Along with reviving The “failing” New York Times, Donald Trump can take credit for having launched an entire academic enterprise, Trumpology: trying to understand how and why he rocketed from a reality show celebrity to the White House. It’s been about a year since I summarized studies trying to answer that question and it’s now about a year before Americans revisit their 2016 decision. What does the last year’s research show?
My previous two entries on Trumpology (“Explaining Trump” and “Explaining Trump Some More”) suggested the following tentative conclusions from earlier studies:
* Explaining why Trump actually won the electoral college is not very interesting. The election was close and any number of minor events could have made the difference. We need to understand why such an improbable candidate won the nomination of a major party and, in particular, what motivated his MAGA enthusiasts.
* The key seems to have been cultural anxieties–Trump’s success in addressing and inflaming worries about race and immigration, clearly, but also worries about feminism and other elements of the “culture wars.” Trump’s attention to economic distress was, at best, secondary.
* Once Trump won the nomination, party partisanship–much greater in 2016 than it was a couple of decades ago–ensured strong support from Republicans and strong opposition from Democrats. Thus, the fall campaign was fought over a very narrow no man’s land, where any event (say, Access Hollywood, purloined emails, or an FBI news conference) could make the difference.
The new research I report below is consistent with these conclusions but fills them out, particularly telling us about Trump’s takeover over the Republican party and his nature as a populist. I am sure that there are many more studies out there, but this is a start. I will review what new we have learned about Trump’s capture of the Republican nomination and of the fall election, place Trump the populist in international perspective, consider parallels to a 1960s-’70s precursor of Trump, and close by speculating about 2020.
Taking Over the GOP
Recent work, notably the 2018 book Identity Crisis by political scientists John Sides, Michael Tesler, and Lynn Vavreck, makes Trump’s victory more comprehensible (see also an article of theirs). Trump catalyzed two major concerns that Republican rank-and-file voters had been increasingly voicing for years but that party leaders had insufficiently responded to. One was protecting middle-class economic security by defending Social Security and Medicare and by fending off foreign competition. Trump embraced these goals while Republican leaders emphasized deregulation, tax cuts, and the free market. The other key concern was the rank-and-file’s anxieties about threatening cultural change, mainly the racial and immigration threats they perceived, but also gender and lifestyle worries. Trump pointed to and denounced these seeming threats more explicitly and loudly than other Republican leaders ever had. Doing so, alone, would not have catapulted Trump to the nomination were it not also for his incredible showmanship and the media’s insatiable appetite for his histrionics, as is also documented by Sides, Tesler, and Vavreck.
Other research also reveals how critical the cultural issues, notably race and immigration, were in Trump’s capturing the party:
* One study found that in the early months of the primary campaign, Trump drew especially strong support from Republicans who lived in counties that had experienced high rates of Latino population growth, but he drew that support only after he had made his immigrants-include-rapists speech and he had promised to build a wall at Mexico’s expense.
* Surveys show that Trump primary voters differed from other Republican primary voters in expressing greater racial, anti-immigrant, and anti-Muslim attitudes, and in saying that being white was very important to them, that whites were under threat, and that being a Christian was a key criterion for being an American (see here and here, as well as the Sides et al book).
* A panel study found that anti-black attitudes–that is, believing blacks to be especially lazy and untrustworthy–as measured four years before the primaries predicted which Republicans in 2016 voted for Trump over other Republican candidates.
* Finally, and most abstractly, one economist constructed measures of universalist versus communal moral values–that is, caring about suffering, rights, justice, and equality versus caring about loyalty, national pride, tradition, and authority. He found that “Trump’s moral language [was] less universalist (or, equivalently, more communal) than that of any other presidential nominee in recent history” and more so than the language of his primary opponents. Voters who shared that philosophy were especially likely to favor Trump in the primaries.
Yet, Trump supporters were not different from other Republicans in income and were only a bit less likely to be have been college graduates.
Trump captured the GOP by hammering much harder than other candidates had–then or before–on the sensitive cultural issues, especially but not only race, issues that many Republicans increasingly cared about, while reassuring them that he would still protect the economic interests of the middle and upper working classes (not to be confused with protecting the interests of the poor).
Taking Over the Country
Once the nominations were decided, party polarization determined how almost all Americans would vote. What remained to battle over were the few remaining swing voters and who could better drive turnout. Trump lost nationwide, but he prevailed in a few key states, particularly winning those states’ undecided voters in the last days of the campaign. A few researchers continue to point to economic motives for Trump votes, but the new evidence continues to favor the explanation that cultural, especially racial, issues made the difference.
In support of the economic explanation, one study located the greatest shift from not voting or voting for Obama in 2012 to voting for Trump in 2016 to be among voters with working-class occupations and those living in working-class counties. Another found that counties that had experienced recent increases in job automation shifted a bit toward the Republican candidate in 2016–a bit, but not that much. A survey found that economic attitudes before the election predicted later Trump vote–but not as much as prior cultural attitudes did. From an economic perspective, one might argue that people develop feelings such as racial animus because they suffer from economic anxieties. It is as likely, however, that cultural concerns shape people’s political views, and their politics, in turn, shape their reading of the economy. (Americans rate how well the economy is doing largely in accord with their party affiliations; they reverse their ratings when the White House flips over from one party to another.) Thus, Sides and colleagues write of “racialized economics”: many Americans view economic issues in terms of whether their group–not necessarily themselves personally–is being unfairly disadvantaged. The authors invoke as illustration Arlie Hochschild’s description, in Strangers in Their Own Land, of Louisiana whites’ resentment of blacks and immigrants being allowed, in their view, to “cut in line” for the American Dream.
New studies continue to show that having heightened cultural concerns, not being economically left behind, was what really distinguished Trump enthusiasts. For example, one study I cited earlier about Republican primary voters also found that white voters who favored Trump over Clinton had held distinctively anti-black views (though not distinctively anti-Latino views) a few years earlier. The consensus among researchers is that Trump was special, at least among presidential candidates of recent decades, in his ability to arouse and profit from racial concerns. (Below, I discuss a comparable candidate of nearly a half-century before.) Moreover, Trump could also draw on whites’ feelings about living with a black president for eight years–years which included, among other touchy racial issues, the Black Lives Matter controversy. Ironically, Hillary Clinton, in her intensive effort to get out the black vote, may have played into the Trump strategy. Sides et al quote Clinton (p. 200) as saying that her campaign “likely contributed to heightened racial consciousness” and so “some white voters may have decided I wasn’t on their side.”
I want to again underline that Trump voters in the national campaign were not just concerned about race and immigration; they were concerned about threats to “traditional” social practices more generally:
* A study of Michigan voters found that voting for Trump was connected–even after accounting for racial views and party identification–with agreeing that “Our country is changing too fast, undermining traditional American values” and with disagreeing that “By accepting diverse cultures and lifestyles, our country is steadily improving” (see also here).
* A study of national data found that the 2016 election stood out from other recent presidential elections in how important hostility to feminism was in predicting male voters voting Republican.
* A different national survey revealed that Trump voters were distinctive in expressing strong feelings of American nationalism.
* Yet another showed that Trump voters were especially likely to endorse anti-elite views and believe in conspiracies.
* Respondents in a post-election survey who endorsed “Christian nationalist” positions (that the government should allow prayer in the schools and should “advocate Christian values” and that “the success of the U.S. is part of God’s plan”) were much likelier to say that they had voted for Trump the year before–even after holding constant racial and economic views and many personal attributes.
* Rural farm areas of the country surged in their support of the 2016 Republican candidate over the 2012 candidate, Romney, consistent with the widening cultural distance between urban and rural communities.
* Finally, the study I mentioned before on universalistic versus communal values confirmed that in the fall campaign voters who answered in a communal direction were especially likely, other things being equal, to vote for Trump.
Commentators on the 2016 election quickly noted the widest gap between white college graduates voting Democratic and white non-graduates voting Republican ever recorded. (The unanticipated size of the graduation difference threw off the pre-election polls in the Midwest.) Many quickly concluded that the gap reflected the different economic fortunes of the two groups. However, more careful analyses show that white college graduates preferred Clinton and white non-graduates preferred Trump because the two groups’ social views, especially on race, differed (see Sides et al, among others such as here).
What About Those Switchers?
The trump card, so to speak, for those who hold to an economic explanation is that a small but critical number of fall 2016 Trump voters had previously voted for Obama. How could they be racists? For one, even racists worry about things besides race, especially when, as in 2008, the the banking and housing systems are about to collapse. Many whites who admitted to hostile racial attitudes nonetheless voted for Obama (see Sides et al). Several studies have addressed the puzzle of the switchers and their results point in a familiar direction: Cultural views mattered more than economics in accounting for Obama-to-Trump switching.
A few studies, using panel surveys, followed the same voters over time. They either found that changes in economic concerns or conditions predicted switching a little bit or not at all. Two researchers using the same survey but applying different methods discovered that it was essentially attitudes about race and immigration that explained who switched (see here and here). Another study found that changes in respondents’ views about the value of inter-group equality, changes in a negative direction, modestly predicted who switched.[1] An analysis recently posted on a blog reported that polls years earlier of voters who would eventually become Obama-to-Trump switchers showed them to have resembled Republicans to start with. This finding suggests that the question about switchers perhaps should be not why they switched, but why they voted for Obama in the first place. Finally, research asking what factors predicted which Midwestern counties shifted from Obama to Trump found that the answer had nothing to do with the economic fortunes of counties, but depended on their racial and educational makeups and with how rural they were.[2]
Thus, studies that have focused specifically on the small percentage of Obama-to-Trump voters also conclude that cultural anxieties explain the paradox.
Caveats
Two qualifications before proceeding: First, these conclusions do not deny that many of Trump’s critical supporters–both in the primaries and in the general election–had economic worries and lived in “left behind” communities. But those economic concerns and conditions did not really distinguish them from other voters. Indeed, some data show that Clinton’s voters were actually more worried about their finances than Trump’s were. It was the Trump enthusiasts’ cultural concerns that marks them.
Second, researchers typically interpret the questions in most of the survey studies as measures of respondents’ racial animus, provincialism, or insecurities. From another perspective (e.g., here and here) what these questions really measure is respondents’ willingness to endorse political correctness; Trump voters were distinctive in standing up to liberal pieties. Even were this a better interpretation, the conclusion remains that Trump mobilized his core constituency on cultural grounds.
Trump, Another International Populist?
Many have interpreted Trump’s ascendance as part of a global story, the rise of right-wing populism, often leading to authoritarian rule as, for example, in Hungary, Poland, Turkey, and India, or to near-victories for such movements, as in France, or in the case of Britain, to the Brexit vote to leave the European Union. The common explanation is that people, especially native-born men, have suffered economically because of globalization, technology, and large-scale immigration. Those feeling injured by these developments have rejected the liberal elites whom they see as responsible and rallied to those on the right who promise to make the nation great again. (For such arguments: here, here, and here.)
Trump, as we saw, does not quite fit; economics was not his main message, nor his core voters’ prime mover. What about overseas? I took a quick look at studies of the 2016 British referendum in which those advocating leaving the EU narrowly prevailed over those who wanted to remain. Economic factors mattered. Districts affected by competition from Chinese imports were especially likely to vote Leave. So were districts that had been hard hit by the Tory government’s cutbacks in social spending as part of its austerity response to the Great Recession. (It is ironic that communities hurt by the Tories around 2010 would six years later support the largely Tory Brexit and nine years later give Boris Johnson a big win.) Some argue that the xenophobic language of many Leave voters was the result of economic anxieties: “Trade policy is not, in general, a topic that can easily stir political passions . . . .[T]o be a politically viable strategy, protectionism is therefore often cast in terms of national pride and self-sufficiency” (here).
Yet, cultural issues mattered greatly in the Brexit vote. The feeling that life had worsened in the last generation, which many more Leave voters reported than did Remain voters, surely included an economic dimension, but Leave votes were more closely tied to concerns about immigration and multiculturalism than to worries about the standard of living (Table 1, here). One review found that objections to immigration were not related to voters’ economic situations, but to their cultural worries, while another study found various kinds of both economic as well as cultural objections to immigrants predicted Leave voting. A 2017 study focused particularly on the importance of English identity (the English comprise 84 percent of the United Kingdom’s population). Strongly asserting a specifically English identity for oneself over a British one predicted who would vote Leave, even when other things that also mattered, notably age and political philosophy, were held constant. The authors conclude that “Brexit was made in England because of England’s population [size] . . . . And England’s choice for Brexit was driven disproportionately by those prioritising English national identity.”
This quick look suggests that economic fortunes and worries may have played a more important role in Brexit (and on the continent) in creating among many, especially among older, less-educated men, a resentment that the modern world was pushing them aside. But their cultural objections were probably at least as important. In the U.S., they were more important.
Trump as Populist: An Historical Comparison
Trump seems sui generis. And in oh so many ways, he is. But the combination of economic populism and xenophobic populism that he campaigned on is not that new. I recently read historian Dan T. Carter’s superb biography of George C. Wallace, The Politics of Rage (1995), and was struck by the parallels (as was Carter himself and as were others, e.g., here and here). Wallace, elected governor of Alabama in 1962, stepped forward to lead resistance to integration (“I say segregation now, segregation tomorrow and segregation forever!”) and he parlayed that role into runs for president, most notably in 1968 when he won 14 percent of the popular vote and 46 electoral votes.
Like Trump, Wallace’s national campaigns were remarkable for the enthusiasm of the crowds that over-spilled his rallies, roared amens to his attacks, and eagerly donated small sums to fund his campaigns. His pitch was both racial and populist. Carter points to “gonzo” reporter Hunter Thompson’s account of a 1964 Wallace rally in Milwaukee’s Serb Hall. The place was packed with working-class men, a few of whom told Thompson that George Wallace “ain’t the same as others. He don’t sneak around the bush. He just comes right out and says it” (Thompson, p. 139). What Wallace said and what his audiences roared approval of were “snarling attacks against hippies, civil rights ‘agitators,’ welfare recipients, atheists, beatniks, antiwar protesters, Communists, street toughs who had ‘turned to rape and murder ‘cause they didn’t get enough broccoli when they were little boys.’” “Wallace’s followers reveled in the performance; they never tired of hearing the same lines again and again”(Carter, p. 346).
One distinction between Wallace and Trump, however, is that Wallace did not capture a major party. But he might have. In 1972, the day after he was shot and crippled, Wallace won the Maryland and Michigan primaries; he was then fewer than 20 delegates behind the eventual winner, George McGovern. Another distinction is that, in Wallace’s time, winning a party’s nomination would not have guaranteed that the rank-and-file members would stay loyal in November (as McGovern learned that year).
A deeper difference is that Trump turned out to be a faux populist, at least economically. As governor, Wallace was happy to raise Alabama’s debt for job-creating and other big government programs (and for some “honest graft”). Trump the campaigner promised a large infrastructure project for jobs and attacked Wall Street interests, but Trump the president borrowed heavily to instead fund tax cuts for corporations and the wealthy and he untied regulations on corporations, including Wall Street firms, while letting the infrastructure promise fade away. Where he has remained a populist like Wallace, however, is in fulminating against the “other.”
Looking Ahead
What does knowing all this imply for forecasting or strategizing 2020?
Barring a truly dramatic event (and it’s hard to imagine more drama than the last three years have brought), Trump will be in a strong position. In our party-polarized era, he starts with roughly 90-plus percent of Republicans who will vote for him, effectively no matter what–and with a roughly comparable proportion of Democrats who will vote against him no matter what. Add to that the absence of a large-scale war and the very rosy employment and wage numbers, and any incumbent president should be a shoo-in. Moody’s, the bond-rating firm, uses a prediction model that includes only the economic fundamentals–consumers’ reports on their finances, the stock market, and unemployment–and it recently projected a solid Trump victory (although in 2016, Moody’s had projected a Trump loss). Nonetheless, he is not a shoo-in. The prediction markets (as of December 15, 2019 during the impeachment process) are leaning Democratic (e.g., here and here) and the polls typically show most Democrats, especially Biden, beating Trump nationwide. Trump seems to be a uniquely disliked president, given the positive fundamentals. He may even have to struggle against “Trump fatigue,” weariness about his official acts and also simple weariness about his omnipresence.
Democrats at this writing are largely torn between two strategies: (a) Assume that almost no one can any longer be persuaded to shift from one party to another and, therefore, concentrate on rallying the bluest of the blue groups–blacks and young women–to get out and vote in huge numbers. This is a perennial hope for Democrats, almost always dashed by reality. Or: (b) Assume that, absent an historic candidate like Obama (Clinton wasn’t historic enough), the bluest of the blue are not likely to turn out in sufficient numbers. Focus instead on converting that thin slice of 2016 Trump voters or 2016 non-voters who might be willing to vote blue.
Encouraging for the first strategy is the 2018 turnout among 18-to-29 year-olds. Although still abysmal at 33%, it was 14 points higher than the average for midterm elections since 2000, a notably larger upward spike in 2018 than for Americans over 45. (Black turnout also grew in 2018 over other recent midterms, but only about as much as white turnout did.) Also encouraging for the turnout strategy is simple demographics: Between the 2016 election and the 2020 election about 3 million non-Hispanic white men 65 years old or over will have passed; over 2 million young women of color will have come of voting age. These unstoppable realities could produce a swing of more than a million net votes from red to blue.[3]
Encouraging for the second point of view is the consistent advantage that Biden, hewing to the moderate lane, has had over Warren and Sanders in head-to-head polls against Trump. The strategy of courting the few persuadable whites in the middle might entail showing that Democrats can listen empathetically to heartland Americans’ cultural worries, not about race, but about their communities, families, and faith. The Democrats certainly can provide, within limits, an economic populism to attract those of Trump’s voters who have started doubting his sincerity on that count.
Whatever the calculations, there is a joker in the deck. In 2016, Clinton beat Trump nationwide by about 3 million votes and she still lost. Clinton even could have, had she won California and New York unanimously instead of just heavily, bested Trump by about 11 million votes and, yet, Trump would still be president. Welcome to the electoral college.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
NOTES
[1] This study got some attention in the media and then a sharp critique in an academic journal. I am persuaded by the original author’s reply.
[2] The linked paper is only about Iowa. But an author wrote me to say that further work showed that the conclusions applied to the Midwest more generally.
[3] This is a crude estimate and I welcome a demographer’s more precise calculations. Death rates are here, p. 28, populations: here. The net vote estimate assumes that the old men vote 67% Republican and young women of color 67% Democratic and then discounts for turnout. The estimate might be double if one assumes 80% party voting and large turnouts.
MADE IN AMERICA 