Here’s another topic I’ve been hearing a lot about and I have a strong opinion on. We’re hearing this type of thing in lots of places, but I’ll just highlight one. This is from the quite excellent podcast Election Profit Makers:*
David: This New York City mayor election isn’t going to be over for months, by the way. I just want to put that out there— with the ranked-choice voting and everything, it’s going to be such a mess. They’re not going to know who the mayor is until, like, Christmas Eve, I bet.
Starlee: Really?
John: There’s going to be lots of ways.
David: I think it’s going to take at least a month. That’s my prediction.
Starlee: I still don’t understand it, the ranked-choice voting stuff.
No, no one understands it apparently, and comedians are having a field day making fun of how crazy and messy and silly it all is. The Daily Show did a whole segment on it, and that’s just the longest parade of jibes I’ve been subjected to in the past couple of months. The message is constant and clear: this a terrible idea that we should all laugh at. And I certainly listen to what my media overlords tell me to do.
Well, most of the time. Because this happens to be an area that I have some personal knowledge of. You see, I used to work on electronic voting systems. Back in those days, we called it “instant-runoff voting,” but it’s the same system. Not only is it trivial to understand, but it’s actually quite good for our democracy. Hasan Minhaj puts it best in this episode of Patriot Act (I encourage you to watch the whole thing, but this quote occurs at around 13:12):
Winner-take-all creates two-party systems. You can’t afford to waste your vote, so you stop voting for candidates who reflect your values, and you start voting for ones you think can win. But when everybody does that, we end up with just two huge mega-parties, even though 57% of Americans want a third party. Think about the way we treat people who vote third party. You’d be like, “Dennis, who’d you vote for?” and he’s like “Gary Johnson.” And we’re like “Dennis! what the fuck are you doing, man?” We treat them like they just left a baby in a hot car. We’re like “what were you thinking?!?”
In fact, I constantly vote for third-party candidates, but that’s primarily because I refuse to let the two-party system win. “Doing the math” and avoidng third-party candidates is what allows the Democrats and the Republicans to maintain their stranglehold on our political system. And you can call that a “conspiracy theory” if you like, as long as you acknowledge that this “conspiracy” is an open secret that is enabled every election by millions and millions of people. So my innate stubborn streak demands that I give the middle finger to all that shit. But, to be fair, I also have the luxury of living in a state where my second choice always wins, so it doesn’t matter who single-little-old-me votes for. If I lived in a more contentious locatio
But IRV (or, going by its new name, “ranked-choice voting” or RCV) solves all that. With this system, if your #1 choice doesn’t have a chance in hell, that’s fine: your vote for the #2 choice still matters. So I really don’t get why the media heaps all this derision on the whole concept. (Of course, I never understood why the media heaped all their derision on Bernie Sanders either. I mean, I understood why the Democratic Party did, and certainly some of that bled over into the media coverage, but you would think at some point someone would have to have the guts to stand up and say “hey, the idea that no one should have to die because they can’t afford health insurance is not a crazy idea that we should be laughing at” ... but that never happened. Colbert couldn’t do it, Poundstone couldn’t do it, Kimmel and Fallon and Meyers couldn’t do it, and they’re all pretty famously liberal icons. Trevor Noah came the closest, but I suspect that he’s about as anti-Democrat as he is anti-Republican: presumably due to his South African perspective. Of course, Minhaj posits that anti-Bernie sentiment is also due to the winner-take-all syste
The concept that IRV/RCV is complex for the person voting is just mind-boggling to me. What’s your favorite food? Okay, now what’s your second-favorite? In other words, if you couldn’t have your first favorit
We could make a stronger argument that it’s complex at the other end, the part where you figure out who won. But, the first thing to note is, you the voter don’t have to understand that part. You vote, and then the winner gets announced. Forget any ranked-choice anything: how much do you understand about voting “the old way”? Do you know how write-ins work? Do you know what a contested ballot is? Do you know the technical details of how the votes are tallied? Sadly, these things are getting more and more media attention as voting becomes more and more contentious, but I’ll still posit that most of you don’t know those things, and even if you think you do because you saw a news story about it, you probably still don’t, because the news story was likely wrong. Also, it doesn’t matter whether you know the things or not: the winner is who the winner is, and, unless you’re one of the few people who has a political or legal connection to those election results, your knowledge or lack thereof makes exactly zero difference.
But let’s say we want transparency in our democracy, because transparency is always good, and so we want to understand how the results work even though we don’t have to. Okay, fine: here it is. You count everyone’s #1 votes. Their #2 choices and #3 and so forth mean absolute squat. You only look at the #1’s. Does the person with the most votes have a majority (that is, more than 50% of the vote)? If so, you’re done. If not, all votes for the person with the least votes are eliminated. If anyone picked that person as their #1, then their #2 is now their #1, and so fort
This is not a complex process. If I wanted to adopt a less conversational tone, I could have used fewer words, but, even so, it’s pretty short. IRV/RCV is about as “complex” as a baking recipe: there may be a lot of steps, and you have to do every step just so, but there’s nothing particularly difficult to grasp here. It’s not calculus, or physics, or computer science. Hell, I would consider most sports to be more complex than this stuff: try explaining to someone how basketball works in as few words as I just used. Can’t be done, unless you leave out a lot of relevant details (i.e. the difference between a two-point shot and a three-pointer, or how fouls work). There’s no details left out of the above explanation. That’s literally all there is to it.
So what about this question about how long it takes to figure out the winner? Well, first we should note that that, despite David Rees’ dire predictions, it did not take “months” for the winner of the New York City mayoral primary to be announced. In fact, it took exactly two weeks (the primary voting closed on June 22 and the final results were announced on July 6). And I would argue that it only took that long because the board of elections had a pretty major fuck-up in that time. But suppose you think that even two weeks is too long to have to wait. After all, we live in a culture that demands everything be faster: we want it all and we want it now. One of my favorite observations on our modern world comes from science populist James Gleick’s book Faster:
Federal Express sold its services for “when it absolutely, positively has to be there overnight.” In the world before FedEx, when “it” could not absolutely, positively be there overnight, it rarely had to. Now that it can, it must.
This is becoming more and more problematic with elections, because we’ve never known the results right away. There’s a reason why “Dewey Defeats Truman” is one of the most famous photographs of the 20th century. It’s supposed to be a cautionary tale about how our obsessive need for speed can lead us into false conclusions. But somehow it’s become a meme about how newsapapers are stupid, and then we go back to throwing fits when we can’t find out who won the presidential election for a whopping 3½ days. And the weird part is, most of this blowback is because of the rejection of electronic voting. One of the major benefits of electronic voting was that we could get the results faster. So most locations implmented that, and then people got used to getting results almost instantaneously. And then there was this big backlash against electronic votin
* Specifically, from episode #95 (“All Hail the Harmonica Ripper”) which released on 5/25/2021 (starting at around 21:50).
** Or do I?
As a huge voting methods nerd, I kind of hate IRV and sort of resent that it has become synonymous with "ranked choice" when, of course, lots and lots of other methods can also use ranked choice.
ReplyDeleteMy short answer for why I dislike IRV is that it works great as long as two parties are dominant and you want to just avoid the spoiler effect. But as soon as a third option becomes a legitimate threat to the top two, the results get erratic, and not in a good way.
A decent youtube clip summarizing some of the issues:
https://www.youtube.com/watch?v=JtKAScORevQ
The person who made that video likes Approval Voting, which I don't like for other reasons, but the criticism of IRV is entirely valid.
My personal favorite method is anything Condorcet compliant. I don't really care how cyclic ambiguities are resolved because they don't happen in real-world elections, basically ever. If I designed the ballot it would be graded ballots (so you give candidates A, B, C, D, or F ratings) and in the vanishingly unlikely event of a cycle you would pick the candidate in the cycle with the highest GPA.
Yes, I know many people prefer Condorcet. However, I feel that the charges of complexity which don't make much sense for IRV actually have some weight when we're talking about Condorcet. And that means (most likely) that it would have even _more_ of an uphill battle to acceptance than IRV. "Condorcet is better" is not a reason to not implement IRV; I'd rather see IRV become more popular and _then_ push for Condorcet as an upgrade once people become comfortable with the general concepts.
DeleteAs an aside, the difficulty and delays in tallying votes is not an issue with ranked choice - it's specific to IRV. It's because IRV votes are not "summable" - you can't condense one precint's votes into a single tally and pass it up the chain, the way you can with essentially every other voting method. (With a Condorcet method, the votes of a precint can be distilled into a single matrix that contains the totals for votes in each pairwise contest.)
ReplyDelete