New Yorker magazine demonstrates a naive faith in social science

David Remnick writes:

How you interpret and prioritize the cascade of reasons for Trump’s reëlection is a kind of Rorschach test. It will require a long reckoning before anyone can conclude which of the leading factors—economic anxiety, cultural politics, racism, misogyny, Biden’s decline, Harris’s late start—was determinative.

Here I don’t want to talk about the content of Remnick’s article—much has and will be written about the new administration, from many political perspectives—, but rather on the implicit faith he shows in an impossible version of social science. An attitude which perhaps should be no surprise, coming from a magazine that regularly publishes Malcolm Gladwell.

Two things from the above-quoted sentence jump out at me.

First, the final word, “determinative.” Trump won the popular vote nationwide and in the decisive swing state by a little less than 2 percentage points, which, yeah, is close, so it makes sent to look at how different factors could’ve affected the outcome. But the idea that one factor was “determinative” . . . that’s just nuts. There’s no “determinative” here. Lots was going on.

Second, the phrase, “It will require a long reckoning before anyone can conclude,” which implies that there’s some true answer that will at some point be figured out. Now, let me be clear on this: often in history there is a true answer, and sometimes, with care and effort, researchers can figure it out. For example, here’s Walter Mebane’s estimate of the actual number of votes for Bush and Gore in Florida in 2000. At the time of the aborted vote counting, there was a huge uncertainty into what the voters had actually wanted, but the actual tally of vote intentions now seems clear. This is separate from the legal question of what should’ve been done in November and December, 2020, but it’s an example where social science can, with time, resolve a potential factual dispute.

In this case, though, there’s no clear question to answer, and the idea that, after some “long reckoning,” anyone can conclude “which of the leading factors . . . was determinative,” makes no sense. Or, to put it another way, “anyone can conclude” anything at any time, but no long reckoning is required; indeed, the takes are coming thick and fast on the op-ed pages every day.

What can social science do?

To step back for a moment, my problem with the above-quoted snippet from the New Yorker article is that it seems to me to imply a faith in a simplistic version of social science, in which: (a) when something happens, there’s one factor that determines it, and (b) that, in time, people will figure out what that factor is.

I’m cool with the attitude that, with time, historians and social scientists can get a better understanding, both about what actually happened (looking at many data sources, not just exit polls and geographic vote totals) and about what could’ve happened otherwise (causal inference, counterfactuals and all that). So I’d’ve been cool if Remnick had just said something like, “It will require a long reckoning before we can have a clear sense of what actually happened during the campaign and the election, and before we can come to informed guesses about how things could have gone differently.”

Picky, picky?

Arguably, that last sentence of mine is what Remnick actually meant to say; he just wrote something that was a little bit sloppy, excusable given that he’s not a social scientist and also he was writing on deadline. His job is not to get all the nuances right; he’s writing for the New Yorker, not the American Political Science Review, after all, and if he uses the phrase “conclude which of the leading factors . . . was determinative” as shorthand for “how things could have gone differently,” then, no big deal, he got his point across.

The Speed Racer principle

So I’m not actually saying that Remnick did anything wrong here. Rather, what’s interesting to me is that I think his phrasing represents an implicit belief in social science as a way of finding the “determinative” factor.

Remembver the Speed Racer Principle? Sometimes the most interesting aspect of a cultural product is not its overt content but rather its unexamined assumptions. I think this New Yorker quote is interesting in revealing an unexamined assumption about how the social world works, and what social science can do.