Diverse Representation: Conclusions

How can people looking to reform voting methods be the most effective at improving the representation of women or racial/ethnic minorities? This is the last post in a four-part series on the effects of voting methods on female and minority representation:

1. Districts vs. Block Plurality
2. Comparing Single-Winner Voting Methods
3. Proportional Representation
4. Conclusions

What should you do if you mainly care about female and minority representation?

The big goal for people looking to improve the representation of women and minorities needs to be proportional representation (PR). PR clearly outperforms the alternatives. It’s much better for women’s representation than single-winner districts. As for racial/ethnic minorities, single-winner districts sometimes can give them good representation, but only if the minorities are relatively large, the districts are drawn well, and the minorities are geographically concentrated such that most minority voters can be in a majority-minority district. By contrast, PR always works. It’s not dependent on geography, and, even when a minority is too small to elect someone on its own, PR makes it easier to elect someone from that minority with the help of a relatively small number of White voters.

Switching from single-winner Plurality to other single-winner voting methods may also help, but the empirical evidence is far weaker and the theoretical arguments are less compelling. Given a yes-or-no choice to stick with Plurality or switch to something else, the something else is the better choice — but we shouldn’t expect a tremendous improvement. Better single-winner voting methods are worth pursuing, especially for singular positions like mayor and governor where you can’t have multi-winner elections, but getting PR will be a much, much bigger difference.

Switching from Block Plurality to single-winner districts is never optimal, but it can still be an improvement. People looking to improve the representation of racial minorities should be careful, however; if the racial minority isn’t reasonably large and geographically concentrated, a switch to single-winner districts won’t help them and it may even backfire.

Since PR is a far bigger improvement than other reforms, people wanting to improve female and minority representation should communicate in a way that emphasizes what is important. In particular, they should stop referring to Instant Runoff Voting as Ranked Choice Voting. For instance, RepresentWomen uses the RCV term for both single-winner IRV and multi-winner Single Transferable Vote (Proportional RCV). They know it’s an important difference, but their terminology obscures it. A common hope is to use IRV as a stepping-stone to STV, but how can they convey the importance of switching from IRV to STV to casual observers when they use the same term — Ranked Choice Voting — for each? I think it’s justified to use the RCV term for STV (perhaps as “Proportional RCV”) since the RCV label is far better-known and more popular. But casually conflating the two in the present makes it harder to campaign for PR in an IRV-labelled-as-RCV future, and we should avoid it.

Not overusing the RCV label is even more important when it comes to Preferential Block Voting, a winner-take-all and non-proportional multi-winner voting method used in Utah. From the perspective of someone casting a ballot, Preferential Block Voting and STV are indistinguishable; they both use ranked ballots to elect multiple candidates. But they yield very different results; Preferential Block Voting isn’t proportional, so it is far worse for representing minorities. Imagine a city in Utah is considering a switch from Preferential Block Voting to STV, and consider the perspective of a relatively well-informed voter who has no special interest in voting methods but is vaguely familiar with some of the claims. To her, both of these voting methods are Ranked Choice Voting, which she’s heard results in fairer outcomes and is better for minorities, and she likes the ranked ballot. But why should she switch from RCV to RCV when she’s heard the version of RCV they’re currently using also gives the same benefits of the new-fangled version that will be more expensive to administer? The difference is subtle, and when everything is just “RCV” it will be invisible to almost everyone. If Preferential Block Voting is better than Block Plurality it isn’t much better, and adopting it will make it more difficult to adopt a legitimately good voting method. Preferential Block Voting should be opposed, especially when it masquerades as RCV.

For electing women and minorities, which proportional voting methods should be supported? My answer is approximately, “All of them”. Semi-proportional voting methods like Single Nontransferable Vote are significantly worse than the genuinely proportional methods, so you probably shouldn’t settle for them if you’re looking to help women. But arguments for why one genuinely proportional voting method is better than another tend to be very speculative, and we should expect the differences between proportional voting methods to be much smaller than the difference between proportional and non-proportional voting methods.

Should activists focusing on voting methods who care exclusively about improving the representation of women or racial/ethnic minorities behave differently than those who want to improve elections in general, without a special focus on underrepresented groups? While much of my advice is the same, I think those focused on gender and racial equity should have somewhat different priorities. Specifically, they should:

• Focus more on getting PR relative to improving single-winner voting methods.
• Care more about a voting method’s political viability relative to how good it is overall; for example, I expect Condorcet methods to be a little bit better than IRV in terms of diverse representation, but I expect the difference to be much smaller than it is regarding other considerations (e.g. Voter Satisfaction Efficiency).
• Place greater value on giving voting methods real-world tests; we can determine things like how well the winners under a given voting method reflect the preferences of the electorate solely on the basis of theoretical analysis and computer simulations, but we can’t make halfway confident predictions as to how they’ll help women or racial minorities.