The Meaning of Trump's Election Has Been Exaggerated
The consequences of the 2016 elections are assuredly significant, but the causes of the surprising outcome have been widely exaggerated. Post-election commentary includes words such as “autocracy,” “civil war,” “tyranny,” “fascism,” and “doom.” Fortunately for our country, the use of such words reflects a misperception of the American electorate and how it voted in 2016. This erroneous conception stems from a common tendency to assume that a consequential election only results when a major segment of the electorate intends those consequences.
The 1964 Democratic landslide, for example, ushered in “The Great Society,” and the Republican wave in 1980 paved the way for the “Reagan Revolution.” In majoritarian systems, however, consequential elections do not require electoral upheavals; even small changes in the vote can produce major changes in institutional control and the direction of public policy. This happened in 2016. Only two Senate seats and six House seats changed hands, and Hillary Clinton won the popular vote by a margin only 1.7 percentage points smaller than Barack Obama’s 2012 margin. If the national popular vote determined U.S. elections, the commentariat would have interpreted 2016 as a status quo election.
In the U.S. electoral system, however, where the votes are cast takes on critical importance. Clinton lost six states Obama carried in 2012. In particular, three states adjacent to the Great Lakes -- Michigan, Pennsylvania and Wisconsin -- voted for Trump by a combined total slightly less than 78,000 votes. Suppose the Clinton campaign had followed the advice reportedly offered by Bill Clinton and devoted more attention to these states, managing to switch 39,000 votes from Trump’s column to Clinton’s. Properly distributed, those votes would have given Clinton an Electoral College victory and post-election commentary undoubtedly would have emphasized the wisdom of the American electorate in rejecting racism, sexism and demagoguery.
Those 39,000 votes constitute a minute portion (.03 percent) of the national popular vote, a fragile basis on which to rest claims about impending autocracy, tyranny, civil war and fascism. Undoubtedly, the 2016 election was consequential, but it neither required nor reflected any sea change in the American electorate. Although little time has passed since the election, enough data is available to cast doubt on the most troubling claims made about Americans who voted for Trump.
Much election commentary reflects a distorted view of the American public. The key to an accurate understanding of elections is the realization that compared to the electorate at large, nearly everyone you see on TV or read about in the newspapers is abnormal -- abnormally interested, abnormally involved and abnormally opinionated (yes, almost every reader of this article is abnormal). Not so the general public. Consider that there are about 235 million eligible voters in this country. Only 1 percent of them subscribe to The New York Times or the Wall Street Journal. About 1 percent watch Fox News or MSNBC’s Rachel Maddow on any given night. Anderson Cooper and Don Lemon of CNN have about the same-size audience as programs that that run on Nickelodeon at the same time. And as print and television news ratings decline, communications studies report that the Internet has not picked up the slack. When people report that they get “news” from social media, they generally mean sports, weather, celebrity happenings and the like, rather than politics, government and foreign affairs. The Pew Research Center reports that less than 4 percent of adults consider Twitter an important source of news. Research consistently finds that campaign events -- like the James Comey letter and the “Hollywood Access” recording in 2016 -- that dominate media coverage have little or no effect, in part because many voters are barely aware of them.
Statistically normal people who are going about their lives and paying little attention to politics are not newsworthy. Far more newsworthy are those who constitute the political class: the small minority of candidates, self-appointed activists, donors, demonstrators, partisan media commentators and other people who are deeply involved in politics. Statistically speaking, these outliers are the public face of politics and government, and however inaccurate, their explanations and narratives become the dominant ones.
Those most dismayed by Trump’s election tend to attribute it to mobilization of the “deplorables.” Vox commentator Jenee Desmond-Harris is representative of this view: “Donald Trump has won the presidency, despite an unprecedented level of unfitness and in defiance of nearly every prediction and poll,” she wrote at the end of the campaign. “And he’s done this not despite but because he expressed unfiltered disdain toward racial and religious minorities in the country.”
Were some Trump voters motivated by prejudice? Of course they were (as were some voters in Britain, France, Germany, Austria, even Denmark and Sweden), but that is nothing new; the question is whether prejudice played an especially prominent role in 2016. It’s a sensitive topic, fraught with controversy even in the academic literature, but the claim that Trump tapped into a deep vein of prejudice is problematic.
According to a widely used measure of racism in the American National Election Study, Trump voters in 2016 were virtually identical to Romney voters in 2012; if anything, they scored a bit lower. Of 676 counties that had voted twice for Obama, 209 voted for Trump. On average, these counties were more than 80 percent white. The American National Election Study and an Economist/YouGov Panel both report that Clinton lost one-third of Obama’s 2012 white supporters. How does racism explain why white voters who had voted for a black Democratic president did not vote for a white Democratic presidential candidate?
The evidence for other forms of prejudice playing an unusually important role is similarly lacking. Given Trump’s insulting remarks about Latinos, many analysts expected a surge in Latino turnout and a plunge in the minority percentage that would vote Republican, but neither happened. Different survey companies have argued about the exact numbers, but it appears that Trump did not do significantly worse among Latinos than Romney. While Latino activists receive considerable attention from the media, Latinos are not as politically homogeneous a group as African-Americans. Many have been in the country for generations, and surveys show that many have moderate to conservative views on the issues, especially social issues.
If racism and ethnocentrism did not propel Trump to victory, then perhaps sexism did. According to liberal writer Christina Cauterucci, this assertion was self-evident: “For anyone who voted for Donald Trump, bald-faced racism and sexism were not the deal-breakers they should have been,” she wrote in Slate. “Hatred of women was on the ballot in November, and it won.” Here again, the evidence is surprisingly weak. The gender gap in the 2016 voting was slightly larger than in other recent elections, but smaller than many expected given the first female presidential candidate. The proportion of white Obama voters Clinton lost to Trump was virtually the same among men and women. Moreover, the exit polls reported that 51 percent of white women voted for Trump.
To explain this seeming anomaly, some feminist writers have argued that women are victims of “false consciousness,” to use Marxist terminology. Other commentators rejoin that women have multiple identities. For a married mother struggling to make ends meet, the gender of the candidates may not rank high on her reasons for voting. Still others suggest that “woman” simply was not the first thing that came to mind when people thought about Hillary Clinton. As women’s studies scholar Jennifer Lawless commented: “People have vehement reactions to her in one direction or another, and have for 20 years. So I’ve often said that if people are fundamentally opposed to her, I’m not convinced that it’s sexism; it could be ‘Clinton-ism.”
Another explanation for Trump’s appeal lies in the policies he advocated: restrictions on immigration, particularly Latino and Muslim immigration, withdrawing from international agreements like trade deals and climate initiatives, and downsizing the U.S. role in international politics. Determining the importance of policy issues is difficult because many voters choose their positions after deciding which candidate they will support. At any rate, policy issues appear to have been more important in the primaries, where they differentiated Trump from other Republicans. They seemed less prominent in the general election campaign. The Clinton campaign focused less on policy than on making the persona of Trump unacceptable. Trump meanwhile relied heavily on generalities like “Make America Great Again” and “Drain the Swamp.” As for some of Trump’s more extreme positions, most were not as widely shared among his voters as pundits presumed. For example, according to the 2016 American National Election Study, only 61 percent of white Trump voters favored building a wall on the border with Mexico, and only 56 percent thought illegal immigrants should be deported. Nearly three-quarters agreed that global warming was happening (at one point Trump had described it as a “hoax”) and less than one- third of those people thought it could be attributed entirely to natural causes.
Polling data suggest that voters were treating some of Trump’s positions more as signals of a change in direction than as specific proposals. For example, an Economist/YouGov survey asked respondents whether Trump would really build a wall. Only 23 percent of Trump voters said he probably would, while 35 percent said he probably wouldn’t (the remainder said “maybe”). Similarly, when asked whether Trump would implement a Muslim ban, only 24 percent of Trump voters said he probably would, while 32 percent said he probably wouldn’t. In September of 2017, when news of a possible bipartisan deal on DACA and border enforcement surfaced, hard-line immigration opponents like Republican Rep. Steve King of Iowa predicted that it would blow up the Republican base. It did not. According to the 2016 American National Election Study, 76 percent of white Trump voters thought that “Dreamers” “should be allowed to live and work in the U.S.” Agreement with Trump’s specific policy proposals was far from the primary explanation for his support. So what was?
One prominent answer is class. After several decades of emphasizing cleavages based on religion, ethnicity, race, gender and sexual orientation, election commentary in 2016 saw a resurgence of social class as a voting divide. During most campaigns political commentary seems to converge on some particular group or demographic as being particularly critical to the outcome. Recent decades have seen the year of the woman (1992), soccer moms (1996), security moms (2002) and waitress moms (2012), the angry white male (1994), office park dads (2002), NASCAR dads (2004) and Joe Six-Pack (2008). In the 2016 campaign the commentariat bestowed pride of place on the white working class, males in particular.
Survey analysts commonly use education as a crude indicator of class. The relationship between educational attainment and voting for Trump is very strong, and election commentary drew attention to the voting difference between the college educated and the non-college educates. But surprisingly, an Economist/YouGov survey indicates that a majority of whites with only a four-year college degree voted for Trump, including a narrow plurality of white female college graduates. Clinton’s support was much stronger among people with advanced degrees and certifications. Measuring class by education puts the Trump/Clinton dividing line between upper middle-class whites and everyone lower, rather than between the middle class and the working class.
Alternatively, in every election since 1952 The American National Election Studies have asked whether people think of themselves as working or middle class. As a proportion of the electorate, white working-class identifiers have declined from about 60 percent in 1960 to 30 percent in 2016. Consistent with election commentary, in 2016 these working-class identifiers stood Karl Marx on his head: For the first time their support for the party of the right exceeded that of the white middle class -- by about nine percentage points. Working-class white women were almost six points more Democratic than men, but that gender gap was smaller than in every election between 1988 and 2004. Working-class women have grown more similar to men in the past three elections. The election of 2016 simply marked the latest observation in a long-term trend. Working-class identifiers have been steadily drifting away from the Democrats since 1964.
Both objective and subjective explanations have been offered as explanations for white working-class defections from the Democrats. Economists point to the economic stagnation experienced by working-class people and areas since the 1970s, stagnation that often leads to cultural disintegration. In contrast, urban elites have done very well economically in the new world order. More subjectively, numerous commentators saw a revolt against the class privilege of some elements of the Democratic Party. Blogger Andrew Sullivan wrote: “Much of the newly energized left has come to see the white working class not as allies but primarily as bigots, misogynists, racists, and homophobes, thereby condemning those often at the near-bottom rung of the economy to the bottom rung of the culture as well. … They [the white working class] smell the condescension and the broad generalizations about them—all of which would be repellent if directed at racial minorities.”
Considerable evidence supports the existence of a divide in the Democratic Party between upper middle-class voters who prioritize lifestyle issues and less fortunate Americans who prioritize “bread and butter” issues. So the argument that 2016 rose to the level of a populist revolt against elites clearly has merit, not only in the United States but in elections in other Western democracies. In my view, however, the early data suggest that a more significant factor in Hillary Clinton’s loss was the candidate herself and the style of campaign she chose to wage. The campaign relied heavily on “analytics,” the use of big data and algorithms to determine personal appearances, ad placements and spending, and reportedly ignored urgent advice from human Democrats on the ground.
Clinton employed many of the same party pros who directed Obama’s data-driven national campaigns, so it stands to reason they’d employ their tried-and-true methods. But more than anyone else, they should have perceived the obvious differences between the two candidates. It was a given that Clinton would not be able to replicate the enthusiasm minority voters felt for Obama, but Clinton managed to hold only two-thirds of Obama’s white vote and did not make up for lower minority support with higher support from white women. The simple fact is that Donald Trump, the most negatively regarded candidate since 1952, had the great good fortune to run against the second most negatively regarded candidate. Gallup data show that Trump obliterated the previous Republican record (Barry Goldwater, 1964) by 16 points, scoring a 63 percent negative rating. Not to be outdone by Trump’s record-shattering performance, Clinton nearly matched it by topping the previous Democratic record (George McGovern, 1972) by 14 points, achieving a 55 percent negative rating.
Data from the exit polls and other surveys clearly show the minimal regard that the white electorate held for both candidates. Although Clinton’s strong supporters believed she was the most qualified candidate ever to seek the presidency, only half the electorate thought she was qualified, only half thought she had the right temperament, and less than a third thought that she was honest and trustworthy or said what she believed. The negativity surrounding Clinton is accentuated by the fact that Americans clearly understood what they would be getting in Donald Trump: Only one-third thought he was qualified, temperamentally suited, and trustworthy. But many such people voted for him anyway (Trump did get relatively high marks for authenticity, where Clinton scored poorly). Half the electorate disliked each candidate and few liked either one “a lot.” Looking at this data, the notion that millions of Americans awoke on Election Day and thought “Finally, a racist/sexist/demagogue I can vote for!” is all the more dubious. More of them woke up wondering whether it was better to vote for “Alien” or “Predator.”
While the consequences of Donald Trump’s presidential victory may be major, even existential, its underlying basis has been exaggerated. Our country is not poised on the abyss of race war, Civil War, or any other kind of domestic war. Nor are we sliding down a slippery slope toward fascism. Normal people -- a.k.a. the general public -- continue to live their lives as they did before the election. The voters changed little between 2012 and 2016, although the small changes that occurred were critical for the outcome. Many of Trump’s voters were not endorsing his draconian proposals so much as sending a message that they were unhappy with the direction of current policy. Finally, many, if not most Americans thought that the parties had given them a historically poor choice. Believing that change was needed, just enough of them in just the right places rolled the dice to send the candidate of the status quo down to defeat.
By no means do such conclusions imply that all is well. In American politics today, majorities are fleeting. For nearly two decades neither party has shown itself capable of governing in a manner that will win long-term support from a majority of the electorate. As astute political writer and presidential biographer Robert Merry has written, “When a man as uncouth and reckless as Trump becomes president by running against the nation’s elites, it’s a strong signal the elites are the problem.”