Psychologically, losing an election by just a few votes is far more disturbing than losing by a landslide. Candidates who fall just a few votes short forever wonder what it was they could have done differently that would have secured victory. Often, the answer is: “Nothing. You came as close as you did because you did virtually everything right.” It’s an answer the head accepts, but the heart rejects. Almostland is a vale of what-ifs that haunt and torment.
Almostland is where county commissioner candidate Jay Scott now finds himself. In an election in which 13,665 votes were cast, he’s just 17 votes behind Gary Krueger, a margin of 0.12 percent. That qualifies for a recount, although Scott apparently would have to bear the cost, an unfair requirement:
13-16-201. Conditions under which recount to be conducted. (1) A recount must be conducted if:
(a) a candidate for a precinct office or for a county, municipal, or district office voted for in only one county, other than a legislator or a judge of the district court, is defeated by a margin not exceeding 1/4 of 1% of the total votes cast or by a margin not exceeding 10 votes, whichever is greater, and the defeated candidate, within 5 days after the official canvass, files with the election administrator a verified petition stating that the candidate believes that a recount will change the result and that a recount of the votes for the office or nomination should be conducted;
But what are the odds that a recount would result in victory for Scott? That’s not easy to calculate — Scott should consult a statistician — but in my experience a 17-vote margin is not easily overcome in elections employing well maintained and properly operated optical scanning vote counting machines. These machines have a very low rate of random error, probably one in a thousand votes if that high. There’s a fair chance that nine votes could flip — but they could flip for Scott, or against him. Five votes might move to his column while four move to Krueger’s column, leaving Scott with a net gain of one — and he would lose by 16 votes instead of 17.
There are, of course, other factors that can change the count. If the oval that the voter is supposed to blacken is not blackened enough, or circled, or underlined, or the choice is clearly marked in an unconventional manner, the machine will not record the vote. But such votes are valid and will be revealed in a hand recount. Is it possible that Scott received 18 votes not recognized by the counting machines? Sure. Is it probable? Don’t bet the homestead on it. Don’t even bet a burger and beer.
Other sources of error include a misprogrammed counting machine, a systemic error of some kind that skewed the count in Krueger’s direction, and fraud. And ballots can be challenged. Again, these things are possible, but not probable.
I doubt a recount will change the outcome of the election.
Still, Scott may decide to request a recount. The hard numbers of a verified loss may bring him needed peace of mind by erasing any lingering doubt he might have about the accuracy of the initial count. But a verified loss will do nothing to relieve him of the angst of Almostland. Those doubts and what-ifs can only be buried when one loses by a landslide.