The thing about electoral politics at the national level is no one knows who’s going to win. Indeed, there are lots of people who say they know. There are lots of people with the apparent authority of data science who say they know. But no one really knows.

We do not know what we cannot know until the moment in which it’s possible to know it. (Indeed, we may not know the moment of knowing has arrived until after the fact.)

At the same time, not knowing who’s going to win feels terrible given that a lying, thieving, philandering sadist occupies the White House, and that the current occupant has already cheated twice to win the presidency. That terrible feeling will grow more acute and more intolerable the closer we get to Election Day. Given these conditions, lots of people say they know for the benefit of people who can’t tolerate not knowing. The problem is basing our voting decisions of what people say they know but can’t.

---

We should focus more on what’s knowable—now—and base our decisions accordingly.

---

If we learned anything from 2016 (and I think it’s still to be determined whether we have), it might be that’s what happened. Lots of Democrats thought Hillary Clinton was going to win, so lots of them didn’t bother showing up on Election Day. They didn’t bother showing up, because they had been told, over and over, that the lying, thieving, philandering sadist occupying the White House would never do any such thing.

In fairness, the people who say they know almost always hedged their bets in 2016. But those hedges were almost always buried somewhere in the 14th paragraph of what was otherwise a very serious and very confident prognostication of future events the prognosticator could not possibly know because the moment in which knowing was possible had not yet arrived. Combined with other factors (an obsession with Clinton’s emails and a flawed understanding of polling among voters), the press corps’ zeal for data science inflated a bubble of magical thinking that popped on Election Day.

None of this is to say we shouldn’t trust polling. It is to say we shouldn’t trust it that much. We should focus more on what’s knowable—now—and base our decisions accordingly. For instance (and I do not intend to insult your intelligence), what we know is Donald Trump is the incumbent, and that incumbents have an enormous advantage in all cases. We know this is especially true in this case. Trump has an international right-wing media apparatus willing and able to lie for him nonstop.

We also know this president has never been popular (under 50 percent approval from the get-go). He has been impeached. He will be acquitted without exoneration. We know he cheated twice, once in 2016 when he invited Russian sabotage, and once in 2019 when he extorted the Ukrainians into smearing his closest 2020 rival. We know, given the Republicans in the Senate failed to punish him, that he will cheat again. We know the Russians, or some other adversarial nation, will try moving public opinion against the Democratic nominee. We know it doesn’t matter who that person is.

We know, given this is an incumbent’s reelection campaign, the Democratic nominee is going to have pretty much the same coalition of voters Hillary Clinton had in 2016. We know once again it doesn’t matter who that person is. We know the president will smear the Democrat no matter who that person is and no matter how hypocritical his smearing will be (i.e., he might call the nominee fat though he himself is quite obese.)

There are some stable things in the world despite the mass of uncertainty we face. Democratic voters should remember these as they participate in today’s Iowa caucuses or watch their results from afar. It probably doesn’t matter who the nominee will be, because any Democrat will face the same challenges and opportunities. What’s key, more than anything else, is maintaining unity after it’s all over but the shouting.

If there’s anything I would say with respect to which candidate would be the best choice, I would say this: Americans tend to be retrospective, not prospective, voters. These are terms political scientists use to describe whether voting behavior is based on past events or future possibilities. For the most part, Americans tend to think about the past, often a nostalgic, sentimental or mythical past. They do not consider the future, generally, unless it is a revival of a nostalgic, sentimental or mythical past.

Moreover, the past tends to be seen as normal. And given that our past is viciously bigoted, bigotry is normal. We should all of us bear this in mind though it may be hard to accept. Most American voters, generally speaking, will choose what’s familiar over what’s unfamiliar, even if what’s familiar is a lying, thieving philandering sadist.

This argument, of course, depends on what “normal” and “familiar” mean. Can Bernie Sanders sell his brand of “socialism” as something familiar? Can Joe Biden convince voters he can lead America back to normalcy? I don’t know. Neither can anyone else.

We can’t know until it’s too late.

—John Stoehr