Does Instant Runoff Voting Violate the One Person, One Vote Rule?

Brett Anderson
djbraski@yahoo.com
 
If a decision is made with a vote the majority should rule.  In order to have the majority rule, each person should get to vote and each person’s vote should be counted equally.  No one’s vote should be counted more than once, or weighted more heavily than anyone else’s.  Simple right?  One criticism of the voting method known as Instant Runoff Voting is that it violates this one person, one vote rule.
 
Instant Runoff Voting goes something like this.  Voters rank their preferences for an office.  Let’s say they can indicate their top three choices and rank them in order of preference one through three.  The top choices of all votes (whether pieces of paper or electronic files) are then tallied.  If a candidate receives a majority of first choice votes, then that candidate is declared the winner and the process ends.  If no candidate gets a majority of first choice votes, then the candidate with the lowest total first choice votes is eliminated.  This losing candidate’s votes are distributed according to the second choice on these ballets.  This process is continued until one candidate has a majority of votes.
 
One criticism of IRV is that it violates the one person one vote rule.  In the IRV system under consideration, each person ranks up to three candidates.
 
Instant Runoff Voting, simulates rounds of voting, where candidates with low vote totals are eliminated.  Multiple rounds of voting, eliminating low vote receiving candidates, and IRV are primarily designed to eliminate strategic voting.   In a voting system with multiple rounds of voting voters are clearly voting more than once.  However in the each round, including the last round, we can trace each vote to one and only one person.  So, each voter had equal efficacy in choosing the outcome, since their vote was counted only once for one candidate.
 
An objector may respond that each voter voted more than once in the whole process, and moreover those prior votes affect the outcome of the election.  So, not only does IRV, if we take each vote couting as the act of voting, than IRB violates the one person, one vote rule,  and this process affects the outcome, i.e. you often get different outcomes than if a non IRV process was used.  So, if one takes “voting” as a process where votes are counted, then each person has voted more than once in an IRV system.  Just as the rounds of voting which IRV is supposed to model.
 
However, if you think of voting as an act where choices are tallied against others, and where one person, one vote means that each person has equal power in determining a winner, then IRV does not violate the one person, one vote principal.  Since at the end of the process we can trace each vote at every round to one and only one citizen, each citizen’s vote, being the set of ranked choices, counted the same as every other citizen’s ranked choices, even though their were multiple rounds of counting.
 
Furthermore, I believe this what the one person, one vote is trying to get at, what we really want is equality of efficacy.  We want each citizen to have the same power as every other citizen in determining the outcome. Using the IRV each citizen has an equal share in the process of determining the choice at hand.  IRV clearly grants each person equal power to determine the outcome.
 
So, if the IRV does violate the one person, one vote rule, it does so only on a certain definition of this, and in a way that doesn’t violate any moral requirements of fairness and equality that lay behind the one person, one vote rule.  Perhaps what we need is an amended rule, something like, “Each person should have equal efficacy in determining the outcome of the election.”
 
But there’s one more complication.  Doesn’t the voter who marks three choices have more efficacy than the person who only votes once?  Certainly, voters who mark their full range of choices have their preferences reflected in the results at a finer grain than those who don’t.  This is what IRV is supposed to do.  But I would argue that each person, including the citizen who only marks one choice, still has equal efficacy in determining the outcome.  Their vote continues to count for the same candidate or choice in every round.  But if the voter who only chooses one candidate has no preference about the rest of the candidates, whether they know nothing about them, or know a lot and still do not care which of them wins, then this voter did not lose out.  Their preferences were accurately reflected by the voting procedure, for there are no further preferences besides the first choice to reflect.  For this citizen’s vote, the ‘one’ that is added to the totals, is counted only once for a candidate, just as everyone else’s.  Even though others had their votes moved from other candidates in previous rounds.
 
But what about people who through a misunderstanding or choice don’t complete the full available list.  They only write down their first choice and nothing else but have preferences that are not reflected on the ballot.  Surely, and especially at first, people will be confused by the voting procedure and will not vote, or make a mistake while voting.  First of all I think most people will quickly understand the voting system, but any voting system confuses some people.  Just because some voters make a mistake, and have their vote disqualified, or just plain vote for the wrong person, or make no mark and walk out befuddled, this is no reason by itself to think a voting system unfair or unreasonable.  There will always be people for whom any system is confusing.  While unfortunate we must face the fact that any voting system will, through mistakes, misunderstandings or complexity, not accurately reflect the preferences of some citizens.  We must be reasonable about the complexity and mental hoops citizens are required to jump though, while balancing the better, fairer, more democratic outcomes.  We must consider the benefits of the system alongside mainly temporary and small increase in confusion.  I believe the benefits greatly outweigh these concerns.  We should not throw away this opportunity for a better voting system because of the unfortunate confusion of some citizens.  For example, the fact that some people don’t understand our current voting system is no reason to trash it for something even simpler. Similarly our desire for inclusiveness should not trump the distinct benefits of IRV.  Also, the idea of democracy demands that we error on the side of trust if a question arises whether citizens can comprehend an idea or a process.  Certainly there is a limit to how complex we can make any public process; we must be reasonable and not exclude too many people.  But I think IRV is not a large step in complexity, and citizens will quickly get used to the idea.
 
So, IRV only violates a trivial, literal version of the one person one vote rule, if you consider a single vote being moved around in multiple steps.  It does not violate idea that each person should have equal efficacy in voting their preference which lies at the heart of the one person, one vote slogan.  Since at the final tally we can trace each vote back to one and only one person, IRV passes this test in that each voter had an equal efficacy or impact in choosing the winner.  Moreover due to the reduction strategic voting, which is IRV’s main benefit, IRV is more effective than the current voting system in the outcomes better reflecting the preferences of the voters.  I have not focused on the benefits of IRV, but only on exploring whether IRV violates the one person, one vote rule.