Following Election 2016’s “shocking” finale, many in academic and journalistic circles have seemed less interested in dispassionately analyzing why Trump won than finding excuses for why Hillary lost.
As far as excuses go, sexism or misogyny (like racism, “foreign meddling,” or “fake news”) is pretty effective: it isn’t that Clinton was a non-charismatic candidate with a lot of baggage and a boring platform who ran a bad campaign — instead, those who didn’t vote for Hillary were driven by irrational and immoral impulses, preventing them from embracing the only ‘legitimate’ candidate in this race.
Therefore, it should not surprise that a vast academic literature has emerged on the alleged role of sexism and misogyny in the 2016 U.S. General Election (given that scholars overwhelmingly lean left). Co-occurrence searches on Google Scholar can provide insight into the scale of this enterprise. Restricting our search to 2016 and beyond, “Donald Trump” and “misogyny” yields 1,480 results to date; pairing “Donald Trump” and “sexism” brings in 2,760 hits; “Donald Trump” and “feminist” has 5,080 entries.
There is certainly some overlap between these, but it is nonetheless clear that a large academic corpus is being rapidly produced on this topic – in a wide array of fields, using diverse theoretical and methodological frameworks. However, surveying the titles and abstracts of these works, it is difficult to find any that meaningfully challenge notions that Trump and his supporters were sexist, that Clinton lost in large part because she was a woman (or a “strong woman”), or that gender played an extraordinary role in this election cycle.
Yet there are many reasons to be skeptical of this consensus position.
For instance, much has been made of the “gender gap” between Republicans and Democrats in 2016: according to exit-polls, the distance between Clinton’s margin of victory among women, and Trump’s margin of victory among men, was wider than it had ever been between the parties. But how much of this effect was actually driven by Trump? Consider, the same was true in 2012: the “gender gap” was larger than it ever had been. In this respect, 2016 seems to be a continuation of trends from the previous cycle, rather than a sudden rupture.
Moreover, Clinton’s margin among women (relative to Trump), while solid, was not historic. According to New York Times exit polls, Bill Clinton won women by a bigger margin in 1996, as did Obama in 2008. Al Gore won women by about the same margin as Hillary in 2000. Nor was Trump’s margin among men unprecedented for Republicans: Nixon (’72), Reagan (’80, ’84) and George H.W. Bush (’88) all won the male vote by a larger margin than Trump. The “historic” gap emerged because both candidates had slightly bigger margins than usual among either men or women, not because Trump or Clinton did amazingly well with either group.
However, margins of victory is a non-ideal way of exploring this question because, for many reasons, exit-polls tend to oversample Democratic-friendly constituents (therefore, Democratic margins of victory are probably overstated across the board, and Republican margins of victory, understated). However, we can control for this bias by looking at Trump’s female support relative to his Republican predecessors instead: assuming the quality and bias of a long-running exit-poll is roughly constant across time, longitudinal differences can be held to reflect authentic changes in support among different constituencies.
Among Republicans, Trump won the lowest share of the female vote since 1996. But of course, this does not imply Hillary Clinton did well with women. In fact, she did poorly with women as well. Going back two decades on the Democratic side, the only candidate who got a lower share of women than Hillary was John Kerry.
How can this be explained?
Many analysts have latched onto race: Trump won a majority (53%) of white women. But this, too, is nothing extraordinary. Going all the way back to 1972, Republicans have won the lion’s share of white women in all but two cycles (1992, 1996) – and even in these instances, Bill Clinton could only muster a plurality of the white female vote. Democrats have never won a decisive majority in this demographic in at least the last 40 years. But actually, Trump did equivalent or worse with white women than his immediate predecessors Romney (56%), McCain (53%), and Bush II (55% in 2012).
In short, Clinton’s poor performance with women was not a result of race being especially salient in this cycle. Placed in historical context, Trump’s performance among white women was middling at best for a Republican candidate.
Nor does it seem to be the case that women had “internalized misogyny” and, themselves, couldn’t embrace the idea of a female president: most female Obama voters who defected from the party in 2016 did not go for Trump (again, his performance among women, including white women, was relatively low) – they went instead to Green Party candidate Jill Stein.
Far from being more genteel or amicable than Clinton (i.e. a more “acceptable” female option for those who could not accept a “strong” woman), Stein was more aggressive and subversive — far bolder than Hillary — in her rhetoric, in her manner, in her policy platform, etc. So it does not seem to be that women just couldn’t support one of their own, or were turned off by an assertive and confident woman.
In fact, Hillary didn’t just get one of the lowest female vote shares of any Democrat over the last six elections among those who did turn out, fewer women headed to the polls this cycle overall. According to the U.S. Census Bureau, female participation dropped by 0.4 percentage points in 2016 as compared to 2012 — and 2.5 percentage points compared to 2008.
Clinton ultimately lost largely because of her poor performance with women. Had Hillary won the same shares as Obama, Gore, or even her husband with this constituency, or if she had equivalent (or especially increased) rates of female turnout, she almost certainly would be president today.
Why didn’t she? Again, women didn’t defect to Trump en masse, they didn’t seem have a problem voting for a woman (given that the lost votes gravitated mostly towards Stein). The problem seemed to be Hillary Clinton in particular: her message, her platform, her character. And of course, the same factors that drove so many women away from Clinton likely also depressed her performance with men. Indeed, had Clinton won, she would have been (like Trump is) the least-popular victorious candidate in modern U.S. history.
Given these realities, unsettling questions emerge about how the election has typically been explored in the literature up to now. For instance, why so much focus on men, “threatened masculinity,” and sexism, rather than exploring how women exercised their agency in this election? There is a moral dimension to this question — shouldn’t we be especially concerned with female perspectives, and female agency, in the age of Trump? However, there are theoretical considerations as well.
For one, the story among women seems more analytically interesting: It is truly striking that Clinton performed so poorly (in terms of vote share and turnout) considering her historic status as the first female candidate at the top of a major party ticket, and given the unending media portrayal of her opponent as a sexist, misogynist, serial harasser with a policy agenda that was just as horrible as his rhetoric. The puzzle grows all the more fascinating in light of the fact that Democrats’ vote-share among women has been consistently eroding across most midterm and general elections of the last decade (attrition mostly to third-parties).
The story among women is objectively more important too: Women made up a majority (52%) of the electorate in 2016 – and indeed they’ve represented the majority of voters for every election of the last 30 years. They consistently represent an even larger share of the Democratic base.
Therefore, if one wanted to understand an electoral outcome on the basis of gender, one should start by analyzing and contextualizing the vote preferences of women and how they’ve changed. It is a priority error to focus on men, given they are relatively less significant to determining how most races shake out. Nonetheless, the discussion on gender and 2016 has overwhelmingly focused on the male vote.
As I demonstrate in an article for the forthcoming volume of The American Sociologist, “Race and the Race for the White House: On Social Research in the Age of Trump,” similar peculiarities hold in the burgeoning literature on the role of race and racism in the 2016 election.
For instance, Trump’s victory is often described as a “whitelash” by voters eager to erase the legacy of America’s first black president. However, this narrative fails to take account of basic election data. For instance, the most decisive votes for the 2016 race came from people who had supported Barack Obama in 2012 (and often 2008 as well) but then switched to Trump. If these were people horrified by a black commander-in-chief, it is not clear why they would have voted to give him another four years to pursue his agenda (let alone have voted for him in 2008).
If the election were a referendum on Obama, as a politician or a symbol, one would expect his popularity to have declined over the course of the race — especially given how it ultimately turned out. Instead, Obama grew more popular throughout 2016, even as favorability for Trump and Hillary tanked. Two years into the Trump administration, Barack’s ratings continue to climb, with 66% of Americans offering a favorable opinion of him.
The “whitelash” theory also suggests a surge white voting. Instead, participation among non-Hispanic whites was stagnant relative to 2012, and down from 2008. In fact, whites made up a smaller share of the electorate in 2016, while Hispanics and Asians made up a larger percentage of overall voters.
More damning: Trump actually won a smaller share of the white vote than Mitt Romney. He was nonetheless able to win because he won more Hispanics and Asians than his predecessors, and more black votes than any Republican since 2004.
As with gender, turnout was low among core Democratic racial constituencies, particularly African Americans. Had Clinton been able to better mobilize African Americans to the polls, or had she even just maintained Obama’s vote share among blacks, Hispanics or Asians (even from 2012, let alone 2008), she likely would have won. In other words, the problem wasn’t that Donald had extraordinary support among whites (he didn’t) – but instead that Clinton was significantly less popular than Obama among minority groups.
In fact, contrary to predominant narratives about the election, Obama significantly outperformed Clinton with whites as well. Hillary got the lowest share of white voters of any Democrat since 1984. But again, most of these votes did not go to Trump (who did worse than Romney among whites, and about the same as George W. Bush in 2004). Instead, as with Democrats’ lost female votes, many whites (especially young people) opted to vote for third-parties instead.
Nonetheless, scholars bend over backwards trying to find ways to “prove” that Trump voters were especially racist or sexist. Such narratives may be edifying for those who count themselves among the “resistance” — however, the real-world costs of politicized research likely outweigh these emotional benefits:
For instance, scholars and journalists can alienate the very voters Democrats will need to recapture in 2020 — calling them racist and sexist, often on weak evidentiary grounds. However, even to the extent these narratives were true, they could not be meaningfully operationalized: there’s little one can do about voters’ sexism and racism other than play to it, or not. Presumably, Democrats have settled on “not.” Consequently, there seems to be little practical value in trying to “prove” Trump voters are racist and sexist. Indeed, these efforts distract from learning lessons that could actually help Democrats prevent another humiliating and costly defeat in 2020.
For those of us who would like to avoid another four years of Trump rather than circulate comforting stories to explain away his continued victories, it may be necessary to fundamentally rethink how we study the President and his supporters.
Musa al-Gharbi is a Paul F. Lazarsfeld Fellow in Sociology at Columbia University and a research associate with Heterodox Academy. Readers can connect with his other work via his website, and follow him on Twitter @Musa_alGharbi.