Identity Crisis: The 2016 Presidential Campaign and the Battle for the Meaning of America
John Sides, Michael Tesler, and Lynn Vavreck
Princeton University Press
Hardcover, 352 pages, $29.95
October 2018
This book is an academic work of political science from three accomplished scholars, published by a top-flight university press. But its authors know how to grab you, right from the first line:
“Rakeem Jones didn’t see the punch coming.”
The book begins with the story of an African-American man who, while being escorted by police out of a Trump rally in March 2016, was assaulted—“suckerpunched”, to quote the authors—by a white man named John McGraw. McGraw, you may remember, actually said on camera that Jones was “not acting like an American,” and added “the next time we see him, we might have to kill him.”
The authors remind us that the video of the assault went viral not just because of the punch, but because Trump reacted to it by wistfully pointing out that “in the good old days this doesn’t happen because they used to treat them very, very rough.” He also offered to pay any legal expenses McGraw might accrue.
After starting with this quite representative event, and then citing other issues that were central to Trump’s rise, such as birtherism, immigration, Black Lives Matter, and Muslims and terrorism, the authors’ made their overarching argument about the 2016 election, namely that it was about identity—what it means to be American, and what America itself means. While identity always matters in elections, the authors argued that group identities—and how some groups of Americans felt about other groups—mattered more in 2016 than in any other recent election.
Relying on careful analyses of survey data, the authors demonstrate that, during the Republican primaries, voting for Trump over any of the other sixteen candidates correlated significantly to feelings those voters expressed about immigrants, Muslims, and African Americans. Particularly relevant was whether voters believed racial disparities resulted from unequal treatment and discrimination or from failings on the part of black Americans, i.e., their supposedly poor “work ethic”, as well as voters’ perception of the level of discrimination white Americans experience. This correlation did not exist in the Republican primaries four and eight years earlier.
In the general election, these feelings determined which way people voted more strongly than either of the times Barack Obama was on the ballot. The authors cited as causes Trump employing a core campaign strategy that consisted of “focusing on racially charged issues,” along with Hillary Clinton’s decision to engage directly by making her campaign “a direct rebuke of Trump” on those issues. In summary, “the election turned on the group identities that the candidates had activated.” This “activation” redounded to Trump’s benefit on Election Day.
On the matter of economic anxiety, the authors note that, overall, it had been decreasing over the course of the Obama presidency, and was largely flat in 2016 compared to the years just before. However, the authors emphasized the connection between the political and economic anxieties that did exist and the political identities of individual voters. In other words, “’racial anxiety’ was arguably driving economic anxiety.”
The first chapter examines in detail how partisan identification drove voters’ views of the economy, as well as of their own personal economic situation. One poll found that Democrats making less than $20,000 per year were more satisfied than Republicans making $100,000 and up. This chapter also looked at the increasing connection over time between racial identity, racial attitudes, and partisan identity, with a particular focus on the attitudes and identity of whites without a college degree. The authors explained that it was not the case that white voters who opposed Obama changed their racial attitudes to fit their partisanship. Instead, it was the other way around, as reinterviewing the same people over time showed that a significant number of racially resentful whites moved from the Democratic to the Republican party from 2008 to 2016.
Furthermore, racial attitudes became more aligned with partisan identity over the course of Obama’s two terms. Trump further exacerbated this, thus he found his greatest support among whites with “racially inflected grievances.” In a key passage, the authors boiled their argument down to this:
Economic anxieties came to matter more when they were refracted through social identities. The important sentiment underlying Trump’s support was not ‘I might lose my job’ but, in essence, ‘People in my group are losing jobs to that other group.’ Instead of a pure economic anxiety, what matters was racialized economics.
The reason why the authors talk about an “Identity Crisis” is because they believe that when debates on policy are suffused with questions of ethnic, racial, and national identity, those debates become more emotional and thus have more of a destabilizing effect on our politics, making them “divisive and explosive.” That’s certainly a reasonable summary, I’d say, of Trump’s effect.
In addition to the discussion of identity and voting patterns, the book also covers the main events of the campaign. There is a multi-chapter, thorough discussion of the Republican primaries—including the failure of the so-called invisible primary to produce a strong establishment front-runner, as well as the impact of the sheer volume of media coverage Trump received.
On the latter, here’s one especially mind-blowing statistic: during the 365 days starting with May 1, 2015, Trump received more mentions on cable news than all the other sixteen Republican presidential candidates combined. He wasn’t even a candidate for the first month and a half of that period. The authors noted that during the 2012 campaign, Mitt Romney—in a far smaller field and in a race where he was the presumptive front-runner the entire time—got only 30% of the cable news mentions, compared to Trump’s 54% in the comparable months four years later.
The authors dug deep into the key question of the Republican primary campaign, which they defined as “which voters ultimately came to support [Trump], and why?” In summary, they argued that Trump did not change the minds of Republican voters, or persuade people that his ideas were the right ones, but rather he “activated long-standing sentiments among Republican voters...Trump simply went where many Republican voters were….He capitalized on an existing reservoir of discontent about a changing American society and culture.” The most important of these sentiments, according to the data, were how those voters felt about particular groups (blacks, Muslims, and immigrants), and particular issues that directly related to race or ethnicity such as immigration or Black Lives Matter.
I don’t want to go too much into methodology, but I will summarize briefly why the survey data these authors used is so valuable. There were about 8000 people interviewed back in 2011-12, thus at a time long before Trump began his campaign and at a point where his views had a relatively marginal influence on public opinion. Then, in July 2016, the survey re-interviewed the respondents to find out whether they had voted for Trump, Ted Cruz, John Kasich or Marco Rubio. This allowed scholars to compare the beliefs and attitudes of Trump voters compared to non-Trump Republicans in a way that eliminates the question of whether Trump changed or shaped those attitudes during the campaign.
The data demonstrated that Republican primary voter support for Trump connected most directly to “white grievances; views of immigration, Muslims, and blacks, and liberal views about economic issues.” The latter refers to Trump’s statements on the campaign trail regarding increasing taxes on the rich, government-provided health care, etc.—positions he has clearly abandoned since taking office in favor of orthodox conservative ones. Other issues, including trade, resentment for the establishment, personal “economic anxiety”, and terrorism, do not appear to have helped Trump in the Republican primary campaign.
Regarding the Democratic contest between Hillary Clinton and Bernie Sanders, the authors measured what types of attitudinal differences separated those who voted for each one. Interestingly, there were relatively small divides on policy issues, ideological issues, gender, and even attitudes about gender, such as feminism. Female Democratic primary voters who identified as feminists gave Clinton the same percentage of their votes as did those who did not identify as feminists.
Surprisingly, there were no significant differences in terms of economic policy positions—even trade—taken by Clinton and Sanders voters, although those who had expressed pessimistic views on the economy back in 2011 were more likely to end up voting for Sanders than for Clinton. In other words, it’s not that Sanders’ voters became convinced to be more pessimistic, it’s that he attracted those who were already more pessimistic. Whereas Trump did no better in his primary among voters expressing economic anxiety than among those who did not, Sanders did.
The key divides separating Clinton and Sanders voters were based on the following three things: the degree of loyalty to the Democratic party they expressed, their age, and their race, with Sanders voters being younger, more likely to be independents, and whiter. One fascinating and related point the authors made is that Clinton’s 2016 voters were “the racial inverse” of her 2008 voters in the Democratic primary, a year when she not only did better than her opponent (Barack Obama) among white voters, but in particular among those who expressed negative sentiments about African Americans. In 2016, Clinton and Sanders voters showed no difference in their attitudes about blacks, immigrants, and Muslims.
The authors also discussed the overwhelming and unprecedented (for a non-incumbent) advantage Clinton had over Sanders in terms of endorsements from party officials and office-holders, as well as the disadvantage she had in terms of media coverage—a contrast the authors attributed in part to the “steady drip of revelations” regarding the email scandal, as well as the matter of the Clinton Foundation. One other interesting fact: almost every single ad put on the air by Clinton or Sanders during the primaries was positive. The exact figure is 99.75%. They literally almost never attacked one another through advertising.
Finally, regarding the ‘myth’ of Sanders voters not supporting Clinton, 79% of those who voted for Sanders in the primary voted for Clinton in the general. This is a good deal higher than the 70% of 2008 Clinton supporters who voted for Obama in the general. Of those Sanders-Trump voters, a paltry 35% said they had voted for Obama in 2012. In other words, only about one-third of them were actually solid Democrats.
As for the general election, it was at bottom a clash between two different visions of American identity. “Stronger Together” was Clinton’s pitch, and it contrasted her inclusive, egalitarian vision with Trump’s, one borne out by his rhetoric about women and Americans of color. The authors note the lack of attention Clinton paid to economic issues—with only roughly 8% of her television ads touching on them, compared to Trump’s much stronger focus on attacking Democrats over the economy—34% of his TV ads addressed economic issues.
On the other hand, 66% of Clinton’s ads attacked Trump personally—his rhetoric, his statements, etc. Trump focused on issues like crime, terrorism, and immigration—all of which had a racial element—and which connected to his vision of change for America. The authors don’t specifically assert that this was a strategic error on Clinton’s part, but they did argue that her choice to focus on Trump’s rhetoric was one reason why group identities were so powerfully activated, a situation that they argue benefited Trump. Ultimately, the campaign came down to one question: “whether the country’s increasing ethnic, racial, and religious diversity was a strength or a threat.”
The book contains a long discussion of the media, and of the various scandals and controversies that surrounded both candidates. Suffice it to say that the authors argue that ultimately Clinton was hurt by these more than Trump, although quantifying the impact is not possible. They noted that, by the end of the campaign, voters saw Clinton as less honest than Trump, something that was “even more striking because Trump was so prone to dishonesty himself.” Indeed.
The authors make an interesting point about the general wisdom that Trump won because voters desired “change.” In fact, the fundamentals of a generally sound economy and President Obama’s strong approval ratings predicted a modest popular vote victory for Clinton—which is what happened. Additionally, they pointed out that it’s “hard to square” the notion of this being a change election with the fact the Democrat got more votes—in this case by 3 million—than the Republican for the third consecutive time.
The key to his victory was Trump’s strength not among whites across the board, but among whites without a college degree, voters who happened to be disproportionately larger parts of the population of some key states such as Michigan, Ohio, Wisconsin, and Pennsylvania. Especially in the battleground states, a significant chunk of Obama 2012 voters expressed views on racial issues that were “arguably closer to Trump’s” than Obama’s own, or those of Clinton. Clinton improved on Obama’s 2012 margin among college-educated whites by a gaudy 10 points. Unfortunately for her, she did 14 points worse than Obama among whites who lacked a college degree. The latter group represented 42% of 2016 voters, compared to 31% for the former. As the authors noted:
‘The ‘diploma divide’ gave Trump votes exactly where he needed them . . . . No other factor appeared as distinctively powerful in 2016, compared to prior elections, as attitudes about racial issues and immigration, and no other factor explained the diploma divide among whites as fully . . . . They were, unsurprisingly, the factor most strongly activated by a racialized campaign.
The authors went into detail about the connection between racial attitudes and economic attitudes, describing what they call “racialized economics: the belief that undeserving groups are getting ahead while your group is left behind.” The data made clear that Trump voters’ views reflected this racialized economics while those of Clinton voters did not. In one telling data point, while two-thirds of Trump supporters believe that “average Americans have gotten less than they deserve,” the number who agreed when the word “blacks” was substituted for “average Americans” was only 12%. Half the respondents got the prompt with the word “blacks” and the other half with “average Americans.” Clinton voters were equally likely to agree no matter which prompt they heard.
Overall, the data showed that the two sets of voters were not divided on the widely agreed upon idea that “average Americans are being left behind. Rather, the divide was whether a racial minority deserved help.” For white voters, their views of economic issues were filtered through their “racialized perceptions of economic deservingness,” which connected to their sense of “white grievances.”
The authors also looked at black turnout and noted that the drop—5 points nationally and even more in some competitive states, such as Michigan and Wisconsin where it dropped by more than 12 points—did hurt Clinton. Furthermore, the data showed that more restrictive voter ID laws had “modest, if any effects.” For example, black turnout fell in places that enacted such laws, but not equally across the board. Some such states saw lower reductions in black turnout than the national average. Additionally, places like Michigan and Washington, DC, where no such laws were enacted, also saw reductions in black turnout. The better explanation, the authors argued, was simply that African American voters felt more positively about Obama than Clinton. For example, in a 2016 survey, Clinton’s rating was 15% unfavorable and 8% strongly unfavorable, while Obama’s was 4% unfavorable and 3% strongly unfavorable.
As for Latino turnout, while it is harder to measure, the data shows that she did not do any better than Obama, “and she may have done worse.” Although Latinos felt less favorably toward Trump than they had toward Romney, they also felt less favorably toward Clinton than Obama four years earlier.
As for gender as a force bringing voters to the polls, “women did not rally to Clinton’s candidacy, but men shifted to Trump—especially men with more sexist attitudes.” The data showed that the relationship between sexism and voting existed among white men but not other men—more sexist white men voted for Trump than Romney, and more non-sexist ones voted for Clinton than Obama, but the shift was more profound among the former. Trump’s gender advantage among men was the biggest since the 1988 Republican victory, a decisive one, and his advantage among white men was the biggest since Reagan thumped Mondale in 1984. Believe it or not, Clinton did worse among white women than Al Gore sixteen years earlier.
After addressing a number of other possible topics that could have impacted the outcome, from campaign tactics on TV and social media advertising, to field operations, to Russian interference, the FBI investigations, and the Comey letter, the authors returned to their prime finding: namely, that voters’ polarized views on racial injustice, Muslims, and immigration defined the election, and reflect the identity crisis in America.
Since the 2016 election, a lot of people, including myself, have done a lot of pontificating about how the result came about, about how Donald Trump became president. Sides, Tesler, and Vavreck have dug deep into the data, both surveys they’ve conducted as well as surveys going back years, and have given us what look like definitive answers to that question. More than that, they’ve interpreted the data and integrated their understanding of the academic literature on race and politics into an analysis that’s both highly accessible yet incredibly substantive. As an academic who has read too many books that fail to do so, I can tell you how rare that is.
Unless someone else unearths some contradictory data, this book is going to remain the definitive explanation of what motivated and differentiated voters from one another in both primary campaigns and the general election in 2016. Anyone—academic, student, or general reader—who is interested in the topic, in other words, anyone interested in politics, race, or the state of our country, absolutely needs to read this book. Moving forward, so does anyone interested in learning what actually happened in 2016 in order to apply those lessons to 2018, 2020, and so on.
Going above and beyond, one of the authors of Identity Crisis, John Sides, has generously agreed to answer a few questions. Five to be exact, as fits with Daily Kos practice. Our interview appears below:
1) We’ve recently learned that a “mini-recession” in 2015 and early 2016 affected the U.S. economy, although it was
concentrated specifically in the energy, agricultural, and manufacturing sectors. The downturn, according to Neil Irwin of the
New York Times, “helps explain some of the economic discontent evident in manufacturing-heavy areas during the 2016 elections.”
How does this fit with your research on the impact of economic anxiety on the 2016 election in the battleground states you mentioned (WI, PA, MI, OH)?
Our data come from national surveys and cannot definitively identify the impact of any localized economic effects. That said, our data do not suggest that economic anxiety was a central factor driving voters’ choices or the choices of “working class” whites in particular. To know that any localized “mini-recessions” really mattered, it would be imperative to show that this really registered in the attitudes of voters and was actionable in their decisions. I think there’s reason to be skeptical, though, based on our analysis.
2) There’s been a debate among some Democrats (perhaps fewer than it might seem from media accounts) arguing over whether Trump won because of economic anxiety OR racial anxiety. What would you say to those who see that as an either/or question?
If we think of the election as a “competition” between these two factors, then our book finds that racial anxiety was the bigger factor. But we also argue that the competition framing isn’t quite right. In particular, it fails to capture the ways in which economic anxiety takes on a racialized character. We find that the potent sentiment in 2016 wasn’t “I might lose my job” but “Whites are losing jobs because employers are giving them to minorities.” We call this “racialized economics” and it’s connected to a great deal of evidence that people think about politics in terms of groups and their deservingness.
3) What does the 2016 Democratic primary contest tell you about how the 2020 primary will play out?
I’d hesitate to draw strong conclusions. The 2020 primary will likely be much more wide open than the Clinton-Sanders contest. The main lesson from 2016 was that people routinely overestimated the ideological cleavages in the party and the extent to which ideological differences distinguished Clinton and
Trump Sanders voters. I imagine that there won’t be a lot of daylight among the 2020 Democratic contenders in their views on policy.
Another lesson from 2016 was the importance of minority voters, who provided the backbone of Clinton’s support. The Democratic party is now nearly 50-50 white vs. non-white. A successful nominee is going to have to build a multi-racial coalition.
4) Similarly, what does the data on the 2016 general election tell you about 2020, assuming Trump is on the ballot again?
The 2016 popular vote was pretty much in line with economic and political fundamentals, which predicted a close race. So there’s every reason to think that those fundamentals will matter in 2020. The good news for Trump is the country’s solid economic growth. The bad news is that he is less popular than he should be given the growing economy. Given those fundamentals, his election is hardly assured, but neither is he clearly the underdog.
Trump’s victory was also predicated on running against an opponent nearly as unpopular as he was. If his approval rating doesn’t improve and the Democrats nominate someone more popular than Clinton was, that’s also a challenge for Trump.
5) I know you're not a political consultant, but I'll ask this anyway: based on this research, what advice would you give to a Democratic nominee? Trump's approach seems to be already pretty set in stone, so I won't ask what advice you'd give him.
A lot depends on what happens between now and then. If the economy were to slip into a recession, then the Democratic message is obvious — they should point that out! But if the economy continues to grow, Democrats need a different message. They shouldn’t remind voters about the growing economy, obviously! And they need a message that might counter Trump’s more exclusionary rhetoric about immigration and related issues. Something like this:
“You don’t need to worry about ‘your place in line’ — no one is going to cut in front of you and get benefits you’ve worked hard to secure. You place in line is safe. We can ensure everyone gets what they need, without taking things away from people who have worked hard, been patient, and might now need a little help. America can do it all.”