When Trump was elected, and for many months following, people kept on talking about why he was elected. What caused it? This conversation irritates me deeply, because people lack a base-level understanding of what causation is. But I’ve waited to say anything because I thought it might be too crass to insert a philosophical discussion into a political one, at least while it was still hot.

Cause and effect is often thought to be a fundamental part of the way the world works, but I and other physicists understand that it is not. For a brief explanation, I recommend this video by Sean Carroll. It is better to think of causality as an emergent property, more in the realm of philosophy than physics.

What does physics have to say about the cause of Trump’s election? It’s everything in Trump’s past lightcone! It was the DNC, it was Clinton, it was Comey, it was Russia, it was neoliberalism, it was identity politics, it was ancient supernovae. This answer is rather naive, but what did you expect from us? Physics can’t provide all the answers.

When we talk about causes, we’re typically just selecting a few things from the past lightcone, and highlighting those things as important. In philosophy, this is known as causal selection. Sean Carroll talks a little bit about causal selection. He says that one way of thinking about it is that a cause is something that has great leverage over the future. But that’s just one way we might think about it.

One of the respectable arguments about causal selection is the argument from caprice: Causes are selected so capriciously, that causal selection is groundless. While I do not think this argument is entirely correct, it is easy to see that causes are indeed selected capriciously, often to suit both the context and goals of the speaker. For example, in one context, Carl Sagan may say, “We’re made of star stuff,” which selects ancient supernovae as the proper cause of Trump’s election. In another context, a person might say, “butter e-mails,” which selects Clinton’s most buttery males as the proper cause.

In these arguments about Trump, they mostly follow this structure:

1. If it were not for the excessive focus on identity politics, Trump would have lost.
2. Therefore, excessive focus on identity politics caused Trump to win.
3. Therefore, we should focus less on identity politics.

The problem with this argument is that it equivocates two inequivalent notions of causation. #1 describes what legal scholars call “but-for” causation: if not for the cause, then the effect wouldn’t have happened. #3 describes a moral notion of causation, where positive consequences are selectively traced to good actions that we should continue doing, and negative consequences are selectively traced to bad actions that we should try to change. For example, if I think buying lottery tickets is wrong, I am more likely to attribute winning the lottery to luck, not buying lottery tickets (even though both are required).  This is because I want to discourage people from buying lottery tickets.

Sometimes people also bring in a third notion of causation, the idea that each event has a unique cause.  This is also inequivalent to the other two notions of causation.  So whenever people argue that Trump’s election wasn’t caused by X because he was caused by Y, that’s yet another fallacious argument.

When we talk about the election, it’s clear that but-for causation is insufficient. The election was really close. You can identify lots of things that might have swung the election, but that doesn’t mean that we should have changed every single one of those things.  The Democrats could have won if they had joined Republicans in nominating Donald Trump, but that doesn’t mean they should have.

Basically, when people argue about the causes of Trump, it’s clearly a moral argument, not just a factual one.