Commentators have offered a deluge of competing explanations for President Trump’s unexpected victory in the GOP primaries and later in the 2016 general election. Fortunately for political opinion writers, there are many plausible ways to explain his wins, and it was easy to choose an interpretation that matched any writer’s ideological or intellectual preconceptions.
Full disclosure, your humble reviewer was guilty of this.
Too often those (still!) discussing Election 2016 are more interested in settling scores than carefully analyzing what happened. Were misogynistic “Bernie bros” to blame for Hillary Clinton’s loss? Was it instead her tone-deaf campaign? Has left-wing identity politics gone too far? Did Russian trolls swing the election? Were Trump voters really motivated by economic anxiety, or was it really all about race from the beginning? Our personal answers to these questions are often determined by our underlying assumptions.
With Identity Crisis: The 2016 Presidential Campaign and the Battle for the Meaning of America, political scientists John Sides, Michael Tesler, and Lynn Vavreck provide welcome insights into these subjects. They carefully examined multiple sources of data, considering the plausibility of various explanations for the election’s results. If they have a thumb on the scale, promoting an ideological agenda, I haven’t discerned it. This is the best, most dispassionate analysis of 2016 that I have seen.
Leaked Clinton Emails Didn’t Hurt Her As a Candidate
Given their clear-eyed account, it is unsurprising that the authors do not try to claim that a single variable determined the outcome in 2016.. However, some explanations have less credibility than others do. They are particularly dismissive of the idea that Russian interference swayed the election. Compared to the amount of political content produced daily on social media, Russia’s online presence was paltry. There is furthermore little evidence that such content sways anyone’s votes. They also found no discernible evidence that hacked Democratic National Committee emails influenced Clinton’s favorability/unfavorability ratings.
The claim that Clinton ran a particularly bad campaign also fails to survive scrutiny. Experimental tests showed that her campaign advertisements were generally effective, and they similarly showed that her vote share was higher in places where she purchased more advertising. That said, as political scientists have long known, political advertising has only a minor effect on election outcomes, and a massive increase or geographical shift in advertising would probably not have swayed the election in Clinton’s favor. It is true that Clinton had a smaller field operation than President Obama had in 2012, but the authors found no evidence suggesting a larger campaign staff would have made a discernible difference—again because of the minimal effects that field offices have on presidential vote tallies.
Clinton’s campaign tactics in the general election were not markedly different from those of President Obama’s campaigns. Yet pundits heralded Obama’s team as political masterminds and condemned the Clinton campaign as “arrogant.” Of course, if a handful of states had voted slightly differently, Clinton’s staff would today enjoy an equally stellar reputation.
The role economics played in the election results was ambiguous. It is true that Trump performed well in the primaries among Republicans who were not free-market purists. There really is a constituency for populist economic policies within the GOP electorate. Many Republicans favor tax increases on the wealthy, economic protectionism, and high spending levels on Social Security and Medicare. This was obvious to anyone familiar with survey data on public policy, but it was apparently a shock to the conservative intelligentsia.
Yet economics was less important than many observers believed (or hoped). The claim that rising economic insecurity explains Trump’s rise, while plausible, has little empirical support. The authors found no strong relationship between favoring Trump and measures of economic dissatisfaction and anxiety. Furthermore, for all the talk about “anger” being the dominant emotion in the 2016 election, Americans were mostly not angry about the economy. The last years of the Obama administration were characterized by increasing economic optimism.
There is a growing disconnect between economic indicators and approval of the occupant of the White House. Trump supporters have been justifiably frustrated that low unemployment and strong economic growth have not led to greater enthusiasm for the President. However, the authors note that this trend began under President Obama. From Richard Nixon to George W. Bush, stronger consumer sentiment was associated with higher approval numbers for the President. Obama broke this pattern; in fact, there was a negative correlation between these variables during his time in office. That being the case, it is unsurprising that relatively strong economic numbers did little to assist candidate Clinton.
Sorting Out the Racial Factor
After demonstrating why other factors had, at most, a modest influence on the election result, Sides, Tesler, and Vavreck conclude that racial attitudes were a key predictor of vote choice. I hasten to add, they carefully note that the relationship of such attitudes to electoral behavior was complicated—the election was not merely a story of white racial animus, or Trump normalizing racist and nativist language.
They also note that, contrary to conventional wisdom among liberal commentators, there is little evidence that Trump increased racism in the electorate. In fact, polling shows that feelings of prejudice among whites have decreased since 2016. This change has only occurred among white Democrats, however—white Republican attitudes have not changed very much in either direction.
Although Trump’s rhetoric did not stir up white anxiety or feelings of racial identity and resentment, he did make these politically salient. That is, there was a weaker correlation between racial attitudes and vote choice in previous elections, including 2008 and 2012, when Barack Obama was the Democratic nominee. Because race was a central element of Election 2016, apparently, racial attitudes were a more important predictor of vote choice in both the GOP primary and in the general election.
The chapter on the Republican primaries emphasized the unusual nature of Trump’s campaign talking points, and how they served to activate feelings of white identity and anxiety. In recent presidential election years, the leading Republican candidates vying for their party’s nomination were mostly indistinguishable on questions related to race; all promoted a formally color-blind conservatism. As the authors note, 2016 was different: “Few Republican candidates for president have attempted to distinguish themselves from their Republican rivals on issues connected to race and ethnicity—until Trump did exactly that.” The authors’ use of longitudinal survey data was helpful, as they were able to examine responses from subjects surveyed in both 2011 through 2012 and 2016. These data were particularly useful for understanding those voters who voted for Obama in 2012 and Trump in 2016.
Those who wish to downplay the role of race and identity in the 2016 election note correctly that a non-negligible number of white voters who supported Trump previously voted for President Obama. As it seems unlikely that people who had voted for America’s first black President had high levels of racial prejudice, this fact increased the plausibility that economics or some other factor was more important in explaining their subsequent support for Trump.
The authors note, once again, that the issue is a bit more complex than that. Not all white Obama voters were racial egalitarians in 2012. Significant minorities of these voters expressed feelings of racial resentment and had very negative attitudes toward immigrants. They voted for Obama, however, because their racial attitudes were not very salient in 2012 – they voted based on their partisan identities, economic considerations, or some other factor. Election 2016 was different because race was a dominant implicit and often explicit theme in the campaigns, and a topic endlessly discussed in media coverage of the election. Thus, racial attitudes were an unusually strong predictor of vote choice.
Have Tesler’s Views Evolved?
It would have been useful if the authors had included more discussion of coauthor Tesler’s previous book, published in April 2016, Post-Racial or Most Racial? Race and Politics in the Obama Era. Tesler there made a strong case that race had massive importance during Obama’s 2008 campaign and subsequent presidency. In some ways, its message seems at odds with that of Identity Crisis, which suggests that 2016 was a dramatic departure from previous patterns. I would have liked to know how, if at all, Tesler’s views on the Obama era have evolved over the last two years.
The authors conclude that politicians may have a limited ability to shape things like racial attitudes in the electorate, but they can indeed influence the political importance of feelings of racial identity and prejudice. Identity Crisis suggests that the decisions made by political elites are largely responsible for the high levels of racial and partisan polarization, and different choices in the future could heal our divides.
This might be correct, but I question the likelihood that politicians and other political leaders will reduce their heated rhetoric any time soon. They have few incentives to do so – as the authors acknowledge. After all, unlike his most recent GOP predecessors, who ran more traditional conservative campaigns, Trump actually won the election, collecting Electoral College votes previously thought out of reach for Republicans. Why would the GOP abandon a successful strategy? Similarly, a Democratic candidate seeking conciliation and moderation would also have little chance in the next round of presidential primaries.
For now, both parties have electoral incentives to maintain our current divisions. I suspect polarizing campaigns will be the norm for the foreseeable future.