The 2020 invisible primary in light of 2016

It’s still early, but the Democratic Party is showing some power.

We are well into the Democratic Party’s inaptly named “invisible primary” for the 2020 presidential nomination. Some candidates have declared officially, some are publicly mulling a run, and many are visiting early primary states and reaching out to potential donors and endorsers. Some have even quit. Rather than focus on what’s happening in the race,, I want to focus here on how we study and describe it, especially in the wake of the tumultuous 2016 nomination cycle.

Last week, Vox’s Matt Yglesias suggested that the fact that Elizabeth Warren wasn’t the runaway frontrunner for the Democratic nomination was somehow evidence that The Party Decides’ theories no longer applied. Dave Weigel did a bit of dunking on those who had downplayed Donald Trump’s chances in 2015. Other political observers seem to be of the mind that because of 2016, we no longer have a theory about how party nominations work.

This strikes me as either triumphalist or nihilistic. Or maybe both. But let’s step back for a second and think about just what we do and don’t know about presidential nominations, and what happened in 2016.

First of all, yes, if you followed the logic of The Party Decides, that would have led you to substantially underestimate Donald Trump’s chances to become the 2016 Republican presidential nominee. And it might have led you to make some wildly inaccurate predictions about how that nomination contest would turn out, as I did.

But what do we do with that information today? If a useful theory leads to a spectacular belly-flop of a prediction, does that mean we throw out the theory? Do we update it? Do we conclude that it’s still a pretty useful theory that had a bad data point?

Before answering, it’s worth noting that the theory nicely explained what happened on the Democratic side. Party elites conferred with each other and made an early determination that Hillary Clinton should be their nominee, and they did everything they could do to signal that they’d made this choice long before anyone in Iowa or New Hampshire had started voting. It kept many viable Democratic candidates from entering the race, and the only real challenge Clinton faced for the nomination was notably from someone barely even a member of the party. For the Democrats in 2016, the party decided, and it got the nominee it wanted.

That doesn’t mean the decision was uncontroversial, of course. And the fact that Clinton lost the general election and faced lingering criticism from a sizable portion of Democratic activists means that it may be harder for the party to make such a decision in 2020.

The simple fact is that we’re not entirely sure what rules and norms are governing the 2020 invisible primary. That’s not the same as saying that no rules or norms govern it. Do Howard Schultz or Michael Bloomberg or Sally Yates have the same chances of winning the nomination as Elizabeth Warren or Cory Booker or Kamala Harris? I doubt it, in large part because party actors still have a good deal of agency in determining who is and isn’t a viable candidate.

They still seem to be preferring traditionally credentialed candidates, like senators and governors, and many party insiders are partial to nominating a woman or a person of color. If the party had no voice, being wealthy and famous might be enough to make one a serious contender, as it was among Republicans four years ago. It’s still early, but that doesn’t seem to be happening on the Democratic side.

Perhaps more importantly, as Jonathan Bernstein and Josh Putnam have noted, the presidential candidates are acting like party insiders still matter. Warren and other top-tier candidates are assiduously courting African-American leaders. They have spoken to various key constituency groups within the Democratic Party in an attempt to prove that their ideas are consistent with where the party is. They have been visiting Iowa and New Hampshire and other early-contest states and attempting to win the friendship and support of longstanding party activists there. Some candidates who had been flirting with a run have already been winnowed out due to a lack of insider support.

In 2015 and 2016, Trump very visibly and repeatedly refused to bend the knee to key interests within the Republican coalition, whether he was insulting John McCain or flouting longstanding party views on foreign policy and trade. The party nonetheless rallied behind him eventually anyway. This is not the dynamic we are witnessing within the Democratic Party so far.

But how should we evaluate what we are seeing? As I see it, there are a number of different potential outcomes for the Democratic nomination contest that could occur, listed here in declining order of party control:

1. Party insiders settle on a candidate before Iowa and primary voters nominate that candidate.
2. Party insiders cannot agree on a candidate but manage to winnow the field to a few traditionally credentialed candidates who are ideologically in sync with the Democratic Party.
3. Party insiders settle on a candidate before Iowa but primary voters nominate someone else.
4. Party insiders cannot agree on a candidate or winnow the field, but Democratic voters pick a traditionally credentialed nominee anyway.
5. Party insiders cannot winnow the field, and Democratic voters nominate a non-traditional candidate who holds views hostile to the Democratic coalition.

The very large number of candidates involved in this race and the lack of an obvious coordination point (such as a candidate named Clinton) make it a different sort of nomination cycle than Democrats have faced in recent decades and make options 2, 4, or 5 more likely than usual. Voter distrust of party elites and lingering resentments from 2016 make option 3 somewhat more likely than usual. But we may well be on the path to option 1, and the activities we’ve seen so far are consistent with that.

So, by all means, let’s evaluate existing theories in light of what happened in 2016, and let those of us who were spectacularly wrong in describing those events be particularly humble and careful in examining 2020. But let’s not just assume that because one party exerted no control in one nomination cycle that parties are functionally dead as organizations. We have too much evidence telling us otherwise.