How It Works: Run-Off Voting (Yes, This Will Be On The Midterms…Well, Some of Them)

If you’re anything like me, the voting booth often feels more like a chess match in which you are running more if/then scenarios than you would in a year-long Dungeons and Dragons campaign.
Publish date:
October 23, 2014
politics, voting, elections, Midterms

Let’s say you live next door to an ice cream shop that, unfortunately, is only open one day per year. Also, they sell two flavors: the utterly repellent rum raisin* or vanilla. Very occasionally, the shop will offer a third flavor, but there are rules: 1) it is half-price and you get twice as much, 2) you don’t get to know what the flavor is before you buy it, 3) it is made in a shop where allergens are present, 4) you have a major food allergy and 5) you have to eat all of whatever you purchase.

So, what are you going to buy? Remember, it’ll be a whole year before you get to make this decision again. Will you pick: something you know you’ll dislike, something boring, or a long shot that could possibly kill you? Would you be more likely to try the long shot if you could be assured it is allergen-free and you could trade it for vanilla if you didn’t like it?

What do hypothetical once-per-year ice cream shop decisions and elections have in common? Think about the last few times you voted. How did you make your decision? Did you vote your first choice: the non-objectionable candidate with good odds? Did you cast a vote for a candidate that primarily served as a vote against another candidate?

Voting should be relatively simple: research your options and vote for the candidate who you think will best represent you. But, we all know it’s not that simple. If you’re anything like me, the voting booth often feels more like a chess match in which you are running more if/then scenarios than you would in a year-long Dungeons and Dragons campaign.

But it doesn’t necessarily have to be that way. There’s a way to vote for your first choice first and also vote for a “safety” option. There’s a way to try the new flavor and still get vanilla rather than nothing at all.

It’s called instant run-off voting (IRV to those in the know, and it’s also sometimes called rank choice); it’s already in use in at least 10 countries around the world and in 6 cities in the U.S. It maximizes the likelihood of the maximum number of voters getting their first-choice candidate. (NB: There’s another kind of run-off voting called Condorcet Voting but it’s far less common.)

Here’s how IRV works:

1) Each voter ranks candidates in order of preference.

2) All 1st choice votes are counted; if in this count any one candidate gets 50%+ of the vote, that candidate wins.

3) If no candidate gets 50% of the vote, the candidate with the fewest first-choice votes is eliminated and the votes are re-counted.

4) In this round, votes are counted again this way: first-choice votes for the eliminated candidate are re-allocated to the voter’s 2nd choice, first-choice votes for all other candidates are counted.

5) Anyone have 50%+ yet? No? Steps 3 and 4 are repeated until one candidate has 50%+ of the votes.

6) First candidate with 50%+ of the votes wins.

You need a visual about now? Good, me too!

OK, here we have a hypothetical primary election to choose a general election candidate to represent my party, The Stick Party. In the general election, the winner of this primary will face the candidate from The Squiggle Party. I have just about 0% compatibility with the Squiggle candidate, and my options are above. If you were in my position, how would you vote? If my first choice primary candidate, Tulip, wins, I am virtually assured that Squiggle will win in the general election. Top Hat has a slightly better chance; I could vote for Top Hat and the cross my fingers and/or volunteer my butt off for Top Hat’s campaign but it’s still more risky than Shoes. It’s probably too risky. The media outlets are saying that Shoes is polling 5-7 points behind Squiggle and my Stick Party friends are telling me we gotta rally around the best hope, Shoes. Shoes is the ultimate meh candidate. Sure, Shoes has a good chance of beating Squiggle, but Shoes only represents me halfway.

Sigh. Why bother? I mean, no matter who wins, I’m not really represented. Voting this way (having to choose the lesser of two evils) could be described by the psychology crowd as “loss aversion.” Simply put, loss aversion means the drive to make decisions based on not-losing something vs. gaining something. In terms of motivation and decision-making, loss aversion is much stronger than the prospect of gain.

It’s enough to make you not want to vote, which may explain why 50-60% turnout is about the best we can do in post-WWII elections. If I’m choosing from rum raisin, vanilla, or a mystery option with a lot of risk, one option is to walk out of the ice cream shop and keep my money. The first choice candidate is like the possibly allergen-laden flavor -- you want to try something new, it’s a good deal, and it’s far more appealing than the other two options. But, if you pick it, something bad might happen.

By something bad, I am of course referring to the spolier effect, or the idea that a third candidate will steal votes from one candidate, allowing another candidate to win. The modern canonical example of the spoiler effect is Ralph Nader’s candidacy in 2000; various political historians and analysts have posited that Nader’s presence on the ballot allowed George W. Bush to be elected over Al Gore. If you want a more current example, look to the 2014 election for U.S. Senate in Kansas. The Democratic candidate (Chad Taylor) requested to withdraw from the election after polls showed that a) he was trailing behind an Independent candidate (Greg Orman), and b) votes for Taylor and Orman combined could beat the incumbent Republican (Pat Roberts). How’s that for playing it pragmatically cynical?

Why would it be worth implementing IRV in more elections? Why am I writing about this when there are so many juicier election stories? This is big deal stuff. And if I learned anything about the net neutrality debates, it’s that an issue really has to seem interesting before anyone will pay attention. Election math is pretty damn boring, so this is me trying to sex it up a bit. It was quite hard to get people fired up about clean elections or Citizens United, but “corporations are not people” kind of makes people stand up and take notice. I like to think of IRV as voting for my first choice every time.

Here are some possible benefits I see when I think about IRV:

1) No spoiler effect. You can vote your first-choice without risking the chance that a 2nd or 3rd choice will lose out to a last-choice candidate if you do.

2) IRV can eliminate the need for primary elections. If all candidates were in one election, voters could rank them in order of preference. This would a) eliminate the practice of voting for the “most winnable” candidate (i.e. Shoes from The Stick Party) in primary elections and, b) save considerable time and money by only holding one election.

3) IRV might engage more voters who will then vote more often. Those who feel apathetic or disenfranchised may be motivated to vote if they had the opportunity to vote for a first-choice candidate.

4) It would make it much harder to buy elections. Right now, big moneyed interests can pour a ton of cash into essentially 50/50 odds and effectively tip the scales. When there are 3, 4, or even 10 candidates on a ballot, it becomes much harder to concentrate money behind any one candidate (or two candidates). Side benefit: mitigating the effect of money might allow people of more average means to run for office, allowing for more diversity in the candidacy pool.

5) There’s some speculation that IRV would reduce attack ads. How? Of course, more mainstream candidates want to be your first choice but they also want to be the second choice of voters who are voting for weaker candidates (who may get eliminated). Negative or attack ads might risk alienating those voters (who might penalize them with a last place rank for their negativity).

6) More representative democracy and more interesting candidates. Right now, because of our 2-party lockdown most candidates keep very bland, centrist platforms to appeal to as many voters as possible. With IRV, we stand the chance of having candidates who can afford a bolder platform without being branded a spoiler.

7) The average citizen would have more voting power. Elected officials would be more accountable to voters because they would need more first-choice votes.

Let’s take a look at our Stick Party primary, if it were done as one IRV election (no primaries):

Under our present system, Squiggle, with 40% of the votes, would have won the election, and Tulip would have been a spoiler, since adding Tulip’s 15% to Shoes would have given Shoes the win. It also means that conventional wisdom would have held true that voting for anyone other than Shoes would have nearly guaranteed a Squiggle win.

But look here – with IRV voting, the outcome is quite different. My beloved Tulip loses the first round and is eliminated. Ah, it was a longshot anyway. In Round 2, Squiggle, Shoes, and Hat keep all their votes from Round 1 and ballots that ranked Tulip as #1 now get re-allocated to #2 selections on the ballot. And it would seem that pretty much every Tulip voter designated Hat as #2 (since 15 + 17 = 32). Now the bland Shoes is out, and we’re down to Squiggle and Hat. In the final round, with both Shoes’ and Tulip’s votes re-allocated, Hat is the winner.

In Round 2, it mattered who Tulip voters picked as a second choice; a 4%+ change would have given Shoes the win over Hat. Don’t you think this would affect the stronger candidates’ campaigns and platforms? Squiggle and Shoes need to appeal to their base but they also need to appeal enough to Tulip and Hat voters to be considered second choices. It means Tulip and Hat voters get a say in influencing platforms of more mainstream candidates and it gives less mainstream candidates better odds of winning -- not by rigging anything but simply by asking voters to put candidates in order by preference.

If this is so great, why aren’t we doing it? Why haven’t we ALWAYS done it?

First, there is a long history of doing everything legally permissible to rig elections, from poll taxes to gerrymandering, to the current debate over voter ID laws. Overall, IRV is a more complex and therefore more difficult system to game.

For a while, math was a problem. IRV elections are, by nature, more complex. But now we have computers to do the math.

Now the bigger problem is that IRV poses a real threat to the entrenched parties; they don’t necessarily want to make it easier for outsider/non-party voices to get in. If you are the DNC or RNC, you aren’t interested in unseating long-term, dedicated elected officials or decentralizing power. If you are a large, wealthy corporation, you’ve learned how to game the system. If you need favors, you write a check. Maybe you write two checks (one to each party) to hedge your bets – either way, there’s an established system for buying influence and you aren’t particularly interested in having to re-learn how to re-establish that influence under a new set of rules. It should be noted that cost is not a factor.

What next?

Well, in some states there exists something called multiple party access. In this process, a candidate may appear twice on the general election ballot, under different parties. For example, in the last New York State Democratic presidential primary, Hillary Clinton was listed as both a Democratic candidate and a Working Families candidate. No matter which Hillary Clinton you selected, she got your vote. So, every vote cast under the Working Families party sent a message to the Clinton campaign about the demands of voters in New York. It also allowed Working Families to stay on the ballot.

In most states, a party has to get/maintain a certain percentage of the vote to stay on the ballot. For all the reasons outlined above, it’s quite typical that voters feel unsafe voting for 3rd party candidates, so they don’t get many votes; thus the 3rd party loses the ability to put anyone up for election. It’s a self-perpetuating cycle that limits true variety in candidates. And while adding mainstream candidate names to 3rd parties is not ideal, it would at least allow 3rd parties to perhaps influence the agendas and platforms of mainstream candidates; it would let mainstream candidates know who was voting and their primary concerns.

If you have the opportunity to vote for a candidate in more than party, cast your vote for the candidate in the party that best represents you. If you have the opportunity to speak up about getting run-off elections in your city, county, or state -- speak up!

The one thing you shouldn’t do is give up and not vote. Not voting is truly the only way your vote is wasted. As President Obama recently said as he picked up his ballot to early-vote, the role of voting citizen is the most important one in a democracy.

So, what do you think? Do you live in one of the six U.S. cities that have IRV in local elections? How do you feel about it? If you don’t, would you be more likely to vote if your elections were done by IRV?

* Yes, this is the second time I’ve used rum raisin as a stand-in for gross or unwanted. I apologize if you actually like this flavor.