Ranked Choice Voting

Ranked choice voting, or RCV, entails voting by placing candidates in order of preference. A voter essentially votes if not A, then B; if not B, then C; and so forth, until all further candidates are equally-worse and thus not ranked. This works for single-winner and multi-winner elections with the voter casting the same ballot either way, although multi-winner elections require more-complex proportional rules.

Two alternative styles of paper ranked ballots based on segmented digital displays.

Early implementations of ranked choice voting for single-winner elections mainly focused around Instant Runoff Voting (IRV), which operates in the same manner as a Majority system with runoffs: rather than Plurality—whomever gets the most votes wins—either a candidate receives more than 50% of votes or the candidates with the fewest votes are eliminated and a new election is held. IRV uses the rankings to count the runoff, rather than holding a new election.

IRV had some unfortunate defects carried over from single-vote systems like Plurality and Majority: given an election between two candidates, a third candidate could eliminate the winner in the second-to-last round and then be defeated. In essence, if A defeats B, adding candidate C can make B the winner—even if B also defeats C in a two-candidate race.

This actually occurred in the 2009 Mayoral Election for Burlington, Vermont, causing the immediate repeal of Ranked Choice Voting specifically because IRV outright eliminated the Condorcet candidate.

We have made major advances in recent years, and now have simple and efficient Condorcet systems such as Tideman’s Alternative, which prevents the above and resists all forms of strategy without requiring complex math and indecipherable tallying methods. Combined with a proportional nonpartisan blanket primary, Condorcet systems reliably elect broadly-representative candidates.