As the Iowa caucuses approach, there’s been lots of talk on the internet about whether Bernie Sanders can win. He’s gained support in the pollshas a lot of campaign money, and earlier in the season earned some coveted endorsements on the left. He’s also decidedly still a candidate who makes the political and media establishment uncomfortable. But if we learned anything from 2016, it’s that the possibilities for who can win a nomination contest and a general election are many. So it’s worth thinking about what kind of president Sanders would be.

This requires us to confront the presidency as it is, in the wake of Trump and all of the developments leading up to that. This isn’t some bipartisan whiskey fantasy bullshit, or an Aaron Sorkin vision of patriotic speechifying. The question of what kind of president Sanders would be goes beyond his temperament and ideology, two topics that I conservatively estimate account for about half the internet at this point. I would also urge readers not to interpret this as a polemic against or in favor of Sanders (Mischiefs of Faction will not be endorsing any number of candidates.) Rather, I want to offer some ideas about how the practice of government in an expansive and ambiguous office is influenced by a fluid and uncertain selection process and a bitterly partisan political landscape.

Let’s start by thinking about Trump’s nomination and presidency. In 2015, Brendan Nyhan observed that Trump’s success in early primary polls could be partly attributed to his ability to draw on appealing, but inaccurate, public visions of the power of the presidency. Nyhan wrote, “Mr. Trump is the purest Green Lantern candidate we’ve seen in recent years. He cleverly exploits the appeal of presidential omnipotence by contrasting his supposedly decisive style of business leadership as a real estate magnate with the compromises, inertia and policy failures that are inevitable in politics.”

While this was certainly an accurate diagnosis of the broader contrast between politics and the private sector, Trump’s presidency has exposed both the potential and the limits of such a presidency. The presidency does, in fact, have a great deal of unilateral policy-making power. Presidential communication also turns out to be important, albeit not necessarily for persuading Congress to pass bills. In their new book, Unmaking the Presidency, Susan Hennessey and Benjamin Wittes describe Trump’s approach as the “expressive presidency.” They define this primarily as personal self-expression at the expense of civil virtue, but if we combine the idea of an expressive presidency with the existing concepts of a permanent campaign and a partisan presidency, we come up with a powerful description of the office that is not entirely limited to Trump. It’s not impossible to imagine a variety of situations in which a president — one who has been especially successful at building a devoted following based on rhetoric and ideas — could continue these kinds of communications in office. Such a president could establish, over and over again, a specific vision of the country and who should lead it, and the ability of a new president to serve as the voice for the underrepresented. This type of communication is better suited to some appeals than others, but it is remarkably adaptable to different ideologies.

It would be a mistake to see this turn in the presidency as merely rhetorical, however. Unilateral action has accompanied some of the top rhetorical priorities. For Trump, these priorities include both conventional Republican items like rolling back regulations, and those specific to Trumpism like immigration restrictions. Legislative accomplishments, even with Republican majorities in Congress, have been more elusive.

It’s not that difficult to envision a Sanders presidency unfolding in a similar fashion. Obviously, the priorities and the messages would be very different. And this isn’t necessarily to criticize Sanders’ weaknesses but rather to assess the way his strengths fit into the office as it exists now. Sanders clearly gives a hell of a rally speech, inspires his followers, and offers a way of thinking and talking about politics that’s clear and consistent. Those things just don’t seem likely, based on the available evidence, to fully unite party factions or build up a durable national policy majority. To be clear, I think the evidence suggests that if nominated, Sanders could pull together the necessary coalition to win; the strength of partisanship means that nearly any competent politician could. But the tactics at which he most excels — and the ones encouraged by our nomination system — are not conducive to the style of legislative governing that dominated the twentieth century. Instead, the presidency offers some tools to make policy change, without needing to cobble together enough support to get Congress to pass a bill.

One major question is whether elite Democrats would warm to Sanders as a candidate and a president the way that Republicans have embraced Trump. If you buy the asymmetry thesis and think the Democrats are a party of groups, it might be a tougher sell, with distinct factions remaining unconvinced that Sanders represents their interests. If the argument that the Democratic Party organizes itself around egalitarian ideology is correct, or if negative partisanship inspired by Trump overrides everything else, then it’s possible that a factional candidate Sanders could create the shallow party unity that Trump has enjoyed. And such unity has been able to do a lot — as we are seeing now with impeachment — but it’s important to keep in mind that what it couldn’t do was produce Congressional majorities to pass sweeping legislative change. It’s an expressively united party under an expressive presidency.

It’s almost certain that the current nomination system allows factional candidates to succeed. What’s more, current party conditions also have the potential to turn any and all of the candidates into factional candidates. A long season with lots of debates has exposed the weaknesses and flaws of the candidates. Biden has the most endorsements, but still claims a shaky front-runner status and has failed, in the words of FiveThirtyEight’s Perry Bacon, to really inspire political elites in Iowa and New Hampshire. Biden’s detractors worry about his age, his record on many issues, and his representation of a return to “normal” that wasn’t that great for everyone. Elizabeth Warren might be able to bridge the gap between the establishment and the emerging left, but, like Pete Buttigieg, she struggles to win support among voters of color and voters without a college degree. Any of the remaining contenders could fit into this model of the presidency, offering appealing messages to their core supporters and conducting most policy out of the executive branch. Most of the discussion of what’s wrong with the nomination system has focused on producing electable candidates or qualified presidents. But it’s possible to have a president who is perfectly qualified and yet unequipped, politically speaking, to build coalitions and govern across branches. The result could be a presidency that is primarily rhetorical and unilateral, alternating with each election between distinct governing visions, producing the exact kind of instability in government that the presidency was designed to avoid.

Julia Azari is Associate Professor and Assistant Chair in the Department of Political Science at Marquette University. This article was originally published by Mischiefs of Faction.