Who took control of my brain? I got up early but I’m not sleepy. Most shockingly I remember what I’ve been planning on writing about, elections. I don’t mean the particular election coming up. I’m not talking about specific candidates. I want to discuss how we decide who wins. If there are only two candidates in an election, it’s easy. The candidate with the most votes win. The problem is what to do when there are more than two. It can be mathematically proven that there is no “fair” method of choosing a winner. Here are two definitions of fair.
The Pareto Condition (PC): If every voter prefers Candidate A to Candidate B, then Candidate B should not be declared a winner.
The Condorcet Winner Criterion (CC): If a candidate would win every single one-on-one race, that candidate should be declared a winner.
Unfortunately no possible system can satisfy both criteria. Let’s look at an example with three candidates. A, B, and C and the percent of people for whom each is the first choice.
A – 40%
B – 25%
C – 35%
By our current, first to the post, system candidate A wins. Seems fair. But what if B and C are more closely aligned and the supporters of one would all support the other. 60% of the voters will be unhappy as their least favorite candidate win. To avoid this situation, some states have instituted rank choice voting. People don’t pick one candidate; they rank the candidates. If nobody gets a majority of first place votes the candidate with the least votes is eliminated and we look at the second choice amongst their voters. In this extreme case every vote for B prefers C to A so now C wins with 60% of the votes. Sounds fair, but it isn’t. Let’s say that every supporter of A prefers B to C. If there were just a two-candidate election between B and C Be would get 65% of the vote, a landslide. But under rank choice he’s been eliminated.
In elections with small groups of people there’s a better way. You can just keep having elections until someone has a majority. That’s not quite the best method. It’s better to keep voting until there is a consensus, perhaps 60% to minimize how many people are unhappy. That’s how we run juries. They keep voting until the decision is unanimous.
There are other methods that try and overcome the difficulties, but none are perfect, and we have to live with the possibility of many unhappy people if there are more than two candidates in an election. This is the strongest argument for a two-party system. Even that kicks the problem down the road. How do we choose the two candidates? In current America that’s where the problems arise, in choosing the candidates. We use a system of primaries that is done virtually nowhere else. In other democracies some segment of the party leadership picks the candidates. In England it’s whoever pays a fee to join the party. Only a small fraction of the electorate joins any party. In Norway the head of the party just chooses the candidates.
In the US our crazy quilt system is largely run be the individual states. Some use caucuses that are cheaper to hold but far less representative. Most people won’t commit, and open-ended time needed to attend a caucus. Once there they use something akin to rank choice. After all the primaries and caucuses, we have a convention. Now it becomes like a jury, a small group can keep voting until someone has a majority. There have been elections where none of the declared candidates won, someone outside the field was chosen as the compromise, it’s called a Dark Horse Candidate. James Polk was the first Dark Horse to be elected president.
What method do you think is best? Here’s the most important thing; to be intellectually honest it can’t depend on the particulars of the elections. When I was young New York State had four significant parties, Democrat, Republican, Liberal, and Conservative. The Liberal party usually backed the Democrat and the Conservative the Republican; usually but not always. The idea of the minor parties was to keep the major parties far enough left and right to satisfy the members of the minor parties. That often led to three party elections. Once that gave New York City the liberal John Lindsay. Once it gave the state the conservative William Buckley using the first to the post system. Sometimes I won and sometimes I lost but I accepted the results. You can’t change it after the fact.
Now think about the upcoming Democratic primary. Think about the last Democratic primary. You can’t say I’d go with first to the post in 2016 but Rank Choice or multiple ballots until consensus but another choice in 2020. I have already seen people contradicting themselves. Don’t be one of those people. Be honest.
I have been first to the post until recently; the experiences of Maine and other places that have tried ranked choice have shown that it works better than I thought it would. People are not gaming their votes. I’m not convinced it’s best, but my objections are smaller. As for choosing candidates, having the convention as a deliberative body when no one has a majority seems wise. I’d like to see the threshold raised to 60% but I’m not sure if that is practical. That’s for wiser heads than mine to decide.
Let’s go back to my original example. If there were a convention with 100 people, my guess is that B would end up the winner. That feels the fairest to me. Everybody’s voice is heard at all steps in the process.
I suggest that you read more on this from better sources than me. Not enough thought is given to how we practice democracy.
Wise Madness 