Category Archives: Voting Theory

Why we need ranked choice voting (RCV)

Maybe you think our current voting method is working just fine.  Just take the time to read this FOX News story, “Massachusetts mayor recalled – and re-elected – amid federal indictments” from Fall River, MA.  The short version is that faced with Federal indictments for tax fraud and stealing from investors, the city’s mayor, Mayor Correia refused to step down so the Council held an election with two questions on the ballot. 

  • Q1: Should the mayor be recalled? 
  • Q2 If recalled which of five candidates should replace him? 

Oddly enough Correia was one of the five names to replace him.  On question 1, he was recalled with roughly 2/3rds of the voters voting to recall.  On question 2, the 1/3 who voted not to recall him of course voted for him as the new mayor.  Because the other 2/3rds who wanted him recalled split their votes among the other 4 candidates the mayor was re-elected as his own replacement.  Even though it was clear that 2/3rds of the voters opposed him, our flawed plurality system put him back in the office against the will of the majority. 

This sort of thing happens quite often in regularly scheduled elections and voters just assume that a plurality winner would beat every one of his opponents with a majority in head-to-head matchups.  In a normal election there’s no way to prove whether this perception is right or wrong.  But in this case the ballot asked the right questions.  We can clearly see how Correia would lose to every opponent in a head-to-head race.  Ranked Choice ballots give the voting algorithm more information about voters’ wishes from which the collective will of the electorate may be determined.  For instance one common method simply eliminates the weakest candidates one at a time until only two remain.  Mayor Correia would never reach the majority needed to win that final round.

For more on ranked choice voting try Amazon’s free preview of my ebook “Finding Our True Political Center – Through the Coming Revolution in Voting.”  On your computer just click on “Look Inside.”

CONSENSUS VOTING

A Typical Ranked Choice Ballot

Its forthright simplicity, directness and familiarity make it much more likely to gain voter acceptance by voters than other more complex forms of ranked choice voting.  You can think of it as plurality voting on steroids.  The ballot is no different than the one we are all used to and if you choose you can simply put an “X” or the numeral “1” in the box next to your top pick.  But you may also rank your second and consecutive choices on the ballot, as few or as many as you like.  This is a “ranked choice ballot.”  Either way, if your top pick gets a majority of first place votes he or she is the winner. 

But a simple plurality doesn’t cut it. 

What’s wrong with selecting a plurality winner?  Well I’m virtually in the process of writing a book on this topic called “Finding Our True Political Center”, but I will give you one example here.  We’ve all seen polling on head-to-head matchups.  If I were to win head-to-head matchups against every one of my opponents, it makes sense that I should win the plurality vote as well.  So it may surprise you to learn that in a polarized race there is a good chance that I will lose to a polarizing opponent.  But wait for it; that same polarizing opponent may actually be the plurality winner even if he loses head-to-head matchups against every one of his opponents.   In voting theory this is known as a Condorcet loser. 

When no one gets a majority of first place votes with consensus voting we add the second place votes to the total to see if some majority of voters rank any candidates in either first or second place.  If so the candidate with the most votes wins (note as many as three candidates can meet this requirement).  If none do, we simply add additional tiers until a winner is named.               

As a moderate you have liked both Klobuchar and Buttigieg for the 2020 Democratic nominee. Splitting the moderate vote you hoped one would drop out soon so the other can pick up their supporters and gain more delegates in March.  This conundrum called the “spoiler effect” is a common problem with plurality voting.  Let’s see how consensus voting eliminates this spoiler effect.

For simplicity let’s make our hypothetical election a nation-wide consensus election for the Democratic nominee. Since we expect no single candidate to have a majority of first place votes we know a consensus winner cannot be named before the second place votes get added.  By ranking one of our favorites first and the other second you effectively cast a full vote for each.  The theory is you give them each the number of votes they would have received from you had the other not been in the race.  This argument actually bears out pretty well in a diverse field of voters but I will leave the proof to the mathematicians.

The term “ranked choice voting” refers to any of the many forms of voting that allow ranking of choices on a ranked choice ballot.  The most popular of these, “Instant run-off voting” is often mistakenly called ranked choice voting, seemingly to the exclusion of a host of other prime candidates.  I submit it is the most popular because it’s the first, not because it’s the best. 

Many ranked choice forms reduce the field one candidate at a time then repeat the algorithm on the remaining field in “elimination rounds.”  The “instant run-off” eliminates the candidate with the fewest first place votes.  The “Coombs method” or the “bad apple sort” eliminates the candidate with the most last place votes.  No matter how good these more complicated algorithms are, the fact that voters are confused by and skeptical of elimination rounds is problematic.  Consensus voting takes a straight-forward approach to selecting one winner rather than eliminating a series of losers.

My personal mission has been to find a more inclusive algorithm that doesn’t scare voters away.

Inclusivity is a measure of how well the voting method weighs the wishes of all factions (including such as racial, ethnic and religious groups in addition to party affiliation) in proportion to their numbers to find the right balance. 

The instant run-off is neither inclusive nor user friendly. The Coombs method is quite inclusive but perhaps even more difficult for voters to follow.  Consensus voting is inclusive, intuitive and familiar.  This leaves us a couple of questions, why do we want an inclusive method and how does consensus voting accomplish this? 

The “Why” takes us back to polarized elections and divisive campaigns. 

A polarizing candidate is one who attracts a lot of first place votes and a lot of last place vote with few in the middle.

Of course this is by design because the objective in a plurality race is to “be the first choice of a lot of voters” so it really doesn’t matter what the rest of the voters think.  His Amazon rating might be 3-stars, because that’s what he gets with a lot of ones and a lot of fives.  In other words we elect a lot of average candidates who are also divisive politicians.  Getting all 4-star ratings gets you squat in a plurality race.  A lone polarizing candidate is possible but more often they come in (polar opposite) pairs. So we need an algorithm that gives weight to those second and third choices.

The best way to understand the “How” is to see it in action especially in a polarized race.  Consider a three-way race between a liberal, a conservative and a moderate.  Both of the extremists have lots of firsts and lots of lasts and certainly one of these will win a plurality race.  The moderate dominates the center but will have the fewest first place votes making him the first-man-out in an instant run-off race.  One of the two more divisive candidates will win promising four more years of divisive and obstructionist government.

If no candidate receives a majority in a consensus race we add in the second place votes making the moderate a certain winner with nearly 100% of the (first or second place) votes.  This introduces a new problem.  What if the guy in middle is a doofus?  He wouldn’t be the first doofus we’ve elected.  But would we really elect a doofus in this situation?  The answer is yes. In a polarized race defeating the opponent is as important, sometimes even more important than electing the preferred candidate so yes, we will always place him last even if it means electing a doofus.

But this is a problem that really isn’t a problem.  Don’t get me wrong; because of this apparent problem it took me a while to find this very simple algorithm and no doubt this is the reason no one has endorsed or promoted it before.  What we need to understand is plurality voting and the instant run-off open the outer lanes for divisive candidates while consensus voting closes these lanes an opens the center lanes for moderate candidates.  With the center lanes wide open many qualified moderates would compete for these lanes and there are lots of qualified moderates.

We are not trying the change the behavior of voters because we can’t.  But we can change the behavior and the incentives for qualified candidates to become unifiers rather than dividers.

INCLUSIVE VOTING – FINDING OUR TRUE POLITICAL CENTER

After reading my most recent article, How Our Voting Method Brings Out Our Worst, my son, Ian said, “Dad, they’re never going to change the voting method.”

He may be right, but if there was ever “a right time to do it” that time is now.  If there was ever a time when the public is ready for change that time is now.  In fact with the support of national organizations like FairVote, local governments across the country are trying out different voting methods in hopes of getting fairer elections. But local governments generally do not face the extreme polarization we see in national elections. National elections are what we need to fix and it’s already happening.

A Congressional House race in Maine was just resolved using a form of ranked choice voting called the “instant run-off” and the plurality winner didn’t win…

Continue reading

John McCain: May His Legacy Live On

Usually we think of eulogies as remembrances, goodbyes or of a passing, but all who spoke for McCain this past week spoke with inspiration and hope for the future, even a call for urgency as in this Hemingway quote from Obama’s eulogy;

Today is only one day in all the days that will ever be. But what will happen in all the other days that ever come can depend on what you do today.

I could find no better words to express my admiration to this man.  But I would like to propose a plan to turn his principles and ideals into productive action for future generations to come.  Continue reading

Voting II – Can Ranked Choice Voting Save Democracies Around the World?

We tend to think of polarization and divisiveness as symptoms of the current administration in the U.S. and maybe they will just disappear with future elections; well that’s not going to happen.  This has been trending for over forty years and getting rid of Trump will not change this.  Putin’s efforts to undermine the institutions of democracies around the world exploits this vulnerability, but this only highlights how bad it can get. With or without Putin’s help polarization will bring an end to many democratic regimesContinue reading

Voting I: Ranked choice voting is the real thing

As published in Grinnell Herald Register on July 9, 2018

The Washington Post recently published an article entitled Give ranked choice voting a shot.  Believe me if we do we’ll never go back.  Ranked choice voting (RCV) is better by orders of magnitude.  The idea of ranking your choices in order of preference is really not that hard for voters to understand.

As a matter of clarification RCV is not the same thing as instant run-off voting (IRV) as implied in the article.  RCV is simply the format of the ballot allowing us to rank our choices whereas IRV is the algorithm most commonly used for sorting out the winner. Continue reading