Predicting the past: Can Trump win again?

[I wrote this piece on Election Day 2020, before the vote-counting started. Less than 24 hours later, it was looking like not a single one of my insights would be borne out by events.]

electoral poster in support of Any Functioning Adult

With the US polls so clearly pointing one way, it doesn’t take a brave man to predict a Joe Biden victory. But my own belief that Donald Trump won’t win pre-dates Biden’s poll lead. I was convinced there would be a change at the White House even at the start of the year when Corona was only a drink and Biden was looking like an also-ran in the Democratic primaries.

Why am I so convinced? Three reasons. Any one of them should be enough to put an end to the Trumpocracy. If all three combine, it could be a landslide against him.

There was, and is, no
way to test my theory

First, the Trump base. The “base” loves Trump and Trump loves nothing more than holding a rally with his base. He seems to love rallying for the right to govern even more than he loves the act of governing, itself. But the base isn’t big enough. It wasn’t big enough in 2016; it just happened to be in places that took the Democrats by surprise. The surprise factor isn’t there this time around and there appear to be no signs that the base has grown. The signs are that it has shrunk.

Sitting here in the UK, I may not be looking at the best data. But what I do see suggests that, whilst many of Trump’s supporters in 2016 still love him in 2020, many don’t. I have yet to see reports that Trump has won many – if any – over to the orange side.

So one key statistic I shall be looking out for as the results come through is whether the absolute number of votes for Trump has gone up or down, regardless of the percentage he gets.

Second, the negative vote. Last time round, it wasn’t just Trump v Clinton. It was also a contest between the ABTs and Anti-Cs. Anyone But Trump on one side and those who loathed and detested Clinton on the other. Well, Hillary isn’t on the ballot this time around. The ACDs (Anti-Clinton Democrats) can vote B this time around. And I expect they will. The same goes for the neutrals who despised both candidates in 2016. Most of them despise only one of the candidates in 2020 and it isn’t Joe Biden.

Third, the numbers who turn out. The number of people who voted early this year is just under 100 million, more than double the early turnout in 2016. What we don’t yet know is whether this is just a shift to early voting because people don’t want to stand in line during a pandemic or a sign that more people will vote this time around. It’s not long until we find out.

If the number of voters is truly up by a significant amount, that is almost certainly the result of a mass realisation that apathy is not a cure for people who are sick of politicians. That can only mean more votes for Biden. (Yes, I do know that there are other candidates. But I don’t see the other 1,223 candidates as the magnet attracting a significantly higher number people to the polling booths – even if all the extra candidates manage to get their Moms and their Dads to vote for them.)

A higher turnout will lead to a result that doesn’t reflect the current polls. That’s because pollsters don’t just ask a bunch of people how they will vote and report the proportions. They adjust the numbers based on the proportions (“weighting”) from demographics who have voted previously. An influx of people who don’t normally vote will seriously mess with the weights.

After Trump’s victory in 2016, I was of the view that a re-run a day later would almost certainly have led to a different result. I reckoned there were a lot of Anti-Clinton Democrats who thought Trump couldn’t win so they just stayed at home. They wouldn’t do that if my hypothetical rerun could become a reality. And, given another month for the Democratic Party to get its act together in the rust-belt states, I felt the likelihood of a Clinton victory would have been even greater.

My theory amounts to predicting the past. There was, and is, no way to test it. Four years of a Trump Presidency may have changed people’s minds. I have argued on this page that the mind-changing won’t have worked in his favour. But if Trump does pull it out of the bag again, this blog post is quickly going to look pretty foolish.