For just over forty years, the Republican Party had a famously orderly succession to its leadership. From 1968 until 2012, every non-incumbent presidential nominee was, in one sense or another, the next in line, as either

  1. the most recent Republican vice-president,
  2. the runner-up in a previous contested Republican primary, or
  3. the namesake son of the previous Republican president.

This lent even raucous Republican presidential nomination races an aura of seriousness and stability. The nomination sometimes went to someone who was once a rebel within the party (such as Ronald Reagan or John McCain), but only after he had put in some time in the organization, been given a thorough second look by the electorate, and developed into something more like a respected elder statesman.

Well, things continue until they don’t.

The 2016 election exposed grave vulnerability and fragility in the American party system. One major party was successfully hijacked by an extremist outsider in the face of initial opposition from a huge portion of the party’s elites and elected leaders. The other party came surprisingly close (if still not objectively very close) to meeting the same fate—and if Bernie Sanders had the advantage over Donald Trump of long experience in the Senate, his relationship to the Democratic Party was even more attenuated than Trump’s relationship to the Republican Party. (Sanders formally became a Democrat for the first time only in order to run in the primaries, and in the Senate he still identifies as an independent who caucuses with the Democrats.)

It is little appreciated how much liberal democracy depends on strong parties, and a revitalized, re-understood liberalism adequate to the moment will have to overcome a traditional distaste for partisan politics.

Seventy-five years ago the political scientist E.E. Schattscheider wrote that political parties are “the orphans of political philosophy.” For more than a decade, a handful of political theorists led by Harvard’s Nancy Rosenblum (full disclosure: my undergraduate advisor) have been working to integrate an appreciation for parties and partisanship into our understanding of what healthy democratic government is like. The election of 2016 showed how urgent that task is; as political scientist Julia Azari has been arguing for months, “the defining characteristic of our moment is that parties are weak while partisanship is strong.”  

I suggested in my previous essay that “voting patterns didn’t change enough between 2012 and 2016 to justify big claims about new national moods or about Trump’s distinctive appeal. I believe the consequences of this election will be deeply abnormal. But the voting behavior that brought it about was, in the end, very normal.” That normality was the strong partisanship mentioned by Azari. 89% of Democrats voted for Clinton, 90% of Republicans for Trump. Those figures are down a touch from 2012both major parties lost more voters to third parties than in 2012but considering the year of headlines about how unpopular both candidates were, the result is stark.

For a while in the Fall it seemed that Trump would alienate enough Republicans to suffer a sizeable loss to Clinton. But those alienated Republicans had mostly shifted their voting intentions to third parties, not to Clinton, and most of them came home again by election day. Which is to say, we should start by understanding that partisans are very, very likely to vote for their own parties’ candidates, regardless of those candidates’ personal merits or indeed the substance of their views. Republicans will even vote for an opponent of free trade and of the postwar western alliance who grossly offends against conservative Christian sexual mores, if that’s who is at the top of their ticket.

The most important political science book of 2016, Christopher Achen’s and Larry Bartel’s Democracy for Realists, shows that we should expect this. Systematic political beliefs, ideas about policies, and views about the state of the world don’t drive decisions about what party to support among most voters most of the time. It’s the other way around: the initial judgment call about which side I’m on (a judgment call that is often shaped by intergenerational habits or membership in identity groups: which party is the party for people like me?) then leads me to bring my beliefs, ideas, and views into line with those that prevail on my side.

This is in part a matter of economizing on information, which we all do, because we have to. No mind could contain the information needed for sufficiently-informed decisionmaking about the countless policies decided by a modern government, and no life contains enough hours to study it, even if the mind could contain it. But it’s not just a matter of low-information voters using parties as a substitute for judgment.

Achen and Bartels show that the adjustment of beliefs (even about factual questions) to match the party’s often increases with greater information and education; it takes some sophistication to correctly understand which side of an issue one’s party is on, after all.

All of which is to say: once Donald Trump got the nomination, there was a real chance of his being elected. (Note that this means no one should ever wish for, and they certainly shouldn’t engage in, strategic cross-party voting to help a major party to nominate a truly disastrous candidate. The other party’s candidate starts with a very large part of the electorate that’s almost sure to vote for them; and that could well put them close enough to a plurality that they’re a stock market crash, a terrorist attack, or an FBI letter away from winning. Although there is now some debate about this finding, Achen and Bartels argue that there’s strong evidence that weather or even shark attacks can jiggle the needle.) In the face of strong partisanship among voters, a lot depends on the strength and quality of the parties as institutions in shaping nomination outcomes.

The question, then, is how did Trump win the nomination?

The immediate answer involves a lot of nervous, feckless, or opportunistic Republican elites waiting for other Republican elites to do the risky work of seriously opposing him, or just taking for granted that he’d implode of his own accord like previous vanity and novelty candidates like Herman Cain had done. In the fall of 2015 Trump flirted with the possibility of running as a third-party candidate if he didn’t like how the Republican establishment treated him; and any of his rivals who took direct aim at him became the target of his mud-wrestling style of politics. (Ted Cruz stood out for the shamelessness of his strategy: hug Trump close and actively praise him, the better to be able to sweep up his support once, presumably, other Republicans had done the work of taking him down.)

Democratic elites, in their capacity as superdelegates, at least faced some pressure to get off the fence and declare for a candidate, allowing Clinton an early show of strength. The Republican Party doesn’t have superdelegates to defend the party’s organizational interests, and Republican officeholders were extremely slow to endorse or rally around any candidate. The RNC seemed so unnerved by the third party threat that its leadership sat on their hands.

Enumerating these bad decisions is just the detailed way of saying: it turned out that the party establishment didn’t have enough organizational strength to be able to repel the invader. An RNC more confident in the party’s strength would have made different choices. Republican officeholders and donors with a clearer sense of partisan coherence might have rallied around one alternative and pressured others out of the race soon enough to matter.

In other words, while we can easily imagine smarter, earlier, and braver action by party elites that would have changed the outcome, suggesting that the party might not have been organizationally impotent, the party elites managed things so haplessly in part because they were organizationally weak and didn’t have tools they were confident in.

In an important new paper, NYU law professor Samuel Isaacharoff argues that the ability of American parties to effectively coordinate political action has been undermined by reforms and legal restrictions including the decline in patronage, the increasing reliance on primary elections (I would add, especially “open” primaries), and campaign finance reform that disproportionately targets highly-visible fund-raising by the parties themselves. The result is “hollowed out” institutions that are vulnerable to “hostile takeover.” Presidential campaigns end up standing substantially on their own in organizational terms, and remake the party afterwards if they win. Trump had notoriously little organizational contact or useful coordination with the RNC in the summer and fall; it turned out not to matter as much as nearly all pundits expected.

Why should we care?

The expectations of the founders notwithstanding, parties aren’t a pathology. There is no case of anything resembling democratic politics in a large modern nation-state that doesn’t involve party competition. Parties aggregate ideas, interests, and ideologies into medium-to-large coalitions, offering information to voters that is even approximately manageable. (Imagine if responsible voting involved not only a mastery of millions of pieces of information about thousands of policies, but also finding out about where the dozens of candidates who appear on a typical ballot—from president to dogcatcher—stood on each of them.) Parties encourage tradeoffs and turn-taking among the groups that make up those coalitions, providing rewards for moderation and patience. And they monitor both their own politicians and the other side’s politicians for gross malfeasance or misgovernment.

A party has a longer time horizon than does an individual politician, and doesn’t want to be tarred with an individual politician’s sins. The individual politician may have an interest in short-term corruption, demagogic popularity, and personal power, but their own party has an interest in keeping them in check. And if their own party forgets about that interest, the opposition is there to remind them.

The elites in the Republican Party spent late 2015 and 2016 showing every sign of knowing that Trump was bad for their party. They simply couldn’t, or wouldn’t, do enough about it. As voters and as people we should hold them responsible for their failure, but as citizens trying to understand the current state of affairs and looking for routes toward improvement, we should worry about the institutional weakness their failure revealed.

Democratic government needs parties; parties do not need democracy. As a method of integrating large numbers of people into organized mass politics, the party has also been a tool of authoritarianism and totalitarianism—indeed, party organization is one of the key ways in which modern totalitarianism differs from pre-modern tyranny.

In other words, the fact that we need parties as such is the beginning, not the end, of responsible analysis here. We need healthy and competing parties with reasonably responsible elites operating under rules and institutions that encourage ongoing responsibility. The breakdown of parties like that is an easily-overlooked piece of the fragility of free societies—for example, when they become vehicles for the organization of mass support for one strongman, instead of coalitions of cooperating groups brought together by groups of intermediating elites. Parties are only semi-formally a part of the state, and at first glance don’t seem like a part of its constitutional machinery, but their weakness or breakdown becomes weakness and breakdown in the system as a whole.

Isaacharoff’s paper analyzes parties through the model of the corporate business firm. This is useful and illuminating. As Douglass North, John Wallis, and Barry Weingast have argued, there’s something important in the early-19th-century conjunction of the rise of mass-membership partisan democratic politics and the transformation of the corporation from a monopolistic tool of mercantilist power into an organizational form open for anyone to access in a competitive market. That is, modern competitive democracy and modern competitive commercial capitalism are both part of what they call the “open access order” in which access to organizational tools that solve collective action problems is widely available.

In classical liberal or libertarian social thought, the commercial corporate firm is an object of constant analysis; but when it comes to firms’ political counterpart, that body of thought looks more like political philosophy than like political science, and parties are nowhere to be found. (A thinker like F.A. Hayek might mention that he identifies as an Old Whig to signal what his commitments are, but the sense of the Whigs as a political party doesn’t play any part in that.) I would argue that this is of a piece with classical liberal social theory’s neglect of democratic political processes and indeed of political science as a whole. Classical liberals rightly don’t believe that politics is the most morally valuable thing in the world, but the liberal order that encompasses such values as freedom of religion and speech, due process of law, and open commerce is structured and protected by its political environment. That order  is acutely vulnerable when those political structures fail. Our colleagues in economics (and in law-and-economics) do constant valuable work analyzing which legal rules and institutional forms best allow the commercial corporate firm to fulfill desirable economic functions—and which are most likely to become perverse institutional failures. Liberal political thought desperately needs to pay as much attention to the firm’s political counterpart.

I haven’t offered solutions here. I hope to develop ideas about solutions in this space over the next year. But treatment comes after diagnosis. We have to start by understanding that the dangers faced by the political and constitutional order have a lot to do with the weakness of the parties that we rely on to to make that order function.


Jacob T. Levy is Tomlinson Professor of Political Theory and Director of the Yan P. Lin Centre for the Study of Freedom and Global Orders in the Ancient and Modern Worlds at McGill University; author of Rationalism, Pluralism, and Freedom; a blogger at Bleeding Heart Libertarians; and a Niskanen Center Adjunct Fellow and Advisory Board Member.