Early voting has come to NYC! This is Ballot Question Number 1:

Ranked Choice Voting (AKA Instant Runoff Voting or IRV)

In 2005 the city of Burlington, Vermont voted to install Ranked Choice Voting for mayoral elections. After a controversial election and a “lame duck” mayor, Burlington reversed itself and became the first locality to overturn Ranked Choice Voting. What happened?

Election Reforms

This is the first year city residents get early voting polls as part of an election reform package that makes it easier for New Yorkers to vote. The first ballot question fundamentally changes the way city conducts elections.

A yes vote is a vote in favor of amending the city charter to do the following:

- establish ranked-choice voting to be used for primary and special elections beginning in 2021

But what is Ranked Choice Voting?

Its easiest to explain in a video with cute animals:

The current system of plurality voting, also known as First Past the Post, has voters select one candidate from the pool of candidates. The system incentivizes strategic voting where you pass your votes not to your ideal candidate, but to the candidate who has the highest chance to beat the candidate you least like. First past the post always devolves into a two-party system in the long run, and is blamed for the wide-spread voter apathy and disillusionment. It is simple in execution, but difficult for the voter to express their true preferences.

Vermont’s troubled mayoral election

But acknowledging a system is flawed and fixing it are two very different things. Burlington, VT offers a stark example of why Ranked Choice Voting matters and what challenges could be ahead for voting reformers. In the mayoral election with five candidates, two of which were eliminated in the first round, the Progressive Party and incumbent gave a strong competition to the Democrats and Republicans.

In the penultimate round the tally looked like this:

In this scenario, under the plurality system the Republican candidate would have won, but he had only a minority share of the vote with 63% of the voters voting against him. How Ranked Choice Voting solves this conundrum is to record voters second choice, and third choice etc. Then when the votes are tallied, the candidate with the smallest share of the vote gets eliminated from the round and their voters are split up by their preference instantly. In this case Democrat voters expressed a 2:1 preference for the Progressive Party candidate. The final tally looked like this:

Any voters who abstained from ranking further are treated as if they hadn’t voted in the final tally. After all, when their candidate lost, there was nowhere else for them to go. This gives the Progressive candidate the lead and the win. When voters were surveyed, most expressed satisfaction with the results.

The Republican party in Vermont, however, was fuming at the lost to what they considered the worst option. Republicans lobbed two criticisms. First, the voting system was too complex. Had their voters known that their candidate would lose in the final round, they could have voted strategically. Second, the voting method failed the Condorcet Criterion.


The Condorcet Criterion is yet another election evaluation metric that mathematicians find favorable. The Condorcet winner is the person who would win a two-candidate election against each of the other candidates in a plurality vote. In this case, since nearly all the Republicans would rather have the Democrat than the Progressive, and the same was true for all the Democrat’s voters, the Democrat is the Condorcet winner. Importantly, a Condorcet winner does not always exist, and there are few voting methods that satisfy the Condorcet Criterion and are also simple enough to implement. Approval Voting is one method of voting and while it does not always satisfy the Condorcet Criterion, it comes the closest. This is how Approval Voting works:

So taking a look at the results again, the final three candidates each would have won their race had the right voting system been in place. The Republican won in First-past-the-post, the Progressive won in Ranked Choice Voting (IVR), and the Democrat won in Approval Voting. As a matter of fact there are multiple other evaluation metrics that mathematicians find useful and there are multiple ways to rank and decide a winner. And each method contradicts the others.

Theres no perfect voting method. So do we throw up our hands. Well it is important to take a step back and look at what we gain and what we don’t from switching New Yorkers to Ranked Choice Voting. First, Ranked Choice Voting gives voters a choice they can live with. Voters who sat out elections because their candidate wouldn’t put an (R) or (D) in front of their name, can truly vote their conscience. Second, it can change the tone of the campaign. Politicians are not just jockeying to be your first choice, but also your second choice and your third choice. Third it ends the spoiler effect, when a third party does enter the race, it is always to the detriment of whichever major party that candidate is closer to (so much so that the opposition major party may even fund the third party to help win). RCV makes it safe to vote third party without helping the opposing candidate.

What Ranked Choice Voting doesn’t address. It’s not proportional. There is only one winner. Sorry. It does not truly promote third parties. Even though it wasn’t one of the two major parties that won Burlington, it would still trend down to two over many elections unless something really shook up voters. It does not eliminate strategic voting. Republicans in Burlington could have ensured the lesser of two wrongs won by shifting their votes to the Democrat. Last we could choose a voting method that better satisfies the Condorcet Criterion but don’t. Why not? Well the Condorcet Criterion favors centrists.

I think it’s important to note that when you read the criticisms of Ranked Choice Voting it is usually from the right even though all the criticisms could have (and indeed have been in the past) made by the Democrats. If we look at their top criticisms again, both the complexity issue and the Condorcet Criterion could be alleviated by moving to the Approval Voting method.

Let’s look at what happened in Burlington, VT. In 2010, a wave Republican year, voters repealed Ranked Choice Voting and went back to a First Past the Post system (with a runoff election if candidates don’t reach 40%). FPtP is the only system their candidate could have won with the help of the spoiler effect (notably Wright the Republican candidate did not get 40% of the vote but in a low turnout runoff against Kiss, anything is possible). So I see these criticisms as motivated to go back to the flawed FPtP system, including the weird Condorcet criticism. Republicans as they are now don’t bother to appeal to the center and changing to a Condorcet method would certainly hurt their chances in the short term. Which is why no one is advocating for it.

So while Ranked Choice Voting is flawed, it is almost strictly better than the current system for the simple reason that it has more information to go on about how voters feel. Its detractors mostly seem vested in the status quo. New Yorkers should look at all their options but they should rest easier if they voted YES to Ballot #1: Ranked Choice Voting.

Political Data Analyst. Professional experience in statistical models and surface and air microbiology.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store