What Really Explains Polarized Elections?

It has been said that the 2016 election gave us the worst pair of presidential candidates the two major parties could have found.  Whether or not that’s true, all the mudslinging certainly made them seem this way.  Beginning with the primaries every candidate’s strategy is to fire up his base and to hell with the rest of us.

We tend to blame social media and fake news stories for the polarization and lack of civility, but I think we are overlooking the obvious culprit, our voting method.

There are literally thousands of voting methods we can choose from.  All are fair and all are consistent with democracy, but each voting method has a unique dynamic that can influence not only the outcome but the very nature of the campaign and the functionality of the government that follows.  The most prominent dynamic of simple “plurality voting” is how it inspires strategic voting, that is we often pick someone other than our favorite.  Voters don’t want to waste their vote on a sure loser nor waste their time voting for a sure winner.  This assures there will be no serious third party challenges and we can expect many more close elections.

Compared to our two-party presidential elections the upcoming French presidential election to be held this April 23rd features ten major parties.  We might think this can be explained by cultural or social differences or maybe something different in the constitutions, but it’s not.  It can be explained by one minor difference in the voting methods.  We both start with plurality voting (one man-one vote).  In the U.S. whoever gets the most electoral votes is the winner.  But in the French election process if no one gets a majority, the two top vote getters (popular votes) are pitted against each other in a run-off election.  Polling suggests that the populist candidate will win a plurality the first election, but a moderate will prevail in the run-off as voters in the other eight parties will likely then back the moderate.  With ten major parties there is greater diversity in political thinking compared to the U.S. where voters get pushed to polar extremes.

As a long-time Republican, I see the party as a hodgepodge of disconnected ideas, sort of a coalition of minorities held together by strength in numbers.

If we used the French voting system we might have a better picture what registered voters actually wanted.  We might see instead a Tea Party, a Populist party, an anti-abortion party, an economic conservative party and (I hate to say it) an anti-environmental party which may or may not all come together in a run-off.  We already see many states enacting anti-abortion measures touting a voter mandate.  And yet according to a 2016 Pew Research Poll, public support for legal abortion is as high as it has been in two decades of polling. Currently, 57% say abortion should be legal in all or most cases, while 39% say it should be illegal in all or most cases.

Mathematicians will point out that the French do not have to go to the trouble of holding two separate elections to get the same result.  Simply allow voters to list all candidates in order of preference on the first ballot and sort it out with an algorithm.  A variation of this has been adopted for the Maine Gubernatorial race, where the candidate with the fewest first place votes is eliminated and successively repeated until only one candidate remains.  This differs from French elections in that it’s possible to win without placing in the top-two plurality spots.  Everyone can acquire new supporters as weaker candidates are eliminated.  If your top choice is eliminated your second choice drops into that spot.  This may seem confusing, so I will devote a future post to voting theory or you can refer to this Stanford paper.  Earlier I said,

Each voting method has a unique dynamic that can influence not only the outcome but the very nature of the campaign and the functionality of the government that follows.

Perhaps the best way to demonstrate this is to conduct a totally hypothetical thought experiment.  You can play along by trying to project the winner.  We’re going to compare two voting methods that sound similar yet get entirely different results.  The first is Maine’s method dubbed “ranked choice voting” and the second is Coombs voting named after the Frenchman who first proposed it.  I prefer to think of Coombs voting as the “Bad Apple Sort” where we go through the apple basket throwing out the worst apple until only the prize apple remains.  Both methods use the same “ranked ballot” where voters list candidates in order of preference.  In the bad apple sort we successively eliminate the candidate with the most last place votes (eliminating the most disliked as opposed to eliminating the least liked).  Advocates of the bad apple sort claim it finds the most satisfactory candidate for the group as a whole.  Let’s continue with our thought experiment to see why.

The most striking thing about the “bad apple sort” is how it mirrors the brain’s own decision making process, successively eliminating the worst choices in a natural extension to group thinking.

Suppose we make a do-over simulation of our recent election, the “nightmare scenario” where the two major candidates are both liked and hated in near equal proportions.  With ranked choice voting the moderate candidate B (in white) is eliminated first with (only 5%) the fewest the first place votes.  The right wing candidate A (in red) would have won in a plurality contest with 48% of the vote.  Even though he picks up another percentage point from the B supporters, he loses to the left wing candidate C who picks up 4% to capture a 51% majority.  Using the “bad apple sort”, A is eliminated first with 51% if the last place votes.  Then C is eliminated with 53% of the last place votes.  Now this is the exact scenario some would use to show the bad apple sort is unfair when “We could actually select B the winner while getting only 5% of the popular vote.”  What these critics don’t understand is how the voting method changes the dynamics.

This bad apple scenario will never happen because if you and I could figure out the outcome so would the candidates.  They would adjust their campaign strategy and goals from “be liked by the most” to “be hated by the fewest.”  Instead of firing up their base and building animosity, they would try to project themselves as promoting an agenda serving for all voters.  “Bad apple sort” voting method has a strong bias against the kind of polarization that has paralyzed Congress is recent years.  Ranked ballot voting would also attract more highly qualified moderate candidates who can draw votes from the extreme candidates.

Am I pushing for a constitutional amendment changing our voting method?  I would love to but realistically this is not going to happen.  What I have learned is there is a way to sort out the bad apples using our conventional two party system, if we are willing.  We should never again have to choose between the two worst apples in the basket.

Look for my next post ” The Surprising Role Independent Voters Can Play in Changing the Course of American Politics”