind and decided not to.

In the “good old days” of poll site voting, Spokane and many other Washington counties used punch-card ballots with tiny perforated squares that lined up with candidate names and ballot measure options. Sometimes a voter wouldn’t push through the square for a candidate or issue hard enough to knock it all the way out. So the square might wiggle back into the space when run through the machine and register as an undervote.

The punched-out squares were called chads, producing the term “hanging chad” that the 2000 presidential recount in Florida made infamous.

Usually, undervotes don’t matter. There may be as many as 40,000 of them statewide in the Senate race and 50,000 in the I-1631 contest, but neither would change the result.

In some cases, however, they can represent a lost opportunity for a candidate or issue that comes up just short.

Consider Spokane’s 6th Legislative District legislative race between Jenny Graham and Dave Wilson. As of Friday morning, Graham was up about 740 votes and barely out of recount range. But 2,684 voters didn’t pick a candidate in that race, even though they cast ballots and presumably made choices in other contests. Had Wilson closed the deal with even half of them, he’d be ahead; if Graham had convinced half of them she was the better candidate, the race could’ve been over Tuesday night.

The best example of missed opportunity this year is the Spokane County assessor race, where Republican Leonard Christian was about 500 votes ahead of Republican Tom Konis on Friday. Admittedly, it can be hard to get voters deeply engaged in the assessor’s race, let alone explain what the assessor does (two words: property taxes). But there were about 3,500 write-ins and more than 46,000 undervotes. Some of the latter might be disgruntled Democrats who would never vote for a Republican, but there’s a good chance some of them might have been persuaded with a bit more effort.

It is, after all, more than twice as many undervotes as the Spokane County sheriff race, in which incumbent Ozzie Knezovich was running against an opponent who had died after the primary.

The state Supreme Court race between incumbent Stephen Gonzalez and Nathan Choi had the most undervotes of any contested race in Spokane County, at 24 percent or more than 51,000. The race with the least is no surprise. The hotly contested 5th District congressional race between incumbent Cathy McMorris Rodgers and Democrat Lisa Brown was skipped by only 1 percent, or about 2,200 as of Friday.