NYT’s Article About Ranked Choice Voting Is More Right Than You Realize

Sass
4 min readMay 31, 2021
A ballot being placed into a ballot box with the word “VOTE” crossed out and replaced with the word “TRASH”.

New York City is about to hold an election for mayor using Ranked Choice Voting (RCV) for the first time. Some New Yorkers have been excited about it, but fears of strange quirks in the system have been creeping up, culminating in a recent New York Times article by Nate Cohn going over how a concept known as “ballot exhaustion” may cause some problems. Nate made an important point, but ballot exhaustion isn’t the only way RCV discards votes; however, there is a positive way forward.

To review, let me give you a condensed, simpler version of what Nate’s article explained.

RCV is a voting system that allows voters to rank multiple candidates (in this case up to 5) in a given race on their ballot instead of just marking preference for one candidate like normal. Then, the ballots are tabulated through a process called Instant Runoff:

  • Add up all of the top-choice rankings.
  • If no candidate receives more than half of the top-choice rankings, the candidate with the fewest top-choice rankings is eliminated from the race, and voters who ranked that candidate as their top-choice get their vote reassigned to the next-highest choice on their ballot who is still remaining in the race, if any. We’ll call this their new “top-choice”.
  • Repeat until a candidate receives more than half of the remaining top-choice rankings. That candidate is declared the winner

(If you’re familiar with RCV, you’ll notice that my description is a bit different from what you typically hear. That’s because it’s more accurate and you’re about to see why.)

This all seems fine and dandy, but let me present you with a situation you might not have considered. Imagine the candidate you ranked first on your ballot is the second-to-last one eliminated in the semi-final elimination round (third place, effectively) and you didn’t rank either of the two finalists at all. All the other rankings that you expressed on your ballot (up to an additional 4) are thrown away and are never used in any capacity to decide who the winner should be. Additionally, you have no say in the final elimination round — when the winner is actually decided.

A two-panel comic is depicted. In the first panel, the foreground character says to an uninterested election official “Wait! You eliminated my first choice but never counted the rest of my rankings!” In the second panel, that same character exclaims “HEY!!! YOU’RE THROWING MY VOTE AWAY!!”

You took all this time to express yourself at the polls and almost none of that information is ever used. Your ballot is fully “exhausted” before the final round and you were never able to give any support to any of the other 4 candidates you ranked. This isn’t even the worst-case scenario, but the problem of discarded votes under RCV is a regular feature, especially in competitive races like this one. How can a voting system that throws away half of the voter preference data come up with an accurate result? It can’t. That’s why almost no other voting system has this flaw.

Enter STAR Voting. STAR is an acronym for Score Then Automatic Runoff, and it’s the ultimate culmination of modern voting science. Here’s how it works:

  • Score candidates on your ballot from 0 to 5 stars, much like an Amazon review.
  • Add up all the scores for each candidate, and the two candidates with the highest scores become finalists who move on to the automatic runoff.
  • In the automatic runoff, your one full vote goes to whichever of the two finalists you scored higher on your ballot. Whichever finalist gets more votes wins!

STAR Voting is simpler and far more expressive than RCV. STAR Voting promotes honest voting and actually uses all of the ballot data, leading it to top the charts in accuracy (which is measured with a quantifiable concept called Voter Satisfaction Efficiency).

A chart depicting the nature of Ranked Choice Voting to trend toward worse overall results than our current Choose-One voting system. The chart also shows that Approval Voting and STAR Voting both perform far better than either of the other two, with STAR performing particularly well without incentivizing strategic voting.
There’s a lot going on in this image, but if you study it closely, you’ll realize that the strategic voting incentive of RCV will, over time, lead to worse election results than our current system. In contrast, Approval Voting (a fantastic, simple method often recommended for upgrading from Choose-one Voting) and especially STAR Voting deliver *stellar* results.

And in case you’ve really been paying attention to this issue, STAR Voting doesn’t suffer from the lengthy and unsecure tabulation problem unique to RCV; STAR can be tallied using simple addition much like we always have.

For this mayoral race, New Yorkers will have to use RCV and will likely uncover other issues RCV brings to the table that I don’t want to melt your brain discussing here. Fortunately, STAR Voting is ready for implementation in NYC and I believe NYC is about to be ready for a real world-class voting system.

--

--

Sass

I’m the Operations Coordinator and a Board Director of the Equal Vote Coalition.