Among the many indignities of the 2016 election, it seems we are cursed to suffer a never-ending debate over why Donald Trump won in which no one can quite agree on the meaning of statistics or causality.
In the Washington Post yesterday, Nicholas Carnes and Noam Lupu became the latest commentators to try to take down the narrative that Donald Trump’s victory can be attributed to his support among the white working class. This account is wrong, they write, for a simple reason: “Trump voters were not mostly working-class people.” If you define working class as below the median national annual income of about $50,000, then, according to the American National Election Study, only 35 percent of Trump’s voters in the general election count as white working class. “What deserves to die isn’t America’s working-class communities,” they conclude, alluding to Kevin D. Williamson’s infamous commentary in the National Review. “It’s the myth that they’re the reason Trump was elected.”
Do you notice the logical leap here? From the premise that low-income white voters didn’t make up a majority of Trump’s voters, Carnes and Lupu conclude that they weren’t the reason Trump was elected. The unstated assumption is that for a group of voters to be the “reason” for a candidate winning, those voters must account for more than half of that candidate’s votes.
But why would that be true? Consider basketball. If you ask any NBA fan why the Cavaliers trounced the rest of the Eastern Conference in this year’s playoffs, the answer will be obvious: they have LeBron James, the best player in the league. But even James doesn’t account for anywhere close to a majority of his team’s points. Does that mean he hasn’t been the reason for the Cavs’ dominance? Of course not. There are at least two ways to explain why. The first is a concept that sports stat nerds call “value over replacement”: if you replaced James with a league-average player at his position—or even, perhaps, another elite NBA player—the Cavs would instantly suck, as they did before he got there. The Cavs scored 111 points per game in the regular season and gave up 107. James basically accounts for that differential.
The second way to understand why James is the reason the Cavs win is to distinguish between two types of causal conditions: necessary and sufficient. A necessary condition is something that has to happen in order to bring about a certain result. In the legal world, this is known as a “but-for” cause: but for X, Y wouldn’t have happened. LeBron James is a but-for cause of the Cavaliers’ success: if he wasn’t on the team, they wouldn’t even make the playoffs. That doesn’t mean, however, that he’s a sufficient condition to guarantee victory. A sufficient condition is enough on its own to guarantee an outcome. Even LeBron James still needs to have some good players around him, and even then, as we’re seeing in these Finals, that may not be enough. One reason the debate around what “really” caused Trump’s victory in November has been so exasperating is that participants never distinguish between, and almost always conflate, the concepts of sufficient and necessary (or but-for) conditions.
The key thing to understand about but-for causes is that for any given outcome they may be infinite. There is only one condition sufficient to winning a basketball game: scoring more points than the other team. But you could come up with dozens of things that had to happen for the Cavaliers to win the title last year: James staying healthy, Draymond Green kicking folks in the nuts, Kevin Love not retiring mid-season to become a professional model.
This is true of presidential elections as well. Jacob T. Levy made this point back in December:
An 80,000 vote margin in a 137 million vote election, about .05%, is susceptible of almost endless plausible explanations. The number of different factors that might well have moved that many votes is very large. So there are a lot of different true but-for explanations: but for Clinton’s failure to campaign in Wisconsin, but for the Comey letter, but for stricter voter ID laws and reductions in the numbers of polling places, but for Jill Stein, and so on, ad infinitum.
If there is a myth about the election, then, it’s that there was any one “real” cause of Trump’s victory. The truth is that there were an endless number of but-for causes, and what matters is which of these factors make the most sense for Democrats to confront in order to regain power. As Nathan J. Robinson puts it in Current Affairs, on the “racism vs. economic anxiety” debate: “The truth about race and economics in the election is easy to grasp. They both mattered, and we can focus on whichever we choose”
So to say that working class whites aren’t the reason Trump won because they were “only” 35 percent of Trump’s electorate in November—25 percent, if you only count people without a college degree—is a complete non sequitur. Sure, the white working class alone wasn’t enough to elect him, but that’s true of any group, just as LeBron James alone can’t win the NBA Finals. The question is whether Trump could have won without his historically strong performance among the white working class, and whether we should care. And on that point, the answer is obvious: no, he couldn’t have, and yes, we should. As Nate Cohn wrote in the New York Times in December: “large numbers of white, working-class voters shifted from the Democrats to Mr. Trump. Over all, almost one in four of President Obama’s 2012 white working-class supporters defected from the Democrats in 2016, either supporting Mr. Trump or voting for a third-party candidate.” Martin Longman digs into this more in the current issue of Washington Monthly, showing how Trump clearly won Pennsylvania because huge percentages of voters in white, rural counties switched from Obama to Trump.
Of course, Carnes and Lupu would rightly point out that not all these white, rural voters count as working class as defined by income. But the myth that most Trump supporters are poor has already been debunked, repeatedly, for more than a year. The important point is that Trump’s overwhelming success among the white working class was a major but-for cause of his victory that, unlike the Comey letter or Hillary Clinton’s email server, actually presents Democrats with an area to improve in future elections. If the 2016 election is still worth analyzing, it’s only to the extent that the left is willing to face uncomfortable questions about Trump’s appeal—or Democrats’ lack thereof—to communities that won’t die anytime soon.