Why IRV

This is an essay by fairvote.org's Rob Richie, published by him in January 2007 and provided as a verbatim copy here. You can also see both Lomax's and our devastating critiques of Richie's essay. Richie has actually published several slightly different versions of this same essay in different places and our critique was written in response to a different (apparently somewhat earlier, but it has the advantage of going into more detail) one. The second Richie Essay is printed here below his first (more recent) version (2 essay-versions in one).

Introduction

Instant runoff voting is the reform community's solution to defective plurality elections for single-winner offices. Every one of the last eight ballot initiatives calling for IRV has been victorious. With newspapers and citizens alike calling for statewide IRV measures around the country, this reform has joined campaign finance and independent redistricting as a main stream essential.

IRV is not the only alternative to plurality elections for single winner races; however, it is the system FairVote feels is most appropriate for electing public officials in the United States. The following paragraphs explain the criteria on which we evaluate voting systems in order to reach this conclusion.

Instant Runoff Voting:
The Best Method for Single Winner Public Elections

There are many ways of electing officials to single-winner offices other than existing plurality and two-round runoff voting systems. But among these options, we believe instant runoff voting (IRV) offers the most politically practical and common sense option for replacing the faulty plurality voting and two-round runoff systems used in nearly all American elections. The benefits of IRV include:

Evaluating Alternatives to IRV

We evaluate other single-winner election methods on these standards, but initially through three criteria that we see as essential in measuring a method's political viability in the United States:

Alternative 1: Range Voting

With range voting, voters score each candidate: for example, they could award between 0 and 99 points to each candidate. The candidate with the most points wins. As of early 2007, range voting has not been used in any public election in the world and by very few, if any, private associations.

Bottom-line: Range voting violates all three of our common sense principles of preserving majority rule, requiring a minimum level of core support and rewarding sincere voters.

Example: Consider a range voting election in which 100 voters have the power to assign a score between zero and 99. There are two mediocre candidates. Of the 100 voters, 98 greatly dislike Candidate B, but decide to express their distaste for both candidates by giving one point to Candidate A and none to Candidate B. The remaining two voters prefer Candidate B and are more tactical. They award 99 points to Candidate B and 0 points to Candidate A. The election ends with B beating A by a landslide of 198 to 98 despite the fact that fully 98% of voters preferred Candidate A.

Explanation: This example illustrates how a tactical fringe can overrule a vast majority of voters when the majority votes sincerely and the minority votes tactically. Tactical calculations rise exponentially with the entry of more candidates, at which point winners also do not need to have been any voter's first choice.

Alternative 2: Approval Voting

Approval voting is a form of range voting, with voters limited to awarding candidates a one or zero. As of early 2007, it has not been used in a public election in the United States. The largest association to use it, the Institute of Electrical and Electronic Engineers, abandoned it in 2002 after most voters started to simply cast plurality voting-type ballots.

Bottom-line: Approval voting violates all three of our common sense principles of preserving majority rule, requiring a minimum level of core support and rewarding sincere voters.

Example: To illustrate how approval voting violates majority rule, consider a primary with 100 voters and two candidates liked by all voters. 99 voters choose to approve of both candidates even though slightly preferring the first candidate to the second. The 100th voter is a tactical voter and chooses to support only the second candidate. As a result, the second candidate wins by one vote, even though 99% of voters prefer the first candidate.

Explanation: This example shows how voting sincerely in an approval voting election will count against your first choice – e.g., if you approve of a lesser choice, you are giving that candidate support equal to your first choice, and that support could cause your first choice to lose. Voters must always be aware of which candidates might win, and candidates have every incentive to ask supporters privately to vote only for them while publicly pretending otherwise. Many voters will bullet vote (e.g., cast one vote for their first choice and no votes for anyone else), thereby reducing even further voters' ability to express their range of views about candidates. In a three-candidate race, a candidate also can win despite not being even a single voter's first choice.

Alternative 3: Condorcet-Type Rules

Condorcet-type voting rules are ones where voters rank candidates in order of choice, and each candidate is compared with every other in terms of how many voters rank one ahead of the other. If there is a candidate who beats all others in these comparisons he or she is the winner. Condorcet-type voting rules have not been used in any public election in the world as of early 2007, but are used to elect the leadership of some private associations.

Bottom-line: Condorcet-type voting violates the principle of requiring a minimum level of core support by permitting a candidate to win who would not win a single vote in a plurality election.

Problem 1: With these rules, a candidate can win without being a single voter's first choice. By putting such heavy emphasis on breadth of support, Condorcet-type systems, like approval voting, encourage candidates to be seen as the “least offensive” candidate rather than leaders who take strong positions that might alienate some voters.

Problem 2: Condorcet comparisons can yield a situation where, in an election among Candidates A, B and C, Candidate A is preferred to B, B preferred to C, and C preferred to A. In this situation, there is no winner, and a “fallback” method must break the cycle. When this fallback is needed, sincere voters can be punished. Finally, Condorcet-type rules are difficult to count by hand in big elections. Hand-counting is important if problems emerge with voting machines or software.

Scholarly Assessment of IRV

Advocates of range voting, approval voting and Condorcet voting sometimes criticize instant runoff voting for (1) being “non-monotonic” (theoretical situations exist in which improving the ranking of a particular candidate can hurt that candidate's chance of winning because it can change the order of which candidates lose for being in last place) and (2) not always electing the Condorcet winner.

IRV advocates dismiss these criticisms. Potential non-monotonicity with IRV is irrelevant in practice and will not affect voter strategy. We also believe that there are times when the Condorcet winner should not win if that candidate is so lacking in core support that he or she would never win even one vote in a plurality or runoff system. To us, being able to lead and represent people effectively makes it important that a significant number of voters rank the ultimate winner as their first choice.

Leading scholars provide scholarly grounding for our views. Here are two of many examples:



Second version of Richie's essay

There are many single winner election methods other than our existing plurality and two-round runoff voting systems. Here are reasons for why we believe that instant runoff voting (IRV) offers the most politically practical and common sense option for replacing the plurality voting and traditional runoff systems used in nearly all American elections.

Evaluating Alternatives to IRV

Range voting, approval voting and Condorcet and variations of these election methods are not used for elections for any public office in elections around the world, but have their strong advocates. Although these systems may have appropriate applications, IRV is preferable for our elections.

We highlight the following criteria to evaluate a single winner system's merits and political viability " IRV upholds all these criteria, while other leading reform options do not.

Here we evaluate other proposed single winner methods.

Range Voting

With range voting, voters score each candidate: for example, they could award between 0 and 99 points to each candidate. The candidate with the most points wins. As of early 2007, range voting has not been used in any public election in the world and in hardly any private elections of any kind.

Range voting violates our three common sense principles: those of majority rule, requiring a minimum level of core support and rewards for sincere voting. Consider an election in which 100 voters have the power to assign a score between zero and 99. There are two mediocre candidates. Of the 100 voters, 98 most dislike Candidate B, but decide to express their distaste for both candidates by giving one point to Candidate A and none to Candidate B. The remaining two voters prefer candidate B and are more strategic. They award Candidate B a score of 99 while giving no points to Candidate A. The election ends with B handily beating A 198 to 98 despite the fact that Candidate A was the preferred candidate of fully 98% of voters. This example illustrates how a strategic fringe can overrule the vast majority of voters and violate the principle of rewards for sincere voting. Strategic calculations rise exponentially with the entry of more candidates, where candidates can also win without being any voter's first choice.

Approval Voting

Approval voting is a form of range voting, with voters limited to awarding candidates a one or zero. As of early 2007, it has not been used in a public election in the U.S., but has been used in some professional societies. The largest, the Institute of Electrical and Electronic Engineers, abandoned it in 2002 after most voters simply cast plurality voting-type ballots.

Approval voting violates the common sense principles of majority rule, requiring a minimum level of core support and rewards for sincere voting. To illustrate how it violates the majority rule principle, consider a primary with 100 voters and two candidates liked by all voters. 99 voters approve of both candidates even though slightly preferring the first candidate to the second. The 100th voter is a strategic voter and chooses to support only the second candidate. As a result, the second candidate wins by one vote, even though 99% of voters prefer the first candidate. In a three-candidate race, a candidate could win despite not being even a single voter's first choice.

This example shows how voting sincerely in an approval voting election will count against your first choice "e.g., if you approve of a lesser choice, you are giving that candidate support equal to your first choice, and that support could cause your first choice to lose. Voters are forced to make decisions based on their calculations of which candidate might win rather than whom they support. As a result, many voters will bullet vote "e.g., cast one vote for their first choice and no votes for anyone else "thereby reducing even further voters' ability to express their range of views about candidates.

Condorcet (Ranked Pairs)

Condorcet voting, or ranked pairs, is a system where voters rank candidates in order of choice and then each potential pairing of candidates is compared on each ballot. The candidate with the best overall record against other candidates is the winner. Condorcet voting has not been used in any public election in the world as of early 2007, although it is used to elect the leadership of some private associations.

Condorcet voting violates the common sense principle of requiring a minimum level of core support because it puts no particular weight on a first choice over a second choice. In a large field of candidates, a candidate can win without being a single voter's first choice. By putting such extreme emphasis on breadth of support, it shares one of approval voting's downsides of encouraging candidates to be seen as the "least offensive"candidate rather than a leader who takes strong policy positions that might alienate some voters.

Condorcet also may not produce a definitive winner. In an election among Candidates A, B and C, Candidate A may be preferred to B, B preferred to C, and C preferred to A. In this situation, a "fallback"method must break the tie. This potential need for a fallback could punish sincere voters. Condorcet also is a difficult system to count by hand when many people vote. Hand-counting is important if there is a problem with voting machines or concerns about software.

Other Voting Methods

Variations of these methods have been proposed, along with entirely different alternatives. One alternative is the Borda Count, where voters rank candidates and candidates win points based on that ranking; for example, a first choice might earn five points, a second choice three points and a third choice one point. Although used for certain elections like picking the winner of the Heisman Trophy, the Borda Count violates all three of our principles: majority rule, minimum level of core support and rewards for sincere voting.

Scholarly Support for IRV Over Alternative Methods

Range voting, approval voting and Condorcet advocates frequently criticize instant runoff voting on such grounds as being "non-monotonic"(theoretical situations exist in which improving the ranking of a particular candidate can hurt that candidate's chance of winning) and not always electing the Condorcet winner. Both critiques are based on mathematical theory but are almost never relevant in practice. Furthermore, we believe the Condorcet winner should not always win if lacking a minimum level of core support. To us, being able to lead and represent people well makes it important that at least some voters rank the ultimate winner as their first choice.

Leading scholars ground our views. Here are two examples:

  • In The Burr Dilemma in Approval Voting (Journal of Politics, February 2007, pgs. 35-36), University of Pennsylvania's Jack H. Nagel explains why non-monotonicity is not a serious flaw for IRV (here called the alternative vote):

    [Exploitation of non-monotonicity] is indeed possible under the alternative vote, but the conditions it requires are extraordinarily restrictive. Note that the kind of strategic voting required to exploit non-monotonicity under the alternative vote demands far more of voters (and organizers) than its counterpart under approval voting. The approval voter who truncates nevertheless votes quite sincerely for his first choice, whereas the alternative-vote manipulator must put her last choice first. Moreover, under approval voting, the act of truncating in itself does not hurt the voter's favorite (although a resulting retaliatory spiral might), whereas the number of insincere voters under the alternative vote must be precisely calculated and controlled, or else the manipulators risk eliminating their favorite. These drawbacks may make strategic voting under the alternative vote less " when and if it occurs; but they also make it, I suggest, far les probable than truncation under approval voting.

  • Virginia Polytechnic Institute Professor Nicolaus Tideman argues that “the alternative vote is quite resistant to strategy” and notes “the lack of monotonicity in the alternative vote should probably not be considered a serious defect since many competing rules [i.e. election systems] lack this property as well"(Collective Decisions and Voting: The Potential for Public Choice, Ashgate Publishing, 2006, pgs. 194-195). He further notes "the lack of resolvability is also probably not too serious, because it seems likely that irresolvable ties would be quite rare."Tideman addresses the Condorcet critique with “the fact that in 87 elections in the sample [one focused on large fields of candidate elected by a proportional voting system with ranked ballots] there were just three in which there was a dominant option [i.e. Condorcet winner] that was not chosen is somewhat reassuring."


  • Return to main page