Comment on “Why the Past 10 Years of American Life Have Been Uniquely Stupid”
This Atlantic article by Jonathan Haidt about increasing political polarization in the US is fantastic, and resonates a lot with me. One of the solutions proposed in the article is voting method reform:
Reforms should reduce the outsize influence of angry extremists and make legislators more responsive to the average voter in their district. One example of such a reform is to end closed party primaries, replacing them with a single, nonpartisan, open primary from which the top several candidates advance to a general election that also uses ranked-choice voting.
Unfortunately, while the concept of holding an open primary followed by a top-5 general is great, every proposal I’ve seen for this (like “Final-Five Voting”) uses single-mark ballots for the primary, and instant-runoff elimination for the general, both of which are clumsily-designed voting methods that can still elect unrepresentative candidates and perpetuate the problem. There are a lot of good voting systems to choose from, but these two continue to dominate the discussion in America, despite being almost as flawed as what they’re replacing.
The fundamental flaw with our current system, First Past The Post, is that voters can only express an opinion about one candidate, so if there are many similar candidates, votes get split between them, making it biased against candidates who are similar to each other and in favor of candidates who are unique (for better or worse). This is the origin of the spoiler effect, closed party primaries, and cycles of increasing polarization.
This is also a problem with open primaries that use single-mark ballots, though. In fact, it’s more of a problem, because they encourage many similar candidates to run against each other, while FPTP encourages candidates to drop out until only two are left. Open primaries take the spoiler effect and crank it up to 11, and can send two members of the same party to the runoff even if they only got a minority of the total vote:
Likewise, ranked-choice voting using the instant-runoff elimination method only counts first preferences in each round, so it also suffers from vote-splitting (despite all the marketing that claims otherwise). It is less vulnerable when candidates are very similar (“clones”), since votes for one will typically transfer completely to another, but if there are three or more strong candidates of different ideologies (which is the whole point of voting reform), the outcome is essentially random. For example, the most-representative, most-preferred candidates can be eliminated first by vote-splitting, which transfers votes outward to less-representative candidates, etc. until the second-worst candidate wins in the last round. This effect is due entirely to the elimination method, not to the ranked ballots themselves. In poor ranked-choice methods like this one, much of the information voters express on their ballots is never actually counted.
A much better solution for a two-round election is an approval voting primary, like the one adopted recently in St Louis. Voters can select as many candidates as they approve of, which makes vote-splitting much less likely. If the general is between two candidates, it can be held using single-mark ballots. If the general includes three or more, it can also use approval ballots.
Even better, a system with more expressive ballots can be used for the general, or can be used to eliminate the primary altogether, such as STAR voting or one of the many Condorcet Ranked-Choice systems. These systems consider all voters’ preferences for all candidates simultaneously, instead of counting only first preferences, making them much more likely to elect the candidate preferred by the voters overall, reducing the polarization and unrest that results from unrepresentative winners.
