Friday, September 22, 2017

What is Ranked Choice Voting and How does it Work?

After Minneapolis first adopted and used Ranked Choice Voting (RCV) in its 2009 elections, the
Minneapolis Elections Bureau asked me to do an evaluation.  In that report, I drafted a description of RCV in terms of the pros and cons and how it is supposed to work.  Now that the early voting in Minneapolis and St. Paul has begun, I decided it was time to reprint this section of the report. 

What is RCV?
In November 2006, voters in Minneapolis adopted RCV as their preferred method for running elections and selecting candidates in their city.  The decision to make this switch came as a result of several years of lobbying and mobilization by individuals and groups such as Fair Vote Minnesota (FVM).  The voting mechanism that RCV voting replaced is known as first past the post (FPTP) or simple majority rule.

FPTP is a simple voting system.  It states that whenever there is an election for a single member seat, such as for mayor or a city council seat, the candidate who receives the most votes wins the seat.  Thus, in a two-person race, the candidate who wins 50% of the votes plus one wins the race.  In races with three or more candidates, the one who receives the most votes wins, even if that candidate wins less than 50% of the total vote.  In non-single member races, such as where two candidates are to be elected, the top two vote getters are elected, again whether or not they receive more than 50% of the vote. 

In voting, citizens are allowed only one vote for each office holder to be selected.  If voting for mayor, each eligible voter is allowed to case one vote.  In multi-member races, such as when citizens are allowed to elect two school board members, each voter is allowed to vote for up to two candidates.  They may not cast more than one vote for a specific candidate.  FPTP is often described as simple majority rule—yet, as noted above, there are no requirements for a candidate to receive a majority of the vote unless that is an additional stipulation imposed by law.  Minneapolis prior to 2009 employed FPTP in all of its elections.

There are several criticisms of FPTP.  One is that the lack of a majority requirement means that candidates elected in single member districts may take office even though only a plurality of voters cast ballots for them.  The concern here is that candidates do not enjoy the legitimacy of majority support if they take office with less than 50% of the voters having supported them.

A second criticism is that FPTP creates disincentives for third party candidates to run for office and get elected.  The reason for this is that voters, fearing that a vote for a minor or third-party candidate would effectively result in the candidate they desire the least being elected.  Thus, third party candidates are viewed as spoilers who voters are reluctant to support.  For example, in the 2000 Florida presidential election, many argued that Ralph Nader was a spoiler.  By that, some argued that votes cast for Nader came from otherwise Democratic votes that would have gone to Al Gore had the former not run.  Thus, Nader votes effectively produced a victory for George Bush, the candidate some contend who was the last choice for Nader or Gore voters. 

While there is mixed evidence that Nader served as the spoiler as described above, the argument is that in general voters will not vote their first choice (a third-party candidate) if they perceive that the candidate does not have a real chance of winning, for fear that it would lead to the election of the candidate they least desire.  This in turn means that it will be difficult for third party candidates to run, thereby insulating Republican and Democratic parties as the two major parties.  Thus, FPTP decreases incentives for voting for third party candidates and effectively limits voters to a choice among the two major parties.

RCV is supposed to address many of maladies of FPTP.  At its most basic, RCV involves three candidates, A, B, and C, all running in a single member district.  Voters are asked to rank choice their candidates, indicating on a ballot who their first, second, and then third choices are.  In order to be elected, a candidate needs to win 50% plus one of all the votes cast.  Assume that there are 100 voters, one could have this scenario.

A receives 51 votes.
B receives 30 votes
C receives 19 votes.

Under this scenario, candidate A is the winner.  However, assume a different situation where

A receives 45 votes.
B receives 30 votes
C receives 25 votes.

Under this scenario using RCV, no candidate is initially declared the winner.  Instead, the candidate with the least number of votes—C—is dropped from the ballot and her votes are transferred to A and B based on the expressed preference of the voters when they ranked their preference for candidates.  Whoever then receives 51 votes is declared the winner.

RCV does not give voters multiple votes; it merely allows voters to rank their candidate preferences.  Votes are then transferred from voters’ first choice to their second choice (or third if there are more candidates) until someone receives the 50% plus one votes.  At no point does any individual voter receive more than one vote or have more than one vote counted in single-member districts. 

Benefits of RCV
Advocates of RCV contend that this voting procedure is superior to FPTP.  They argue first that RCV addresses the problem of candidates being elected with less than 50% of the vote by mandating a real majority to take office.  This addresses the issue of minority (less than 50%) candidates taking office, thereby ensuring that majority rule is actually honored.  Second, the argument is that RCV deals with the spoiler psychology.  Specifically, because voters know that they can rank choice their candidate preferences they have a greater incentive to vote their first choice, even if a minor party candidate, because the candidate cannot be a spoiler.  If no candidate in the first round of voting receives more than 50% of the vote then votes are transferred from one’s first choice (a third-party candidate, for example), to the remaining candidates.  Assume for a voter her first candidate choice is C, second choice A, third choice B, and assume:

A receives 45 votes.
B receives 30 votes
C receives 25 votes.

Under this scenario, since C received the least number of votes, that candidate is removed from the ballot and votes for her are transferred to the voters second choice.  In the case of the voter above, her vote would transfer to A.  If enough second choice ballots from those who voted for C are transferred to A then, A wins.  If more second choices go to B, then B wins.  There is no guarantee that a voter’s second choice will win, but RCV supposedly creates an incentive to support one’s first candidate choice and not view a ballot cast for her as a wasted vote.

As a result of RCV, voters are potentially encouraged to support third or minor party candidates, minority parties are therefore encouraged or given better opportunities to run and win, and voters are therefore given more choices beyond the two major parties.

Criticisms of RCV

There are also criticisms frequently directed against RCV.  The first is that this voting mechanism is biased in favor of political parties on the left.  Second, that this voting procedure is non-monotonic.   Specifically, using RCV it is possible to vote in such a way that a specific voters’ ranking that their candidate will yield election of a lesser-ranked or preferred candidate. Third, that RCV violates the one person, one vote standard for the counting of votes.  In response, to these criticisms, there is no empirical evidence that RCV favors candidates of one political persuasion or another.  Second, while RCV is non-monotonic, so is FPTP.  Third, no voters have their ballots counted more than once (by that, each voter gets only one vote per round counted in a single-member race) and there is no double counting of ballots. 

1 comment:

  1. Dear Prof DSchultz,

    My work in publishing services brought me to your profile on behalf of a client seeking considered reviews for his book.

    Perhaps you’d agree that the challenges the United States and its citizens face have not accrued because people are open-minded and proactive. Unfortunately, the bias today tends towards the unhappy idea that the country is broken and nothing can be done.

    This prevailing zeitgeist is also greatly reflected in the many books being published on the topic that promote rather negative conclusions based almost exclusively on poor outcomes that little consider the processes that lead to those outcomes; this is not useful!

    What would be useful is a discussion focused on verifiable alternatives and this needs to be encouraged.

    Thank you,
    PR Review Team