Why drawing lots is a sensible way to settle a tied election

Let’s imagine a simplified scenario.

You have a town with 100 adult residents, each of whom is allowed to vote for the town’s mayor. One year, the result is an even split, 50-50. There is no winner. So what do you do?

You could have the candidates compete in some way, like playing a game of pool. Of course, that doesn’t have anything to do with the job at hand (much less the ability to campaign, which is really what elections measure). It certainly isn’t democratic.

You could also hold another election. It seems likely that, in that case, the results might end up the same, given the initial split, which would simply start the problem over again.

So you keep having elections, until the vote is different. Why is it different? Maybe someone would decide not to vote or not be able to vote, making the result 50-49 and giving us a winner. But is that a better result? It’s a more definitive result, but particularly in the event of someone being unable to vote, a less representative one. Because someone couldn’t make it to the polls, the person that voter opposed is now the mayor?

The results might also change because people change their minds. This happens, obviously. One might argue this is a better representation of the will of the people, since people have to consider the question again. But . . . is it? Part of the democratic system means narrowing a lot of issues down to a choice between a handful of candidates that is then determined at a random-but-specific time. Had the Senate election in Alabama happened Nov. 1, it’s likely Republican Roy Moore would have won. Revelations about his past conduct probably contributed to his defeat, but would a Nov. 1 election have been less democratic or representative than a Dec. 12 one? Had the election been kicked out to this week, wouldn’t that have been more representative still?

There’s a practical consideration that town would need to consider, too. Elections are not free. They take time; they cost money. For a town of 100 people, that is probably not a big deal. For a district of, say, tens of thousands of people spread over multiple municipalities, the costs are larger: Staff time, polling places, distributing machines, etc.

In the case of one such district — Virginia’s 94th District in the House of Delegates — the determination of a tie in a November race meant the state enacted its tiebreaking system: Literally drawing a name from a bowl. (The Republican, David Yancey, won.)

Many of the responses to that name-drawing have lamented it as being fundamentally undemocratic. It isn’t.

In the case of our small town, we have one half of the population which prefers one candidate and one half which prefers the other. Drawing lots in that case does precisely and unequivocally what we described in the paragraph above: It breaks the tie.

Compare that with the other options above. Is drawing a name less democratic than hoping one of the 100 people who voted the first time around doesn’t vote this time? Is holding another election in the context of a prior tie necessarily a better reflection of which person people want to see serve as mayor than the 50-50 split the first time around? Why is a non-tied result, however it is obtained, necessarily better than a tie in which the result is settled by a coin toss?

In actual elections, though, everyone doesn’t vote. That makes things more complicated.

This all draws attention to the broader fiction that envelops our elections. We spend a lot of time counting votes and obsessing over percentage-point differences between candidates, and that’s all well and good . . . until the results are close. At that point, all of the edge cases — improperly completed ballots, weird issues at the polls, people who didn’t vote — move into the spotlight and make our elections broadly seem as though they’re like throwing a dart at the wall. There has always been a gray area around those hard vote totals in the way that zooming in on a photograph makes it seem blurrier and blurrier. When an election is close, that’s when the gray area becomes important and scrutinized. When an election is close, it makes us consider turnout: If 60 percent of an eligible population voted in an election, how do we know the victor is actually the preferred candidate of the majority of that population?

Our elections necessarily include a hard-to-measure margin of error. Normally it doesn’t matter. Sometimes it does.

It’s analogous to how first downs are measured in football. Referees eyeball where the ball-carrier was when he was tackled and usually it’s obviously well past or well short of the first-down marker. In some cases, though, it’s close, and that’s when the fiction of precision evaporates: People run out from the sidelines with a chain and drop the markers in about the right place depending on the chain and how they ran out and so on. The referees then cogitate and chin-stroke over whether the tip of the ball placed in the approximate place it was downed is past the end of the chain that was placed in its own approximate place. It’s precision theater, just as is evaluating hanging chads or, in the case of that Virginia election, mismarked ballots.

The Virginia race, like any close race, is littered with asterisks and what-ifs. Should the Democrats not control a chamber for which they received far more popular votes? Should that one ballot have been counted? Etc., etc. There are debates to be had about all of that, certainly.

More broadly, though, the pick-a-name-from-a-bowl solution is a good one precisely because it recognizes the imprecision of the electoral process. The race was specifically a tie after all of the votes were counted, but it was effectively a tie on the night of the election. Voters in Virginia’s 94th District were essentially split on who they wanted to have represent them in the legislature even if one or the other of the candidates had a few more votes one way or the other. Not all close races should be settled by drawing lots, of course, but there’s nothing wrong with using that system to settle races in which the generally arbitrary accumulation of votes resulted in a tie.

Elections are imprecise, which only matters on occasion. In the event of an actual tie, it’s better to accept elections are imprecise and we can’t perfectly measure the will of the people than to continue to pretend they aren’t and force a different — and equally imprecise — result.