Outside the Beltway

On American Parties

More on primaries with a foray into Madison and the general politics of power-seekers and incentives.

Steven L. Taylor · Sunday, May 8, 2022 · 13 comments

To follow on recent conversations (especially here) on the role primaries play in our politics, let me try and be direct and explicit about what I trying to point out about American political parties and the role played by primaries and nominating mechanism.

Let’s note from the get-go that collective action is always a challenge. Keep in mind that we talk about Democrats or Republicans as office-holders, we are speaking about thousands across the country and when we talk about Democrats or Republicans as voters we are speaking in terms of tens of millions. Given these numbers, it is no surprise that coordinated action is not as common as we tend to want and that coherence is often lacking. The way we use primaries in the United States exacerbates all of those problems because it utterly decentralizes control of party labels.

This hits, directly I would note, the debate that often rages in the comment sections of this site: the demand that Democrats fix their message. Of the reasons that I often push back on these message arguments is because it is utterly unclear who, to pick one party, “The Democrats” are in these conversations. Is it Biden? the White House? Pelosi? House Democrats? Senate Democrats? California Democrats? Texas Democrats? Is it AOC or Joe Machin? Is it the commentators on MSNBC? Liberal Twitter?

Perhaps we see the point?

Beyond all of that, however, let’s focus for a minute on a main role of parties: to be a signaling device to voters as to what they are getting from a given candidate, at least in theory. But the less coherent the process of selecting candidates, the less clear that signal is.

A quick look at someone voting in Travis County, Texas this coming November shows 41 offices being contested. In 2020 my Montgomery County, Alabama ballot had 16. (And that comports with my general recollection that the TX ballot tended to be longer than the AL one). The degree to which the typical voter or even the atypical one (i.,e., one with a Ph.D. in Political Science who care deeply about this stuff) knows, with clarity, the biography and background of all the candidate on the ballot is quite low, if not basically zero. So, they use the party label to help sort it all out.

Voters rely on party labels to signal to them important information about the candidates they are voting for. But, the candidates are not actually vetted by the parties in the US. Instead, they get to use the party labels because they choose to do so and then managed to win some small number of voters in a primary to let them move on to the general election ballot.

This is not a coherent way to construct a party if the goal is to create even a semi-cohesive goal to govern and it illustrates the way that the parties can change and evolve without anyone noticing until after it has happened.

Keep in mind, that voters tend to vote their party identity, and so when faced with only two viable choices, they are guided more by label than anything else. And, I would note, since there is some level of collective behavior once in office, a legislative body, in particular, there is some reasonable logic to state that if I am a D for some set of reasons that those reasons are more likely to be acted on in a way I find acceptable if I vote for a D instead of an R, even if I don’t know all the details about a given D. But in the US that collective coordination is a lot lower than in other democratic countries.

Because, of course, a given D might be a Joe Manchin or an AOC because the party does not have overall control of who puts a “D” after their name. Indeed, in a different nomination structure, it is rather unlikely that those two would be in the same party at all.

It should be fairly easy to see how the parties, instead of evolving and operating in a coherent fashion, are really general collectives made up free agents who answer not to the party, but to primary electorates. I would note that McDonald’s franchisees have to adhere to a long list of behaviors and policies to use the name “McDonald’s” but a Democrat only needs to win a local primary to use the name “Democrat” (and to get on the primary ballot they only need to self-select).

Let me get a bit more meta for a moment (or, perhaps, several moments).

Keep in mind two fundamental facts. First, humans are imperfect. Second, humans respond to incentives and constraints related to their goals.

One of my favorite passages from James Madison is in Federalist 51 and in many ways encapsulates my general approach and understanding of the political engineering of institutions. In the passage, Madison is in the middle of discussing the role of what would come to be known as separation of powers and checks and balances. He is describing how such a mechanism would constrain certain kinds of behavior. The details of those matters are not my primary concern here (indeed, for reasons I have discussed before, it didn’t quite work as planned) but this discussion hits on my two points from above and provides a general theory of constitutional design (emphasis mine):

It may be a reflection on human nature, that such devices should be necessary to control the abuses of government. But what is government itself, but the greatest of all reflections on human nature? If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary. In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and in the next place oblige it to control itself.

Human beings aren’t angels. They are not ideal, compassionate creatures who serve the divine law to perfection and who are cast out of their positions of power if they fail to fulfill the missions for which they were created. As Madison notes, “If men were angels, no government would be necessary.” Why would we need a government if we were all divine beings? No, humans are imperfect and we need government.

But, alas, if government is to be run by imperfect humans, what can we do? This is, I would note, an ancient question in the annals of human philosophy. It is the central question of Plato’s Republic and to all the tomes that followed until the present day.* The imperfection of humanity and its (in)ability to govern itself justly is encapsulated in the quote attributed to Churchill: “Democracy is the worst form of government, except for all the others that have been tried from time to time.”

So, if humans and democracy are flawed, the best we can hope for is to design the system in ways that take this into account. And primaries, while they sound good on paper, do not create great incentives and therefore do not move politicians into behaviors that we might want.**

Take Kevin McCarthy as a prime example. We know from recordings that were recently released to the public that the Minority Leader of the Republicans in the House saw Trump for who and what he was in the wake of the 1/6 insurrection. Yet, not long thereafter, he was at Mar-a-Lago kissing up to Trump and has subsequently defended the Big Lie about the elections.

Why?

Well, the simple answer is that he wants to be Speaker of the House of Representatives, and to do so requires three things: his re-election (which requires his re-nomination), a majority of House seats going to Republicans in November of 2022, and for that majority to want to keep him in leadership.

This requires McCarthy not being primaried (his re-election is not in question if he is re-nominated), the re-nomination and re-election of his allies in the House, and a Republican caucus that still sees him as leader. To achieve these goals, he needs to both toe the line on the Big Lie/how great Trump is for his own personal re-nomination, but also to help his allies get re-nominated so that he has the votes to become Speaker.

Where in any of this is the incentive to buck Trump? And, note that Trump is in the position of power that he is in owing to a combination of the way we nominate presidential candidates, the way we elect presidential candidates, the importance of the presidency, and the way partisan identity shapes voter behavior in the main. The first three items on that list are the result of structure and the fourth item directly interfaces with structure to produce very real outcomes.

Now, yes, we could all want McCarthy to let his more angelical nature rise to the surface and to sacrifice his political aspirations in service of the truth and in the vain hope that he could change the direction of the party, but he knows (and we know) that such a move would lead to his loss of power. He might be primaried and replaced by a trumpista and, without any doubt, even if he survived that process, the Republican caucus that returned in January of 2023 would not elect him Speaker (because most of them are also responding, sincerely or not, to the threat of being primaried).

I suppose if Kevin McCarthy were an angel, this post would not be necessary, but many of us who have even a passing knowledge of human history or, failing that, deal with human beings on a regular basis, know better.

What readers (and people in general, really) often want, whether they realized it or not, is simply better humans in politics. But, the reality is that, for the most part, people are people. Moreover, the kind of person who wants to be in politics is, by definition, more than a bit egotistical and, also by definition, someone who wants power. So, while it would be great if the people who go into politics would be the most angelical amongst us, that is rarely the case, and indeed it is often the opposite. So the goal should ever be to construct the machinery of politics in a way as to incentivize, as best we can, more optimal behavior. Or, at the very least, construe the structures to diminish what we know to be undesirable behavior.

I would note that finding and producing “better humans” often ends poorly. And that the reason that democracy is the worst form of government except for all the others is that all the others often try to find a way to get those “better” humans in power (e.g., monarchy, theocracy, or other ideological systems) and they end up often being, well, the worst humanity has to offer instead of the opposite. Indeed, any system that empowers someone and tells them that they are in charge because they are the best (which is what “aristocracy” means) or because God put them in power (theocracies or divine right monarchs) we see quite well that humans ain’t angels.

Getting beyond the question of specific politicians, consider that the incentive structure of American primaries as it pertains to new factions. A new faction (e.g., the Moral Majority, the Tea Party, MAGA, the Squad, etc) starts as ideas in the heads of certain political actors who want power. These power-seekers know that power requires election to office. That requires getting on the general election ballot and winning the most votes (with the winner’s party baked in, often, by the ways the lines are drawn). The easiest route to the general election ballot is via nomination by a major party (R or D), and that is via primaries. So, most power-seekers will take the path of least resistance. Instead of forming a new party, they will compete in the primary of one of the two parties closest to their preferences. If they can win they have an automatic pathway to the general election ballot. And the free agency aspect of it means that voters may think they are getting one kind of signal from a D or an R, but may be getting something very different (but then partisan identity kicks in and the cycle can then repeat itself).

We will never get the incentives needed to form new parties as long as this pathway is so easy (and I would note that Top Two process like in California makes it even easier to acquire the R or the D because the first round is just a free-for-all of self-nomination).

The basic alternative is a system (like is basically all other electoral democracies worldwide***) wherein parties exercise control over who uses their label, which allows those parties to present a more coherent message to the electorate. In practical terms, this means that power-seekers have to decide if they want to conform to that coherent message, or not. If they wish to conform, they have to stay conformed, within reason, or else they will lose the label and their pathway to staying in office. Or, if they do not wish to conform, they will have to form a new party (or give up on office altogether).

tl;dr: primaries are vital to understanding the behavior of parties, and of elected officials in general. And the US system of primaries is unique in the world. Our primary system often creates poor incentives, empowers very small parts of partisan bases, and contributes to somewhat incoherent parties and often to poor messaging.

I would note, too, that the best way for pressure for real reform to take place is for new parties to form and to start insisting on their fair share of power and influence.

I would conclude by noting that while this post is mostly about primaries, it really does include a lot of the fundamental underpinning of pretty much everything I write about concerning institutions/political structure.

*I expound a bit on this in a 2017 post, Democracy and Institutional Design I: A Basic Preface on Regime Type, which includes a discussion of the Fed 51 passage as well.

**By which I mean that, at least in the abstract, voters want coherent choices that lead to their interests being pursued. And, also in the abstract, voters want honestly, not subterfuge by politicians saying what they think people want to hear. Consider that the people who want McCarthy to support Trump want him to be sincerely doing so, not because he is willing to lie to get into power.

**Exactly how it works case to case, of course, varies. But it is not an exaggeration to say that no other country uses primaries the way we do. It is a definitive example of American exceptionalism. And it is fair to say, in a general sense, that even beyond the issue of primaries, most parties worldwide exercise far more control over their labels than do US parties, and this is especially true in well-established democracies.