Lee Drutman, author of Ending the Two-Party Doom Loop, recently joined Ezra Klein for an interesting podcast conversation. Drutman is a longtime advocate for proportional representation. While I am not completely against proportional representation, this post covers some of the reasons I think there are better electoral reforms.
Let me start with the part where I think Lee Drutman is exactly right: we need more political parties and more political competition. That absolutely needs to be the goal, because in my view everything else can only come afterwards. Only once the political system is competitive and responsive to the will of the voters will we be able to get money out of politics, deal with massive wealth inequality, reduce polarization, and restore the separation of powers across branches of government.
But is proportional representation the best way to achieve political competition? I’m not sure.
Concerns about proportional representation
The first problem I have is that proportional representation only applies to the U.S. House of Representatives. Of course we do need more competition in the House. But there are so many other positions that will always be determined by single-winner elections: president, senators, governors, and many local races like mayors. We need competition in all of those other places as well! We need reforms that can be used just as well in places where multi-member districts are not applicable.
Drutman does have proposals for those single-winner elections. He says that fusion voting should be used, which is a method that allows for multiple parties to endorse the same candidate. But to me it is a weakness and liability for election reform efforts when different methods are needed for different seats. I simply think it is not worth it, especially when forms of ranked voting can bring competition while maintaining single-winner elections.
The second problem I have with proportional representation is that it does mean larger geographic areas in which to campaign, and this increases the barrier to entry for candidates. One example Ezra and Lee discuss on the podcast is Louisiana, and Drutman says that the state’s six congressional seats could be part of one huge multi-member district. What this means though, is that candidates are going to need to effectively run campaigns across the entire state. While this might be fine for the two established parties, this is a real barrier to independent campaigns and nascent third-parties.
The third problem I have with proportional representation is the complicated ballot design. The main issue here is that for a single-winner election, you’d like to have at least three or four options to provide meaningful choice. Now suppose you’ve combined six districts into one huge district that will be used to elect six individuals. How many candidates should be on the ballot to provide a similar degree of choice? Suppose we’d want three options per district – that’s 18 candidates on the ballot for that single race.
Now, there are different ways of designing a ballot, and some are worse than others. Single Transferable Vote (STV) is the method of proportional representation that FairVote advocates, and this is a really bad one. Suppose a district with two seats available, and each of three political parties ran two candidates each. You already have a ballot with six names, and the example ballot on the Protect Democracy website asks for ranking of all six names. That is already quite a cognitive load for a single race. Now imagine the district had three or four seats. The complexity just seems too high.
Lee Drutman, to his credit, prefers a different form of proportional representation called open list, where parties provide a slate of candidates, and voters choose their preferred candidate. Seats are allocated proportional to the party vote, and seated by the most popular candidate within each party. I actually think this open list PR method is reasonable in terms of ballot complexity, and much prefer it over the STV method. But I still worry that for a multi-member district with four seats, you’d want at least 10 or 12 candidates, and that is a lot of information for a voter to process.
A simpler solution: Ranked Condorcet methods
To me, a Condorcet method with ranked ballots like Bottom-Two Runoff (BTR-IRV) or Consensus Choice are a simpler way to bring political competition and open the door for more parties.
Bottom-Two Runoff would still make gerrymandering a useless strategy. The reason that gerrymandering works is because with only two parties and no feasible entry by independents, it is fairly straightforward to predict the partisan lean of different neighborhoods and draw district lines to maximize seats for a party. Once ranked voting makes third-party and independent campaigns viable, because they no longer risk vote-splitting and spoiling an election, gerrymandering will become ineffective. Because it will be very hard to predict what will happen once entry threats become viable.
And it is this threat of entry by candidates outside the Big Two Parties that is the whole key to unlocking true democracy. It gives real choice and creates a more stable electoral system.
When I say a “more stable electoral system”, I mean something very precise. As I’ve discussed before, the current election system is fragile in the sense that small changes in vote shares can put the country on drastically different policy trajectories. Just think of the presidential elections in 2016, 2020, and 2024; in all of them, shifting a fairly small number of votes in some key areas would have put the country on a completely different path.
It is the possibility of independent entry that would have made these elections more robust. In 2016, there was widespread dissatisfaction with both major party nominees. Ideally, an independent like Michael Bloomberg or Jim Webb would have entered to give another alternative to Trump. But the current system meant that giving more choice to voters would have ensured Trump’s victory, and no such independent entered the race. We need election reforms that make such increases in choice helpful, and BTR-IRV would achieve that.
This would give voters better choices and reduce polarization. Ezra Klein laments what he views as the main problem with politics (min 58:30): “The left of center parties suck… I just look at the way the Democratic party is acting, and it is making just, in my view, terrible strategic decisions one after another.” The problem with the two party system is that Democrats making strategic mistakes has benefited only Trump’s Republican Party. A multi-party system would mean that another campaign or party would have an opportunity to take votes from Democrats when they make mistakes, without it being only Trumpism as an alternative.
That is how election reform that allows for entry would provide for more robustness in the political system. While proportional representation is a step in that direction, ranked Condorcet elections in single-member districts are a simpler way to achieve that result.