Cum hoc ergo propter hoc
How can we begin to understand the causes of things?
The sociopolitical world is a complicated place. Do people behave according to predictable patterns? Why do certain phenomena precede and succeed other things? Why do people do what they do?
Science says that we can understand these things as long as we test them rigourously. We first must adopt a scientific attitude: impartiality, humility, and scepticism are crucial. So is a commitment to proposing falsifiable theories, trying to disprove them, and remaining open to new evidence.
Science also tells us that even if you find patterns in the data—in politics, this might look like a trend in messaging—you can’t be certain that this pattern caused some other phenomenon—perhaps electoral success?—with which it’s correlated. Correlation is not causation. Unfortunately, many progressives have relied on this logical fallacy to explain the MAGA phenomenon. And in doing so, they’re able to wash their hands of any complicity in MAGA’s success.
Bigotry vs. economics as an explanation of MAGA
I’ve already written extensively on why I think Trump’s success, and that of the MAGA movement writ large, can be better explained by economics (particularly a form of retrospective economic anxiety that gripped much of de-industrialized America since the 1990s) than by racial resentment. Individuals vote for a plethora of reasons, and certainly some do it out of racism. But the broad trend is clear: the data confirmed in 2016 that economics was more important than racism as a predictor of voting for Trump, and it’s even more clear following Harris’s defeat (in which voters of colour, a majority of whom still voted blue, swung toward Trump compared to 2020). And as the working class firmly joins the GOP camp, economics looks like an even better explainer of American politics.
This is not to scapegoat these voters or any others for the election result. People who vote for the candidate you don’t like are not evil. It’s hard to understand why people vote the way they do, but chalking it all up to some kind of ultimate, Manchaean malice is unscientific.
My original essay on this subject was inspired by an article by James Risen in The Intercept, in which he listed all the ways in which Trump was horrible and bigoted and then asserted that people voted for him due to his horribleness and bigotry. Risen didn’t cite sufficient data for his claim. So, I looked into the data, and it turned out that Risen, and many other mainstream liberal commentators, were wrong about this. Economic anxiety did a better job predicting support for Trump in 2016 than racism.1
Risen was committing the cum hoc ergo propter hoc fallacy: with this, therefore because of this. Correlation equals causation. Trump is racist; people vote for him; therefore they vote for him because he’s racist.
Science
Causality is very hard to determine. You need to establish correlation, temporal precedence, theoretical plausibility, non-spuriousness, and replicability.2
We’ve been dealing with the tough sociological question of why people like Trump since 2016, but some people still refuse to look at the data or address the statistical criteria for causality. Roxane Gay recently wrote:
Mr. Trump’s election demonstrates how American tolerance for the unacceptable is nearly infinite. There are hundreds of absolutely mind-boggling things I could point to from the past decade — the suggestion of bleach injections to potentially treat the coronavirus and the wild QAnon conspiracy theories infecting millions of Americans, including politicians, and insulting veterans and making fun of the disabled. But three elections in a row, Mr. Trump has been a viable presidential candidate and our democracy has few guardrails to protect the country from the clear and present dangers he and his political appointees will continue to confer upon us.
Clearly, Mr. Trump is successful because of his faults, not despite them, because we do not live in a just world.
No. It’s far from clear that Trump won because of his faults. “We don’t live in a just world” is not sufficient evidence for this claim. I’m frustrated the supposedly legendary New York Times published this. You cannot merely list a bunch of qualities you dislike about Trump and then assert that people voted for him because of those qualities.
What can you do, and what Gay is quite right to do, is point out that “Mr. Trump’s election demonstrates how American tolerance for the unacceptable is nearly infinite.” I agree. Huge swathes of America are willing to tolerate a shocking of sexism, racism, nativism, incompetence, and dishonesty. I’m not writing off Trump supporters as potential partners in building a better, more just America one day. But these trends are concerning.
What to do?
Trump went up against the qualified, smart, and earnest Harris. Unfortunately, she was a mediocre candidate3 who had the wrong electoral strategy and only ~100 days to campaign.4 And I think her silence on America’s complicity in heinous crimes around the world is highly condemnable. But I don’t doubt her good intentions, and her policies were vastly preferable to and more popular than Trump’s.
So, Trump’s victory and popularity probably require an outstandingly powerful explanation. I will continue to keep my mind open to new explanations that can be gleaned from data, and I encourage others to do so as well. Otherwise, we risk writing off our country and giving up completely:
Time and again, we hear the wild lies these voters believe and we act as if they are sharing the same reality as ours, as if they are making informed decisions about legitimate issues. We act as if they get to dictate the terms of political engagement on a foundation of fevered mendacity.
Gay seems to forget that no matter how much you hate Trump voters, they—as the electoral majority—do get to dictate the terms of political engagement. They still have the right to vote just like us, for which I’m eternally grateful. I’m a big democracy advocate.
Perhaps Gay has great ideas on how to solve the MAGA problem. I will attentively read them when she publishes them and learn from them what I can. For the time being I’m convinced that we should strive to right the ship of state through Constitutional, nonviolent, democratic means. We should change our messaging, policies, and candidates to better reflect the wishes of the electorate. I think it’s far from clear that Democrats have to abandon the just causes of advocating for persecuted ethnic and gender minorities to take back the electoral majority.
For example, we could start by canning the corporate lobbyist who’s our party chair. We could run candidates with net positive approval ratings. We stop morally grandstanding and condemning people for widely held social views. We could listen to the vast majority of Americans who are pessimistic about their future wellbeing, instead of scorning them for not understanding macroeconomics.5
Yes, Trump voters believe insane things. Yes, they tolerate bigotry. Are they the most coddled voting bloc ever, as Gay writes? Perhaps. But “coddle” is another way of saying “appeal to.” And in a democracy, politicians must appeal to the majority if they wish to govern.
If you’re not convinced, please read my essay on the subject. I go in depth on the linear regression analyses used by political scientists and sociologists to understand voter motivation.
Correlation: things happen together in a patter.
Temporal precedence: the causes precedes the effect.
Theoretical plausibility: there’s a logical explanation of how the cause makes the effect.
Non-spuriousness: other plausible explanations of the effect have been ruled out. This is the important one.
Replicability: the causal effect is seen over and over again.
She lacked the populist appeal and congeniality of Obama, the policy chops of Bernie, and the experience of Biden. I write this all wearing my Kamala Harris t-shirt.
Which is why we should have had a primary!
To be clear, these suggestions seek to address possible confounding variables that could explain Trump’s success other than racism, which I view as largely spurious and not causal.
Very thoughtful as always. It remains to be seen if the dems will learn from this in 4 years. Didn’t happen the last time for sure.