Debating “Two Cheers for Two Parties”

I recently watched Intelligence Squared’s debate on the two-party system, and I was shocked by the outcome. Before the debate, 35 percent of the audience agreed with the motion, which was “Two Cheers for Two Parties” (by the way, not exactly a paragon of a clearly stated proposition), and 27 percent disagreed. A plurality of 38 percent were undecided. At the close of the debate, 65 percent supported the motion, with only 28 percent opposed and 7 percent undecided. The “for” side improved by 30 points, compared to only one point for the “against” side.

The debaters were all people I know something of and deeply respect. Opposing the resolution were Lee Drutman, author of the brilliant new book Breaking the Two-Party Doom Loop, and Katherine Gehl, a business executive and reformer who serves on the boards of Unite America and New America. I knew I’d be rooting for this side, as the author of a book of my own that argues for essentially the same reforms as Drutman.

What surprised me was to find Norm Ornstein, a longtime advocate for political reform, and Yascha Mounk, an insightful German-born political scientist, arguing for the motion. I use quotes from both in my book. I have found both to be very knowledgeable and articulate on the subject of politics. But in this debate I don’t think they made a strong case.

Consider the following words from Ornstein (at about 24:20):

We agree with our opponents on the need for major structural reform. We agree on the importance of bringing us ranked-choice voting for Congress and for the presidency. We believe in either eliminating or vastly changing the Electoral College. We’d like to see multi-member districts in the House of Representatives and an enlargement in the House. We want to see major voting reform that will eliminate voter suppression and make it easier to vote, and changes in the money system in our politics. But we believe they have to be made within the structure of an existing system that’s built around having two major parties.

Except for the very end, these seem more like Drutman’s talking points than those of someone arguing for the motion. Drutman advocates multi-member districts and ranked-choice voting for the House (along with an enlargement to 700 representatives). And the thing is, if these changes were made, “you’d have more than two parties”, as Gehl interjects later (@~1:11:17). What’s puzzling is that Ornstein seems either not to know this, or to neglect it. Duverger’s law says that when you have multi-member districts and a proportional method of election, you soon get a multiparty system because the spoiler effect, a consequence of plurality voting, no longer deters voters from voting their true preference.

Mounk says he favors ranked-choice voting, but claims (@~1:11:23) that it would not move us beyond a two-party system:

It allows people to say, hey, my first vote goes to Jill Stein, but obviously she’s not going to get 30 or 40 percent of the vote, so it’s going to get eliminated and redistributed to the Democratic candidate. What ranked-choice voting allows you to do is to have some expression of your strong preferences, if you have more extreme political views, but we would still effectively have two parties. That’s why we are completely in favor of ranked-choice voting. But that is not abolishing the two-party system, it is fixing the two-party system.

This is mistaken in two ways. First of all, the fact that people could vote for a Green party candidate as their first choice means that they would have less need to worry about the spoiler effect. They could consider the platforms of all competing parties without having to think so much of tactical questions. This would give the Greens, and other minor parties, a chance to win, or at least a path to someday winning, and thus a chance to gain traction in the marketplace of ideas. It would make them a viable choice, and so even if they didn’t win, they’d make for a more competitive system.

Second, and perhaps more importantly, the race for the presidency involves a single winner, and so cannot be proportional. House elections provide a much stronger opportunity to move beyond two parties, by using multi-member districts and multi-winner ranked-choice voting, such as the single transferable vote (STV). In an election of six representatives for a given district, for example, if the Greens won a sixth of the vote, they would win one seat. And if we had a multiparty House but still only two parties capable of winning the presidency, that would count as a multiparty system.

Mounk also claims that New Zealand, cited by Drutman as a positive example, does not have proportional representation (@~1:01:20). He seems to use a very narrow definition of PR, where “you vote not for a particular candidate but for a political party” (@~34:00). But that’s what most people would call party-list PR — which is not the only kind. Many people agree that STV is a proportional electoral system, and under it, you vote for candidates by name (alongside party labels). It’s proportional because parties get seats in rough proportion to how many votes they get.

New Zealand adopted the mixed-member proportional system in a 1993 referendum, which combines local, single-winner, constituency elections with a “topping up”, at the national level, of additional seats given to parties as required to match the national vote proportions. So it’s not party-list PR, and it’s not STV, but the makeup of the legislative assembly reflects the vote, proportionally. Isn’t that representativeness the main point of PR? Isn’t that the kind of proportionality that matters?

The major concerns voiced by Mounk and Ornstein concern how well government would function with a multiparty legislative assembly in which coalitions would be needed to achieve a majority, and how well things would work with extremist parties in government.

Mounk suggests that with PR, we’d get “a white supremacist party with 10 to 15 percent of the vote” (@~38:23). Really? I’m not sure what definition of “white supremacist” he’s using here — the term has become loosely defined, in some circles — but I find it hard to believe that a party led by David Duke would get that many votes. In an electorate of around 150 million voters (depending on turnout, which would probably increase with a multiparty system), that could be more than 20 million people. There may be that many voters who want to keep immigrants out, but that’s hardly the same thing, I would argue, as white supremacism per se.

Drutman points out that extremism exists in every country, regardless of party system. The question is whether extremists should be formally represented in political discourse, or shunted to the margins and deprived of the option to vote for a party they agree with. If the latter, then their energies may be channeled into gaining control of one of the two major parties, which can be very dangerous, as Drutman explains (@~1:00:22):

We have a strange party system in which Donald Trump ran as a Republican. Probably got about 30 percent of the vote in the primary and about 40 percent of the people are Republicans. So, Donald Trump is about a 12-percent party, as many far-right populist parties in Western Europe are. Now, by winning the plurality of a plurality, Donald Trump becomes president, takes over an entire Republican Party. And now we have one party that, as Norm has described, has gone to a very extreme place. That is the danger of a two-party system.

James Madison warned about factions in Federalist no. 10 — and the worst kind of faction, he wrote, is one that commands a majority — exactly what our current party system promotes by enabling one or the other party to control each chamber of Congress.

By allowing extremist ideas to have formal representation, we can deal with them more openly. Daylight, as they say, disinfects. If an extremist leader is elected and gains a platform for their ideas, others can listen and respond in whatever way seems appropriate. If they are ridiculous ideas, they can be rebuked or made fun of, for example. Isn’t that better than letting such ideas develop less visibly, below the surface, infiltrating one of the major parties, where they have a chance of being enormously amplified if their activists manage to hijack the party?

This leads me to the subject of governing coalitions, which comes up a few times in the debate. For example, at about 1:15:00, Ornstein says, “If we have a number of parties and we’re getting coalitions and the David Duke party is there inside the tent able to use its leverage, I don’t think that’s going to be a healthy process.” But having a few seats doesn’t necessarily mean being “inside the tent”, if by that he means influencing government policy. Leverage depends on number of seats and ideological positioning. A party with a few seats, positioned near the ideological center, may have some leverage when one side or the other needs its votes. An extremist party, with just a few seats, can often be ignored because its votes aren’t needed to form a majority on a certain issue.

In contrast to parliamentary systems, where, when no party has a majority, a durable coalition of parties must be formed to select a prime minister (and cabinet), in the United States we elect the chief executive separately. This means there is no requirement of coalition-building among parties; majorities can be formed on an issue-by-issue basis. And it doesn’t mean that an extremist party would be likely to get cabinet posts, as Ornstein suggests (@~26:54). Sure, it’s possible that you’d have legislative gridlock, but we have that under two-partyism, too. It might actually be better, in the long run, for us to have a lesser quantity of legislating and a greater quality, in the form of laws that command strong public support because they will have been drafted in a more multilateral way that depends more accurately on majority opinion. Meanwhile, in the case of gridlock, we’d have a working executive branch capable of governing day-to-day.

There doesn’t seem to be much recognition in this debate of the crucial differences between parliamentary systems and presidential, or separation-of-powers, systems. The most prominent of the world’s presidential democracies are in Latin America, where a history of colonialism, huge wealth inequality, and other differences make direct comparison with the U.S. difficult. It’s true that things have not always gone so well there, in terms of democratic governance. Mounk says (@~52:13):

What you get is chaos, because a president can never agree with Congress or the parliament. You can never actually pass any laws. And what you end up doing is you get these strongmen who are coming in and saying, none of this is working. We always have political rancor. We can never actually pass any laws. What you need to do is to give all the power to me so I can sideline Congress and do what I want. And that’s why Latin American political systems have been much more prone to democratic breakdowns than political systems elsewhere.

But many of these countries have only been practicing continuous democracy for a few decades — it’s very much a work in progress. Despite this, as of 2019, three Latin American countries (Uruguay, Costa Rica and Chile) rank higher in the Economist Intelligence Unit’s Democracy Index than the U.S. All three are presidential and multiparty, and, so far as I know, doing well economically — Chile’s GDP per capita has gone from about $2,500 in 1990 to over $15,000 in 2017 (in current U.S. dollars). I don’t see how these facts square with Mounk’s “chaos”.


I think of the two-party system as a “meta-problem”. It’s a big part of our country’s inability to solve other problems. Instead of engaging sincerely and working towards solutions, politics on the “seesaw” is a zero-sum game in which the objective is to make the other party look bad. This does nothing to solve problems, but it works electorally — when there are only two choices. When there are more choices, people can choose something to vote for instead of just voting against. So in order to solve the big problems we face as a nation, it will help to solve this meta-problem, which blocks reform on many fronts.

Ornstein and Mounk don’t seem to appreciate this connection. Ornstein names several major problems — voter suppression, money in politics, climate-change denial, tribal media and the toxic culture and polarization it promotes — with no hint of how our party system affects them. Yet it does.

In countries that don’t have two-party duopoly (which is to say, virtually every other democratic country in the world), a competitive electoral marketplace means parties work to get more voters, and don’t benefit from suppressing votes. That strategy only makes sense on the seesaw. Change the incentives and you change the outcome.

In a competitive electoral marketplace, a party could find traction in the electorate by refusing big-money donations, and (if that principled stance proved popular) perhaps go from minor-party to major-party status. When there are only two parties, this strategy is unilateral disarmament for one party, and too risky to try out while the other party is raising huge amounts.

In a competitive, multiparty system, where no single party would likely have a majority on its own, the need for working together would be more clear; electoral competition would be less of an existential battle between “good” (our side) and “evil” (their side). This would gradually affect culture: Media outlets might come to accept the role of referee to a greater extent, and value objectivity more highly than they do in our current system, where they have incentives to increase “us vs. them” polarization.

Gehl emphasizes the marketplace analogy, pointing out that a system friendlier to political entrepreneurship would allow parties, as institutions, to innovate to capture untapped segments of the electorate. Competition is a key to accountability. It says a lot about our system’s lack of it that more people identify as independent than either Democrat or Republican. It’s a reflection of massive political homelessness.

The two-party system allows the dominant parties to “rig the rules of the game to protect themselves jointly from new competition”, as Gehl says. Her focus on competition is spot-on, in my view. It’s well-known that monopolies are unresponsive and unaccountable — duopolies aren’t much better. We need the innovation, fluidity and responsiveness that only a truly competitive electoral marketplace can give us. I think if the Framers had known that parties were essential to representative democracy, and especially if they had known about electoral methods like STV, they would have agreed.

It’s too bad the “for” side in this debate doesn’t see this — and that the audience apparently didn’t either. I hope this subject continues to receive attention; I think the more people hear about it (and read about it, hint), the more they will catch on.

One thought on “Debating “Two Cheers for Two Parties”

  1. I could not believe that the “for” side wone by such a wide margin. It seemed to me that they had very little reasoning for their side. They basically just kept saying that a multi-party system would not work with our culture and that we had bigger problems to worry about like the “cult” that is the Republican party. Honestly, because they were in San Francisco I cannot help but think that the only reason the audience voted so heavily for the “for” side it was because Norm kept on bashing Trump and the Republicans, otherwise I am at a loss for why more people thought the “for side debated better.

Leave a Reply to Caleb Cancel reply

Your email address will not be published. Required fields are marked *