Outside the Beltway

On the Number of Parties 4: The Oddity of Two, Part 1

A real example of American exceptionalism.

Steven L. Taylor · Sunday, January 22, 2023 · 5 comments

Recent conversations in some of my posts have led me to think about how to better talk about why I find the structure of the American party system to be a problem for the quality of American democracy (the fact that the question intersects with a conference paper I am working on helps spur me on as well). As usual, there is an awful lot to say and I struggle, to a degree, as to how much I should focus on a very narrow aspect of the discussion, and how much I should try and make broader connections to various institutional variables. My goal here is to try and place the rigidity of American bipartism* into a comparative context.

It also strikes me that this provides a way to extend a series of posts that I have meant to return to, “On the Number of Parties.” The previous entries in this series:

• On the Number of Parties 1: Basic Counting, Part 1
• On the Number of Parties 2: Basic Counting, Part 2
• On the Number of Parties 3: More Sophisticated Counting

I would note that part 2 has quite a bit about some of the references I often make in posts and in the comments on an ongoing basis, such as the difference between the German party system, as shaped by its mixed-member proportional system and Israel’s system as produced by closed-list PR in one national district. It is supposed to be an illustration, as I often try to underscore, that not all multi-party systems are the same.

Part 3 defines some terms that I will use below (specifically the effective number of parties). It also promised that part 4 would discuss Duverger’s Law and the relationship between party systems and electoral rules. That will have to wait for a later installment. I expect it likely to be more prominent in part 5 (or maybe 6?).

Let me try and present as briefly as possible a basic theoretical underpinning of why I think this is important. I will stick to mostly just declarative sentences.

1. Parties are an almost natural consequence of representative democracy.
2. Party labels function as signaling devices to the population to provide at least a basic notion of what politicians who adhere to a given label will do in office. Note that this is often all the information a given voter might have about a given candidate.
3. Parties serve a key role in organizing government.
4. Ideally, elections should provide a chance for voters to reward or punish parties based on whether they performed as their labels suggest they should.
5. The number of parties in a given system is a function of how cleaved a given society is (e.g., ideological, linguistic, ethnic, religious, etc.) plus how the prevailing electoral rules translate votes into office.
6. It stands to reason that a country with substantial ethnic, religious, and other cleavages (e.g., India) would have more parties than a less diverse country (e.g., Japan).
7. The ideal number of parties is an open normative question, with the theoretical bounds being one party for every single voter to one party only. In more practical terms we can discuss the question of how much fragmentation is too much versus not enough. Too much fragmentation would make governing difficult and not enough fragmentation means that the various interests of a given society are being inadequately taken into account when governing decisions are being made (or that some factions have undue influence on outcomes relative to their actual size in the population).
8. For representative democracy to function optimally, representation needs to do an adequate job of approximating the political views of the society as a whole, within reason.
9. Ideally, if a set of interests in society feel left out, they should be able to form new parties to compete for a piece of power to further their interests. (The cost of enrtry should be reasonable).
10. Accountability and competition should create some level of churn of who is in charge, and who seeks office.

As to point #7, I would argue that it is an empirical matter that two parties are insufficient to represent the broader interests of the diversity of opinion that exists in the US. Allow me to quote Henry Droop’s critique of American representation from 1869(!):

As every representative is elected to represent one of these two parties, the nation, as represented in the assembly, appears to consist only of these two parties, each bent on carrying out its own programme. But, in fact, a large proportion of the electors who vote for the candidates of the one party or the other really care much more about the country being honestly and wisely governed than about the particular points at issue between the two parties; and if this moderate non-partisan section of the electors had their separate representatives in the assembly, they would be able to mediate between the opposing parties and prevent the one party from pushing their advantage too far, and the other from prolonging a factious opposition. With majority voting they can only intervene at general elections, and even then cannot punish one party for excessive partisanship, without giving a lease of uncontrolled power to their rivals.

As well as his general view of representation, with which I generally concur:

It will, I believe, hardly be disputed, that the claim of a representative assembly to have the decisions of a majority of its members accepted as the decision of the whole country, depends upon the theory that these decisions do in general correspond to what the majority of the whole body of electors in the country would decide, if they had leisure sufficiently to investigate each of the questions to be decided, and an opportunity to vote upon it.

Indeed, I commend to you all my friend and sometimes co-author Matthew Shugart’s run-down of Droop’s views (from whence I took the above quotes).

With at least some theoretical notes about the role of parties and the issue of representation, let me shift now the number question specifically in the US.

Let me be as straightforward as possible: the remarkable stability of a two-party system in the United States is an anomaly in modest-sized-to-large (in terms of population) democracies. The only reason that most Americans don’t realize this, and if they do kind of know it don’t fully grasp the significance, is the same reason that fish don’t realize that water is wet.

Without getting into electoral rules, I went through Michael Gallagher’s dataset of the effective number of parties (electoral and parliamentary) of 141 countries. See part 3 for a discussion of those indices. Here is a list of outcomes from the most recent election in Gallagher’s list. I picked every case wherein both the ENEP and the ENPP were under 3.0. I included only countries deemed “Free” by Freedom House.

Again, part 3 of this series talks about what ENEP and ENPP mean specifically, but I will summarize here by saying that they are indices of fragmentation and that ENEP is about how many parties are competing for votes and ENPP is about how many parties win seats.

Note that the above is 23/141 cases (16.3%). I did leave out a handful of authoritarian cases (e.g., Nicaragua and Venezuela) and a couple of “partly free” cases.** Even if I included all of those cases we would find that less than a quarter of all cases on Gallagher’s list would have had ENEPs and ENPPs under 3.0.

What is striking about this list is that almost all of these cases are micro-states, often with assembly sizes more akin to city councils than to countries. It seems reasonable to suggest that countries with very small populations are probably not great comparative cases if we want to understand the United States.

If I filter the population to a million (still pretty small) or more we get:

But even doing that, we see some very small countries with small assemblies. Only Ghana, South Africa, and the United States have populations in at least the tens of millions and with assemblies in the triple digits.

South Africa is a bit of an anomaly in and of itself, insofar as far more than two parties win seats in the parliament, but the ENEP and ENPP figures are distorted by the huge percentage of votes and seats won by the African National Congress party. That really leaves Ghana as perhaps the only democracy of reasonably comparable size in the world that has a party system mathematically similar to that of the United States.*** That’s 2/141 cases (1.42%).

Note that my list includes no country of anything other than minuscule size from Asia, Europe, Oceania, or the Americas.

What’s more, were I to take a longer-term view and include longevity and stability of results, the US sticks out even more. Here’s the US House since 1946:

Note that even Canada and the United Kingdom, which both have similar (although not identical) electoral systems produce outcomes with important differences:

The fact that both systems produce ENEPs of over three means that there are a lot more vote-earning parties in those systems than in the US (2.10 in 2020). Likewise, both countries’ ENPP indicates that more than two parties win seats in parliament (as opposed to the US number of 2.00 in 2020).

Indeed, if we look at established democracies of any size (in terms of population), we find that no other case has the two-party stability that we see in the United States, especially over the length of time it has been the case in the US.

In the next part, I will look more specifically at some of these numbers over time, and also look specifically at electoral systems. We will see that while a number of former British colonies (Canada, India, and the US) use similar methods to elect their first chambers, the party systems are different, especially over time. I will also look into how New Zealand, another former British colony that changed from single-seat, first-past-the-post (the commonality of all these British colonial cases) to MMP in the 1990s and the resultant changes that we can observe in their party system.

But let me reiterate: the US is almost entirely alone when it comes to a stable two-party system both in terms of electoral competition and electoral outcomes. The main cases of similarity are micro-states with very small (sometimes tiny) assemblies that undercut any real comparative value. Once those are set aside, the possible cases of similarity are South Africa, which is actually far more different than the ENEP/ENPP number suggest initially, and Ghana, a case about which I know very little. But even if I knew more, it is still one case out of well over 100, which underscores the oddity of the US case when we realize that all other democracies of any size have multiparty systems of one size or another.

Perhaps this is because the US simply cleaves cleanly into two segments, although this is rather unlikely (and there is plenty of evidence to suggest that it is not the case). Or, perhaps this is because the general structure of our institutions forces multiple cleavages into two containers.

(More to come).

Resources:

Freedom House.

Michael Gallagher’s Electoral Systems web site.

Worldometer’s world populations.

*”Bipartism” as in two parties (not bipartisan, as in two parties working together). This is to be contrasted with multipartism.

**If I expanded the criteria to include “partly free” (and there might be an argument to do so, at least in some cases), the list would have expanded to include:

***At a minimum, I need to give the case a bit more of a look to see what might be the reason for these outcomes.