It's race and immigration, stupid

Under prevailing theories of economic voting, “Make America Great Again” should not be a good slogan for a time when poverty and unemployment are falling, and household income is rising, as the latest Census report shows. The fact that Trump’s faux-populist appeal resonates at all despite this should be evidence enough that it’s not about the economy. In any case, Trump’s more explicit promises about walls and religious tests should make it clear what American greatness really means to him.

Yet there are many who take Trump’s claims to an economic populist mantle seriously, including some on the left. So eager are they to find any evidence at all of a burgeoning American class consciousness that they project it onto an unlikely source, one who spends more on hair transplants than most Americans make in a year.

The argument, made by Frank, Taibbi, Moore, and others goes roughly like this: trade is a huge issue ignored by Washington elites. Trump supporters are those who have been economically left behind, particularly by trade. Liberals who say it’s about racism and nativism have their heads in the sand. Trump supporters are simply voting their class interests. This is why Trump will win.

I don’t believe any of these things are true. For better or worse, trade is not a major issue in American politics. Neither American voters generally nor Trump voters specifically rank it among their top concerns: only 36% of Trump supporters polled said trade was a critical issue to them, well behind terrorism and immigration. Trump supporters are not, overall, economically worse off. A Gallup study found Trump supporters have higher median incomes than non-Trump supporters, and are no more likely to live in areas affected by trade. Curiously, wealthy Trump supporters report feeling the same level of economic anxiety as poor non-Trump supporters – which is to say, they think they’re doing worse than they are. This is probably because, as political scientist Michael Tesler’s research shows, voters’ perceptions of how the economy is doing have a lot more to do with their party identities, and attitudes on race, than how the economy is actually doing.

And when you test economic v racial issues against each other as predictors of Trump support, the significance of the economic ones disappears. It isn’t about trade, it isn’t about jobs, it isn’t about economic performance. It’s about demographics, and the sense that the country is looking less and less like what a portion of the electorate imagines it should look like – a perception that is also exaggerated, since Trump supporters are also no more likely to live in areas affected by immigration.

Let’s be clear what this does not mean: it does not mean that Trump supporters are all cross-burning racists. It does not mean that racial resentment is exclusive to Trump supporters; Daniel Byrd and Loren Collingwood found some (though not as many) of Clinton and Sanders supporters reported similar attitudes. It does not mean “America is already great” or that working class Americans haven’t been left behind. It does not mean that there are not many people who support Trump because they feel (real or imagined) economic stress. It means that for every person who votes for Trump because they feel economic stress, there is another person who votes for Clinton for the same reason. If all you know about someone is how hard they were hit by the recession, you’re no more likely to predict their vote than a coin toss. But if all you know about them is how they feel about black people and Muslims, you can make a pretty good guess.

So why the credulity on the part of some pundits? It may be a suspicion of polling evidence left over from the primaries (remember, the polls didn’t get Trump wrong, the polling analysts just ignored them), or part of a broader, leftists v liberals culture war over identity politics. Those to the left of the Democrats have been pushing back on what they see as a disproportionate focus by liberals on race, gender, and sexual identity to the detriment of economics. And in this election, there is a justifiable sense that efforts to caricature Trump supporters as Klansmen is an easy way for Democrats to excuse their neglect of their working class base. But replacing that caricature with a new one, that of saintly, if slightly confused, characters from a Steinbeck novel, has even less basis in reality. It’s one thing to say that Americans should vote their class interests more. It’s another to say they do. One is a valid normative argument (which I agree with), another is observably false.

Americans have never truly voted along class lines; a working class vote divided by racial issues is not the exception in American politics but the rule. But Trump has a lot of people confusing what is with what ought to be. A better case can be made for a pro-worker platform than fixed ideas about Trump voters. Politicians should cater to working class interests more not because there is, or ever has been, such thing as a unified working class vote, but because it’s the right thing to do.