You are here


The following unpublished oped was submitted to a Boston, MA

newspaper prior to the referendum in that state on introducing

Ranked-Choice Voting into the state's elections, which fortunately

failed.  It does a good job of demonstrating the serious flaws of

the algorithm used to determine an election's outcome given

voter's preferences, or rankings under Ranked-Choice Voting..



Ranked-Choice Voting, A Flawed Choice


Dear Oped Editor:


To move away from our dysfunctional two-party system, we need to change our voting system.  Ranked-Choice voting (RCV), aka Instant Runoff Voting (IRV), has been touted as the answer. However, IRV is a fundamentally unfair voting system that does not do anything its proponents claim and violates more fairness criteria than existing plurality voting. 


IRV has entrenched the two-party political system wherever it has been tried.  If a voter puts a third-party candidate as his or her first choice, it can hurt the chances of the voter’s second choice major party candidate, who could potentially be eliminated in the first round, causing that voter’s last choice to be selected for office. 


The piece in the Winter/Spring 2017 issue of Green Horizon Magazine by the Executive  Director of FairVote following the adoption of IRV in Maine elections is an example of the  misleading advocacy by IRV’s supporters.   In his article the author states "Starting  in 2018, Mainers will be able to vote for the candidates they like the most without helping  elect the candidate they like the least," which is patently false as illustrated by the numerical example below. The article fails to inform the reader about a serious flaw of IRV-- monotonicity failure. In other words, ranking a voter’s first-choice candidate LAST could cause that candidate to WIN as opposed to ranking the first-choice candidate FIRST, which could result in that candidate LOSING!


Instant Runoff Voting is non-monotonic because increasing a vote for a candidate may decrease a candidate’s chance to win. Voters should have the right to know how to rank their first-choice candidate - first or last or in between - in order to help their first-choice candidate win.  Unfortunately, this is not the case with instant runoff voting. Here is an example:  


 #Voters  Votes 1st  2nd  3rd

       6                 B >  A  > C                 

       5                 C >  B  > A                 

       4                 A  > C  > B                


Candidate C wins this contest because candidate A is eliminated in round one, giving 4 more votes to candidate C, resulting in 6 votes for B and 9 votes for C in round 2.


 Yet, if two additional new voters whose real preference is B > A > C vote, so that candidate A is now preferred over candidate C by 12 to 5 voters:


#Voters  Votes 1st  2nd  3rd             

      8                 B >  A >  C                 

      5                 C >  B >  A                   

      4                 A >  C >  B                 


Then candidate A is eliminated first and their least favorite candidate C wins with 8 votes for B, and 9 votes for C.


However, if these same two voters voted A > C > B (ranked their second favorite candidate  A first, their least favorite candidate second, and their favorite candidate last) then their favorite candidate B wins: 


 #Voters  Votes 1st  2nd  3rd

       6                 B >  A >  C                  

       5                 C >  B >  A                   

       6                 A >  C >  B                  


This time C, their least favorite candidate loses the first round, resulting in 11 votes for B and 6 votes for candidate A, and their favorite candidate B wins.


In other words, if these two new voters want their first-choice candidate B to win, they must  rank candidate B as their last choice instead!   Work done by Professor Robert Norman and Joseph Ornstein at Dartmouth College suggests that this monotonicity failure where increasing your vote for a candidate X, may cause X to lose can occur in a fifth of close three-man IRV elections.  However, it is important to point out, as Warren Smith has done, that the more candidates in an IRV election, the greater the possibility for such failures.


Approval voting is an alternative voting method that is simpler to count and to manually audit machine-counts for accuracy and does not suffer these problems.  Under approval voting in multi-candidate elections voters may give one vote each to the candidate or candidates they support and the candidate having the most total votes wins.  Approval voting reform would eliminate the wasted vote, the spoiler role, the necessity of vote splitting and give fairer outcomes in our elections, especially for third parties.


P.S. Kathy Dopp, M.S., received a master’s degree in mathematics from the University of Utah (1989) and is the author of a chapter on post-election auditing mathematics in an upcoming book on numerical methods being published by Springer.  For more information on the flaws of the IRV method of counting rank-choice ballots please see: Realities Mar Instant Runoff Voting