This post’s purpose is to put down thoughts related to a conversation I had last night. One point concerns the defense of parties. Another concerns the nature and explanation of the two-party system.

Political scientists ‘like’ parties but for potentially different reasons.
Comparativists define them as institutions (more precisely organizations) linking citizens to government. They also see parties as the raw material of government formation. Some further see parties as agents of democratization. Many are skeptical of internal party democracy. There is a normative bent in favor of social-democratic parties, although recent research highlights the importance of conservative parties for the emergence and maintenance of liberal democracy. Comparativists have formed these views by studying multiparty systems.
Americanists see parties squarely as institutions and teach them as such (often under the banner of American National Institutions). I think I understand why. The American constitution came before the country’s political parties.1 It is the other way around, to varying degrees, in other countries. The American party exists to cope with the separation of powers. It also exists to hammer out the deals we otherwise would associate with a formal process of government formation. Hence the American party can be thought of as a parallel institution — parallel to those laid down in the so-called Madisonian constitution. And I think political scientists who study the United States are, for these reasons, more comfortable with internal party democracy than are comparativists. The long-standing bent in favor of the Democratic Party is well-known, although I think many would say we also need the Republican Party.
Which of the defenses is more likely to satisfy a person on the street? My bet would be the first. The Americanist vision implies a condition in which it’s okay for voters to have just two choices. And it might well be. Yet nobody likes the idea of not having choices.
The second topic we touched on was how to explain the two-party system in a classroom. It’s not Duverger’s Law. It might be the unit-rule allocation of electoral votes, and that is where I used go. But I think this misunderstands the American party. It is not an organization but instead a parallel institution in the sense of social choice. One can compute the effective number of organizations. By that metric, the U.S. is an outlier. But I don’t think this metric really captures what is happening in the U.S. I guess I am suggesting that the ‘party’ concept does not travel.
- I do not agree with the truism that the Framers failed to plan for their emergence. The best discussion of this is in Taylor et al. (2014), where the authors discuss rejection of the Virginia Plan and the intellectual gymnastics this imposed on Madison. I sometimes twin this reading with one by Pope and Treier (2015).