05 March 2020

Ranked Choice Voting

Currently, most elections for a candidate race in which only one candidate may win, declare the winner of a plurality of the vote to be the winner (also called "First Past the Post" or "single member district plurality" systems). When that plurality is also a majority of the votes cast, this is unproblematic. But, this can be problematic when there are three or more candidates running and no one candidate receives a majority of the vote.

For example, in the status quo, if three candidates are running in a race where two candidates are left leaning, and one is right leaning, the left leaning candidates will tend to split the vote of the voters inclined to vote for a left leaning candidate, potentially causing the candidate least favored by a majority of the voters the win. 

We have a two dominant political party system with political primaries in the United States within political parties, precisely to reduce the likelihood of this happening. But, third-party and independent candidates in general elections create the same problem. In our system, an independent or third-party candidate hurts the major party candidate that they are most similar to, so voters who vote for someone that they most prefer undermine their own objectives when they vote for an independent or third-party candidate.

First past the post electoral systems have the natural effect of disfavoring having more than two political parties in any one geographic region (the UK and Canada managed to have more than one political party nationally because their political parties historically were mostly geographically distinct such as nationalist parties in Quebec, Northern Ireland and Scotland), and actively incentive voters and political activists to shun third-parties.

In Ranked Choice Voting (also called more evocatively "Instant Runoff Voting"), when applied in single member districts, each voter selects not just a first choice, but also second and sometimes third choices in a single member candidate race. If no candidate wins a majority of first choice votes, the candidate with the least first choice votes is eliminated and that candidate's votes are allocated to the next choice of the voters who voted for that candidate. This process is iterated until one candidate has a majority of the vote and is proclaimed the winner.

As a result, in ranked choice voting, in contrast, in theory, the scenario that played out above in a plurality voting system wouldn't happen. If a majority of voters are left leaning, and the left wing vote is split between two candidates, and a minority of voters are right leaning and represented by only a single candidate, in ranked choice voting, voters will, in theory, vote for their first choice, one of the left wing candidates will be eliminated, and the other left wing candidate will win with second choice votes, even though the right win candidate had more first choice votes.

I've used examples of three candidate races, but the analysis generalized to races with more than three candidates.

Ranking choice voting makes it much less important to vote strategically in situations where you have more than one preferred candidate, or there are third-party or independent candidates. This, in turn, makes third-party and independent candidates something that our system can embrace, instead of fighting because it undermines the ideologies that those candidates support.

Despite this fact, while ranked choice voting is a good idea in theory, and is superior to the default simple plurality voting system, it is not my preferred system.

This is because the average voter is bad at making informed choices hypothetically and it is much more burdensome for someone to sift though enough information to make not just an informed first choice but also an informed second and third choice. 

People inform themselves about their second and third choices less when they engaged in ranked choice voting than they do when their first choice has actually been eliminated and they have to consider the candidates who remain. 

Therefore, I prefer actual runoff voting in the event that no one candidate receives a majority, between the top two finishers, held a few weeks later, as Denver, France and Louisiana all do.

An actual runoff system leads to a better informed choice than a ranked choice voting determination would. Also, the complications involved in counting the vote ranked choice voting delays election night results, reduces transparency, and increases the odds of somebody screwing it up.

Therefore, while ranked choice voting is superior to plurality voting, the simpler system of holding an actual runoff between the top two finishers if no one candidate receives a majority, would be a better solution. This produces the same results as ranked choice voting in the hypothetical world where all voters are perfectly informed apart from some very esoteric scenarios where it still comes closer to the ideal result than the default plurality system. It still prevents third-party and independent candidates in general elections from having perverse effects. But, it is easier to administer, more transparent, harder to screw up, and is easier on voters with scarce time to inform themselves, particularly if they have a hard time thinking hypothetically, which most voters do.

Both Ranked Choice Voting and two round elections, make it possible to have more than two viable political parties (without having to have regional "nationalist" type parties to achieve that end). And, a system with multiple political parties is also more likely to eventually implement proportional representation.

Both Ranked Choice Voting and two round elections, also slightly favor more moderate candidates relative to a strict two party system single member district plurality system with political primaries, which is our status quo, because the "elimination round" isn't segregated by political party in the same way that primaries enforce. In a Ranked Choice voting or two round election system for a general election, a non-partisan centrist has a much greater chance of success than in the two party system with political primaries that tends to favor clear partisans over centrists.

Another approach is to have multi-member districts, or no districts at all. 

In a multi-member district plurality system where each voter gets to vote for as many candidates are there are seats up for election, this system suppresses candidates who represent localized majorities that are a minority in the district or jurisdiction as a whole. 

For example, if a district with five elected officials is 40% black Democratic and 60% white Republican and voters are residentially segregated, a multi-member district plurality system where each voter gets to vote for as many candidates as there are seats up for election is likely to elect a full slate of Republicans. But, in a single member district system, it is likely that there will be three Republicans and two Democrats elected, giving Democrats a voice and a chance to sway a swing Republican on some issues, even if the Democrats will still usually lose.

But, unlike single member district systems, multi-member districts are less vulnerable to gerrymandering, which is a very difficult problem to solve with election laws that has great potential for abuse that has been demonstrated frequently in U.S. political history.

Also, in multi-member districts, various forms of proportional representation are possible, that avoid both tyrannies of the majority, and gerrymandering risks. Most countries use some form of proportional representation to elect multimember legislative bodies.

No comments: