Were Rust Belt Trump voters racist?

A study entitled “Vote Switching in the 2016 Election: How Racial and Immigration Attitudes, Not Economics, Explain Shifts in White Voting”, though in pre-print, has already been triumphantly cited to show that racism is the real cause of working-class white votes for Trump. It shows nothing of the kind. Indeed, at least where racism and economic motives are concerned, it shows nothing at all. What follows explains why.

Racial Factors:

To see the study’s weaknesses, you have to look at its foundations. Where racial attitudes are concerned, these foundations lie in the survey designed to root out ‘racially conservative’ views. The details of the survey are buried in appendix D. There we find the following test ‘items’:

The racial attitudes scale was constructed of three items in the CCES, listed below, (!=0.68).These items have an average inter-item correlation of 0.42 and all load highly together on a single factor (Q1: 0.61, Q2: 0.72, Q3: 0.62).

• “I am angry that racism exists” (5=strongly disagree, 4=somewhat disagree, 3=neither agree nor disagree, 3=somewhat agree, 1=strongly agree)

• “White people in the U.S. have certain advantages because of the color of their skin” (5=strongly disagree, 4=somewhat disagree, 3=neither agree nor disagree, 2=somewhat agree, 1=strongly agree)

• “Racial problems in the U.S. are rare, isolated situations.” (1=strongly disagree, 2=somewhat disagree, 3=neither agree nor disagree, 4=somewhat agree, 5=strongly agree)

The jargon here should not distract from the untenable assumptions underlying the test.

As for the first item, it is not racist (‘ racially conservative’ is an obvious euphemism) not to be angry that racism exists. You might believe racism deplorable without experiencing the emotion of anger. That holds even if your anger is concentrated on your own lousy situation. What’s more, even black people may not experience as much anger as the researchers apparently think appropriate. This also bears on the third item, concerning the rarity of racial problems. According to a 2009 ABC/Washington Post survey, “ Far Fewer Consider Racism Big Problem”. And this applies to blacks as well as whites:

44 percent of blacks and 22 percent of whites continue to see racism as a large societal problem. In 1996, 70 percent of blacks and 52 percent of whites held that view.

Conversely, 28 percent of whites and 15 percent of blacks in the new survey said they see racism as a small problem, or no problem at all.

This decline isn’t too surprising, given the election of a black president and a 500% increase in intermarriage since 1967. It may be wrong not to feel the anger; it may be wrong to believe that racial problems are rare. What it isn’t, is racist. Working three cashier jobs with no hope for the future is probably not great at inflaming indignation about others’ difficulties.

The racist-Trump-voter thesis is, in the researchers’ eyes, bolstered by “the evidence to suggest that white voters are increasingly perceiving the Democratic Party as the party of racial and ethnic minorities.”(p.9) But is that belief the preserve of the ‘racially conservative’? What about (black) pollster Cornell Belcher, “ who previously worked for President Barack Obama’s presidential campaigns and the Democratic National Committee”? He is reported to hold that “ Democrats should focus on turning out minority voters — starting with more political groups run and staffed by minorities”. He is quite clear that this involves putting the white working class in second place:

Despite President Donald Trump’s success with the white working class, Belcher said courting these voters shouldn’t be progressives’ first priority. “When people say they want to take their country back, we should stop pretending we don’t know what that frickin’ means,” he said. “Democrats, you’re not going to win blue-collar whites anytime soon. You’re not.”

Maybe he’s right, but he’s not racist. That makes it hard to condemn as racist white voters who believe that the Democrats are heading in the direction he recommends.

What then of the second item, the affirmation of what is often called ‘white skin privilege’? The question here is whether the respondents have any experience of such privilege. Living in an area where good jobs for white people have vanished, and where an exhausting string of crap jobs are all the future seems to hold, failure to acknowledge that privilege may be small-minded, but it too isn’t racist.

The evidence, then, does not support the view that racism is behind Trump support, even if the study’s ‘racial attitudes’ are good predictors of that support. That’s because those racial attitudes have no discernable link to racism.

Economic Factors

What then of the study’s dismissal of economic motives for supporting Trump? Its economic arguments fare no better than its racial arguments, because the assumptions behind its data collection are entirely inadequate for assessing economic motives.

The researchers find no statistically significant relation between economic status and switching to Trump. Little wonder: economic deprivation and insecurity are measured by various methods over the period 2000–2014. To adopt this time-frame is to ignore pretty much every salient economic reality that would influence rust best voters and their like.

Why? Economic dislocation goes back much further: NAFTA, an important job killer, was ratified in 1994. But the despair and frustration of working class Americans goes back much further still.

You would think that someone interested in Rust Belt votes would devote a passing glance to the Rust Belt. Apparently the researchers have no such interest. After all, according to Wikipedia,

The Rust Belt is a pejorative term for the region of the United States, made up mostly of places in the Midwest and Great Lakes, though the term may be used to include any location where industry declined starting around 1980. Rust refers to the deindustrialization, or economic decline, population loss, and urban decay due to the shrinking of its once-powerful industrial sector. The term gained popularity in the U.S. in the 1980s.

So the researchers find no great decline in the fortunes of areas whose great decline occurred largely before the period examined. So it is no surprise that people trapped in shit jobs did not see their fortunes decline after 2000; they had declined, catastrophically, already. The same holds for a younger generation who fortunes never declined, because they never had a chance at the good jobs for which these areas were once known. All the economic data shows is that a call center job in 2014 was about as good as a call center job in 2000.

That was never in dispute. The claim was never that these regions were reacting to a post-2000 slump. The story of poster child of the Rust Belt — Flint, Michigan, Michael Moore’s point of departure for his belief that Rust Belt disappointment would propel Trump to victory, is described as follows. In the 1960s, and 1970s, things were fine.

But then came the recession of the 1980s, along with high gas prices, that tanked sales for US automakers. Meanwhile, Japanese companies saw their fuel-efficient cars get a boost in the American market. GM’s employment began to drop by 1988, with plant closures throughout the 1990s and 2000s. As thousands of people in Flint lost their jobs and many were forced to move away for employment, White said the decline left a visual mark with a new cityscape of empty factory lots and abandoned houses.

So if there were economic motives for backing Trump, the researchers’ time frame is bound to overlook them. Those motives have to do, not only with real or perceived worsening of economic status, but also with an economic status that worsened earlier, or, for many, with an economic environment already bleak at the dawn of this millennium.

But why, you might ask, would working class voters suddenly move to Trump, when they could have voted Republican much earlier? Was it because, at some time in the Obama presidency, they suddenly rediscovered their deep-rooted racism? That odd hypothesis isn’t needed. Trump was the first anti-globalization candidate, Democratic or Republican, to come along in many decades. Someone concerned about jobs, whether or not Trump would actually deliver them, had reason to hope he would, and reason to back him.

The study examined here, so triumphantly promoted, has the air of a study written to be triumphantly promoted. That’s counterproductive. Opposition to Trump will not be served by misinformation about the roots of his victory. Perhaps swing voters were all virulent racists; perhaps they had no economic motives. But the study doesn’t advance those claims a single inch.