Matthew Yglesias’s Vox article on proportional representation is almost great. But I believe that by tying it to one particular form of PR (or, in a more search-engine-friendly term, pro-rep), he missed out on the real potential of the idea.
My argument in a nutshell
- As he said, PR is a solution to gerrymandering in the short term and a solution to polarized zero-sum politics in the longer run.
- This could happen as soon as 2021 if the Democrats embrace it on a federal level. State-by-state action would be much slower and involve fighting on almost-by-definition less-friendly ground.
- In order for congressional Democrats to embrace pro-rep in the numbers it would take to pass, it has to be relatively nonthreatening to their own reelection prospects. That means STV is not viable, but something like DMP or PLACE (explained below) would be.
“I am not a crank”
I’m a PhD student at Harvard; a board member of the Center for Election Science; the inventor of the E Pluribus Hugo voting method used by thousands of voters for nominating for the Hugo awards; and the primary organizer of the BC Symposium on Proportional Representation, which was highly influential in helping design the structure of the referendum that will start next Monday in BC.
Non-disruptive “district-based” proportional representation is possible
Voting theory and mechanism design have come a long way in the last 20 years. One idea that’s gained currency in that time is “biproportional” (or in layperson’s terms, “district-based”) methods. These methods balance two simultaneous constraints: party proportionality and geographic representation. For instance, PLACE voting has exactly one winner per district, just as today, while still ensuring proportional representation at a statewide level.
There have been various serious proposals of such district-based methods. DMP (Dual Member Proportional) is one of the options on the BC referendum. I’ll explain my own proposal, PLACE (Proportional, Locally-Accountable Candidate Endorsement) voting.
Details of PLACE voting, an example of a district-based pro-rep method
Your ballot lists the candidates running in your district. You can either choose one of them, or write in a candidate you’re more excited about who’s running in some other district.
Winners are then chosen by an STV-like process, where candidates are eliminated and the votes they hold are transferred. Whenever a vote is transferred, it goes based on the endorsements of the candidate who originally received it. Before the election, that candidate will have endorsed some “faction allies” within their party; if any of those remain, the one with the most direct votes gets the transfer. If no “faction allies” remain, then the same-party member with the most direct votes gets the transfer. And if no same-party members remain, the predesignated “coalition ally” with the most direct votes gets the transfer.
To begin with, all candidates with less than 25% of the votes in their local district are eliminated. This relatively-high threshold prevents parties from breaking down into a confusing plethora of short-sighted single-issue subparties, as in Israel. But note that the threshold is by candidate, not by party, so a charismatic third-party or independent candidate still has a good chance. Also, since your vote for an eliminated candidate still transfers, you are safe casting such a vote, and third parties can grow to viability over time.
Then, as in STV, eliminations proceed bottom-up, so that votes move from weaker candidates to stronger ones. The exceptions are when it’s necessary to ensure one winner per district: the last candidate from any district is never eliminated, and if one candidate wins a district the remaining candidates there are eliminated immediately.
If a candidate ever holds as many votes as the average district total, they win immediately (as with a “Hare quota” in STV). Any excess votes they hold are transferred onward. And as I said, other candidates in that district are then eliminated.
When no more eliminations are possible, there will be one winner per district remaining. These candidates are seated. To ensure that all voters have a sympathetic representative, parties assign the districts they didn’t win to one of their winning candidates as “extra territory”. So a Republican in a Democratic district, or vice versa, still has a representative from their party whom they can petition.
As with STV, this could be passed by Congress, with no need for a constitutional amendment. In fact, Congress could pass a law saying that only states with a gerrymandered “efficiency gap” of over, say, 3%, would have to use PLACE. That would mean that most Democrats voting for this new law would be utterly unaffected. A few Democrats in gerrymandered red states would see their own elections switch to PLACE, but keep the same districts; because PLACE fixes gerrymandering, it would only help their re-election chances. And a couple of Democrats from the one state with a Democratic gerrymander — Maryland — would have a bit harder road to re-election.
This plan would have advantages besides non-disruption. Because PLACE doesn’t require redistricting, and can use existing ballot-counting machinery, states would have little grounds to challenge and/or delay this exercise of Congress’s power over “times, places, and manner” of federal elections.
But STV has a movement…?
Of course I’m aware that groups like FairVote at the national level, or Voter Choice Massachusetts at the state level, are pushing for “Ranked Choice Voting” (which includes STV as one form). In fact, I’m on cordial terms with people like Rob Richie and Adam Friedman, the heads of the two orgs I mentioned. Meanwhile, the organization I’m a part of, the Center for Election Science, is a fraction of the size of these. But we are growing, with a budget of over half a million dollars this year and probably more next year.
I think, however, that the logic I sketched at the beginning is undeniable. Jurisdictions like British Columbia; Santa Clara, CA; and Lowell, MA; have all rejected STV in recent years, largely because of opposition from incumbent politicians. But the logic of proportional representation is strong, and now British Columbia is having another pro-rep referendum, with more modern methods replacing STV as the options.