A Trump rally in 2016. Brett Carlsen/Getty Images
Hint: It has to do with race.
One of the most puzzling elements of the 2016 election, at least for a lot of Americans, was the millions of voters who switched from voting for Barack Obama in 2012 to Donald Trump in 2016. Somewhere between 6.7 million and 9.2 million Americans switched this way; given that the 2016 election was decided by 40,000 votes, it’s fair to say that Obama-Trump switchers were one of the key reasons that Hillary Clinton lost.
The existence of those voters has served as evidence that the most plausible explanation for what happened in 2016 — that Trump’s campaign tapped into the racism of white Americans to win pivotal states — is wrong. “How could white Americans who voted for a black president in the past be racist,” or so the thinking goes.
“Clinton suffered her biggest losses in the places where Obama was strongest among white voters. It’s not a simple racism story,” the New York Times’s Nate Cohn wrote on the night of the election. This typically segues into an argument that Trump won by tapping into economic, rather than racial, anxiety — anger about trade and the decline of manufacturing, or the fallout from the 2008 Great Recession.
A new study shows that this response isn’t as powerful as it may seem. The study, from three political scientists from around the country, takes a statistical look at a large sample of Obama-Trump switchers. It finds that these voters tended to score highly on measures of racial hostility and xenophobia — and were not especially likely to be suffering economically.
“White voters with racially conservative or anti-immigrant attitudes switched votes to Trump at a higher rate than those with more liberal views on these issues,” the paper’s authors write. “We find little evidence that economic dislocation and marginality were significantly related to vote switching in 2016.”
This new paper fits with a sizeable slate of studies conducted over the past 18 months or so, most of which have come to the same conclusions: There is tremendous evidence that Trump voters were motivated by racial resentment (as well as hostile sexism), and very little evidence that economic stress had anything to do with it.
This isn’t just a matter of historical interest or ideological ax-grinding. Understanding the precise way in which racism affected the 2016 election should shape how we think about the electorate in the run-up to the 2018 midterms. More broadly, it helps us understand the subtleties of America’s primordial divide over race — and why racism will continue to fracture the country politically for the foreseeable future.
The study found strong evidence for racism — and little for economic anxiety
The three scholars who wrote the study — UCLA’s Tyler Reny, UC-Riverside’s Loren Collingwood, and Princeton’s Ali Valenzuela — drew on a database that has information on more than 64,000 American voters. Inside that huge sample, they restricted their analysis to white voters who switched their presidential vote from 2012 to 2016 (most commonly from one major party’s candidate to the other’s, but occasionally from a third party in 2012 to Clinton or Trump).
They then split the sample of white voters in two, between working-class and non-working class voters, and then tried to figure out what the vote switchers ran in common. To do so, they ran tests on three different types of question: scores on a test measuring attitudes towards racial minorities, hostility to mass immigration, and measures of economic stress (e.g., whether a person’s family income was lower or higher than the median income in the county where they lived).
The results were quite striking. First, attitudes on race and immigration were crucial distinguishing characteristics of both Trump and Clinton switchers. The more racially conservative an Obama or third party voter was, the more likely they were to switch to Trump. Similarly, the more racially liberal a Romney or third-party voter was, the more likely they were to switch to Clinton.
Second, class was largely irrelevant in switching to Trump. Keeping racial attitudes constant, white working-class voters were not more likely to switch to Trump. The white working-class voters who did switch tended to score about as highly on measures of racial conservatism and anti-immigrant attitudes as wealthier switchers.
Third, the correlations between measures of economic stress and vote switching were either weak or non-existent. There’s just little evidence supporting the “economic anxiety” or “economic populism” explanations for the Trump surge.
“We find a much stronger association between symbolic racial and immigration attitudes and switching for Trump and Clinton than between economic marginality or local economic dislocation and vote switching,” Reny et al. write. “In fact, we find marginally small or no associations between any of our economic indicators and vote switching in either direction.”
These findings reveal the subtlety, and importance, of America’s racial divide
A 2018 demonstration in Chicago supporting the conviction of police officer Jason Van Dyke, who shot 17-year-old Laquan McDonald in 2014. Joshua Lott/Getty Images
The Reny et al. findings may seem counterintuitive: How can people who wanted a black man to run the country somehow become attracted to Trump because of his racial demagoguery?
The unspoken premise behind this question is an assumption of a certain kind of white redemption narrative: By voting for Obama, white America exorcized its racial demons. But the truth is nothing of the sort. For one thing, Obama lost the white vote by 12 points in 2008 and 20 points in 2012.
For another, voting for Obama once or even twice doesn’t automatically mean that someone is not prejudiced against black people or immigrants. It’s possible to support Obama in particular while maintaining overall anti-black or anti-immigrant attitudes. In those cases, some other factor, like the Iraq War catastrophe or financial collapse, may have predominated over white voters’ racial hang-ups in the 2008 and 2012 election.
The 2016 election was different.
One reason is that Obama’s second term featured a significant amount of racial conflict. The Black Lives Matter movement was founded in 2013. The 2014 killing of Michael Brown in Ferguson, and subsequent week of protest and unrest, kicked off a massive and racially polarizing national debate over police violence against African Americans.
A second reason is that Obama’s very presence in office was racially polarizing. Michael Tesler, a scholar at the University of California-Irvine, has documented in detail how Obama’s very presence in the White House polarized America along racial lines. It would make sense that this effect would grow stronger the longer Obama was in office, setting the stage for a major backlash in his final year.
Third, and arguably most importantly, the two candidates turned the election into a kind of referendum on American race relations. Trump kicked off his campaign by calling Mexican immigrants rapists and vowing to build a wall between the US and Mexico. He vowed to ban Muslims, and described black life in America as a hellscape of violence and poverty. Mitt Romney’s 2012 campaign was not nearly so overt, which means it was less likely to attract voters who held latent racist and anti-immigrant attitudes.
Clinton, for her part, positioned herself as a champion of racial justice. While Obama’s rhetoric on race was typically post-racial, positioning the country as more united than divided, Clinton got out front on issues like police violence and immigration. There are plenty of valid reasons for this — Clinton was more worried about failing to turn out minority voters, Obama was more worried about alienating skittish whites, and there was no way to respond to Trump’s campaign without tackling race head-on.
The result, though, is that racial issues became the key political dividing line in a way they were not in either 2008 or 2012.
Now, Reny et al.’s statistical analysis can’t show all of this on its own. You should never draw conclusions this large from one statistical analysis, as it could suffer from any number of problems.
However, this analysis of the election is supported by a wide and deep body of research, the vast majority of which shows that concerns about identity and race were the decisive issues in the 2016 election. This was true in the Republican primary and the general; it’s also consistent with research on far-right parties in Europe whose xenophobic appeals are similar to Trump’s. There is a complete lack of statistical evidence, by contrast, for the “economic anxiety” theory.
American politics are likely to only get more polarized on racial lines. Trump and Trumpism are, for the time being, the core of the Republican Party; the Republican message on race and immigration will match his as such. California Rep. Duncan Hunter, for example, is running a nakedly anti-Islam reelection campaign against Democratic challenger Ammar Campa-Najjar (who is a Mexican-Arab Christian by background).
The implications, both in 2018 and in the long term, could be significant. Reny et al. compare this period to the post-civil rights era, a period where the historically Democratic South transformed into modern-day red America primarily in backlash to the Democratic embrace of civil rights:
History suggests that significant changes in voting across party lines, particularly for the presidency, precede changes in party identities, the basis for realignments. This sequence of events played out during the Southern realignment (i.e., Democrats voting for GOP presidential candidates but maintaining their party attachment) and here we provide evidence that it may be happening again after two terms with a black president and during an era of mass demographic change due to immigration. Racial conservatives and those with the most punitive immigration views are moving right and were the most likely to switch to Trump in 2016. Our data suggest the same is happening in the opposite direction as those with racially liberal or pro-immigration views may be sorting into the Democratic Party.
This prediction may or may turn out to be accurate. But it’s plausible, and there’s no use burying our heads in the sand by pretending this is about class when it isn’t.
