The Politics of Losing

Home > Other > The Politics of Losing > Page 10
The Politics of Losing Page 10

by Rory McVeigh


  FIGURE 5.3 Education and primary support for Trump.

  Source: Authors’ analysis. See appendix for more on data sources and methods of statistical analysis.

  FIGURE 5.4 Unemployment and primary support for Trump by percent college educated.

  Source: Authors’ analysis. See appendix for more on data sources and methods of statistical analysis.

  Remember median household income had no relation to voting for Trump even though Trump, more than other Republican candidates, appealed directly to poor and working-class white voters. Here again we see that the relationship between median income and the Trump vote depended on the percentage of college graduates in the county. Median income, in fact, had the expected strong negative effect on the Trump vote, but only in counties with relatively few college graduates. Figure 5.5 shows that where only 10 percent of county residents graduated college, the vote for Trump was high in counties with low income—but the predicted vote for Trump dropped quickly as median income increased. Yet like unemployment, median income was unrelated to the Trump vote where high percentages of county residents were college graduates. Again, these findings indicate that Trump distinguished himself from Republican competitors in communities that were not only economically stagnant but also had few residents well situated to benefit from the postrecession recovery.

  FIGURE 5.5 Median income and primary support for Trump by percent college educated.

  Source: Authors’ analysis. See appendix for more on data sources and methods of statistical analysis.

  We also argued that Trump’s unusually explicit misogyny did not concern his core supporters, and this was because of links between family structure and the global economy. A traditional local economy goes hand-in-hand with a traditional family structure, in which men are primary breadwinners and women are either homemakers or work in low-paying supplemental occupations, like working part time at the supermarket. Trump did well in counties where relatively few women worked. Figure 5.6 shows that this relationship was particularly strong in counties with relatively few college graduates. The same pattern keeps appearing. In this case, Trump did well in counties where only 10 percent of adult residents held a college degree and the percent of women in the labor force was low. And again, the percent of women who work was unrelated to Trump voting in counties where 30 percent of residents held a college degree. These findings once again underscore how Trump’s proposed solutions provoked very different responses, depending on whether there was a highly educated workforce in the county.

  FIGURE 5.6 Female labor participation and primary support for Trump by percent college educated.

  Source: Authors’ analysis. See appendix for more on data sources and methods of statistical analysis.

  At first it seemed that there was no strong correlation between voting for Trump and marriage rates. Sociologist David Autor outlined how the economic deprivation of those who have not benefited from globalization has depressed marriage rates, since men’s reduced earning power, along with attendant behavioral issues, has left them ill-prepared for long-term relationships.22

  The graph in figure 5.7 suggests that marriage rates did matter. Trump did well in counties where marriage rates were low and education levels were low. Did Trump’s rhetoric and behavior appeal to men in places where disconnected economies and too few college degrees undermined male dominance in the household? These same conditions could lead to support from women who regretted the dearth of suitable traditional marriage partners in their communities and who took hope from Trump’s promises to bring jobs to their communities. In these places, those voting Republican because they wanted to protect male privilege rather than class privilege could now align both interests within their support for Trump.

  In the United States, conservative evangelical Protestants have aligned themselves with the Republican Party, in part, because of the party’s conservative positions on abortion, contraception, and LGBTQ issues, including same-sex marriage. Given that Trump’s opponents devoted significant energy to courting evangelicals, and given Trump’s own lifestyle and prior liberal views on abortion and same-sex marriage, why did he fare so well among evangelicals? While he did switch to an antiabortion position when he ran for president and promised to appoint conservative justices to the Supreme Court who would overturn Roe v. Wade, this didn’t distinguish him from any of his Republican opponents.

  FIGURE 5.7 Percent married and primary support for Trump by percent college educated.

  Source: Authors’ analysis. See appendix for more on data sources and methods of statistical analysis.

  Previously in the chapter we showed that, controlling for other county attributes, Trump gained more support in deeply evangelical counties. This finding is clearer when we consider how evangelical voting differed, depending on the percent of county residents with a college degree. Once again, we see in figure 5.8 that the educational divide not only predicts the Trump vote but shapes how other county attributes affected the vote. The proportion of evangelicals in a county strongly predicted Trump voting, but only in counties with relatively few college graduates. In such counties, Trump made it easier for evangelicals to align their religious beliefs and their economic grievances with a vote for a single candidate.

  FIGURE 5.8 Percent evangelical and primary support for Trump by percent college educated.

  Source: Authors’ analysis. See appendix for more on data sources and methods of statistical analysis.

  Finally, we return to the issue of race. In October 2017, Ta-Nehisi Coates, writing in the Atlantic, called Trump “the first white president,” by which he meant that Trump was unusual in how he explicitly presented himself as the representative of white America: “It is often said that Trump has no real ideology, which is not true—his ideology is white supremacy, in all its truculent and sanctimonious power.”23 Was Trump a candidate by and for white Americans? We see that the percent of nonwhite county residents had no relation to the Trump vote in the primaries and caucuses. Keep in mind that the vast majority of Republican primary and caucus voters were white. Pew data indicate that 86 percent of registered Republicans in 2016 were white.24 We include the race variable, therefore, to ask whether the racial composition of the county was relevant to white Republican voters when choosing among Republican candidates. The percent of nonwhite county residents increased Trump voting in counties where a relatively high proportion of those in the labor force worked in manufacturing (see figure 5.9). This may indicate that Trump’s brand of racial politics was particularly appealing to white Republican voters who found themselves in racial competition for manufacturing jobs, as opposed to primarily competing with foreign labor. Keep in mind that this only reflects the votes of Republican primary voters, who are predominantly white. Their support for Trump was especially high in counties with large nonwhite populations and with high proportions of manufacturing jobs. As we see in the next section, race was a strong factor in determining the outcome of the general election.

  FIGURE 5.9 Percent nonwhite population and primary support for Trump by percent employed in manufacturing.

  Source: Authors’ analysis. See appendix for more on data sources and methods of statistical analysis.

  UNSTABLE ALLIANCES IN THE GENERAL ELECTION

  While Trump’s rise within the Republican Party was a serious concern for many orthodox Republicans, his victory in the primary campaign confronted regular Republicans with a dilemma. They could support Trump even though his temperament and qualifications could prove disastrous for the party and even though he might take the party in directions that party leaders did not want to go, particularly toward economic protectionism. Or they could support Hillary Clinton, who opposed their entire agenda. Soon after he received the nomination, most Republicans fell in line behind Trump and, by the time of the general election, voted for him.25

  By comparing the communities that supported Trump in the primaries and caucuses to the communities that supported him in the general election, w
e can see how fragile and unstable these Republican alliances were. The core counties that supported Trump in the primary were not all the same as those that supported him in the general election. This is because orthodox Republicans had mostly lined up behind Trump for the general election, and Trump was facing a Democrat rather than fellow Republicans. This shows deep divisions in the Republican Party made deeper by Trump’s candidacy, during which his core supporters backed him, hoping he would deliver on his economic promises to those in struggling communities, while orthodox Republicans also backed him in the general election, hoping that he would continue to fight the traditional Republican battles once in office.

  To see this, we first examine how the same set of variables used to predict primary voting is related to Trump voting in the general election. The full regression results are in the appendix (table A.3, column 1). To keep things simple, here we mention only the key differences when comparing the primary vote to the general election vote. In the general election, Trump gained more support in sparsely populated counties—this is the rural and urban divide—but that variable was not a significant predictor of the primary vote. While during the primaries, Trump gained support where the median age was high, the opposite was true in the general election. Median household income was not a significant predictor in the primaries, but it was in the general election, with Trump getting more support in the wealthier counties. Even though he gained his core supporters by appealing to economic hardships faced by white voters, in the general election, like Republicans always do, he received more support where median incomes are high. This means that regular Republicans who turn to the party for its economic policies that benefit the wealthy turned out for Trump, despite Trump’s protectionism. Similarly, while he gained support in the primaries in counties with high unemployment, the opposite was true in the general election. Hillary Clinton gained more votes than Trump, net of other variables, in counties where unemployment was high. Trump secured the nomination by appealing to voters in the hollowed-out economies of the heartland, but he won the general election like a normal Republican candidate.

  Trump did well in counties with traditional family and gender arrangements. He gained support where relatively few women were in the labor force, where men were disadvantaged relative to women in education, and where relatively few adults were married. These conditions made Trump’s promises to bring back manufacturing jobs attractive, especially for men and for women who were interested in maintaining—or entering into—traditional marriages. But when we look at the general election, we see something very different. Trump won counties where high percentages of adults were married and where men were better educated than women. The percent of women working had no correlation to the vote for Trump versus Clinton. Once again, we see that normal voting patterns took hold when we got to the general election because most Republicans—not just his base—backed Trump.

  Trump fared well in counties with high proportions of evangelical Protestants in both the primaries and the general election, but the proportion of Catholics in a county had no effect on the vote in the general election. In the general election, neither the percent employed in retail jobs nor the percent employed in manufacturing jobs predicted the voting outcome. However, in the general election—like the primary and caucus elections—Trump did much better than his Democratic opponent in counties with relatively few college graduates. While to some extent it may reflect different cultural tastes between those who are college educated and those who are not, it is a key indicator of whether a county had enough college-educated residents to link their community to the global economy.

  It was with no small amount of trepidation that traditional Republicans voted for Trump. The correlation between the vote for Trump in the general election and the vote for Romney in 2012 is extraordinarily high, but there is virtually no correlation between the vote for Trump in the primary and caucus elections and the vote for Trump in the general election. In other words, knowing what percent of voters sided with Trump in the primaries in a particular county would be useless when trying to guess what percent voted for him in that same county in the general election. This is because Trump set himself apart from other Republicans in the primaries in ways that made him attractive in the kinds of counties that don’t normally vote Republican in the general election. In the general election, traditional Republicans got behind Trump, making the voting results look very similar to how they would have looked if we were examining the vote for Mitt Romney in 2012 or John McCain in 2008. Satisfying these two very different constituencies—Trump’s base and orthodox Republican voters—would challenge Trump when he entered the White House.

  TRUMP VERSUS ROMNEY

  Even though the types of communities that supported Trump in the general election are very similar to those that supported fellow Republican Mitt Romney four years earlier, there are a few notable differences. To assess those differences, we run the same analysis, but this time controlling for (or statistically holding constant) the percent of the vote for Romney. This way we can identify attributes of counties that predict either more or less support for Trump compared to the vote for Romney. The full results are in the second column of table A.2 in the appendix.

  As expected, even after controlling for other variables, the vote for Romney was a strong predictor of the vote for Trump in 2016. Trump tended to do better in counties that weren’t densely populated and where men had no education advantage over women. It also appears that Trump’s racial appeals made a difference in the general election. Although Republican candidates normally do well in the general election in counties with mostly white residents, that was especially true of Trump. Even after considering the prior Romney vote, Trump did exceptionally well in predominantly white counties, and exceptionally poorly in counties with higher concentrations of minority voters. Finally, and importantly, we once again see the importance of education. Taking into account the Romney vote, we see a strong increase in Republican voting (i.e., for Trump) in counties with relatively few college graduates.

  To recap, in spite of the similarities in the vote for Trump and Romney in the back-to-back general elections, Trump nevertheless fared much better in whiter counties and in counties with few college graduates.

  * * *

  In chapter 4 we describe how the lost economic power of white Americans who missed out on the riches of globalization created the preconditions for white nationalism. This disrupted the tenuous alliances between Republican constituencies that have been in place for forty years. Although the Republican Party consistently advocated policies beneficial to wealthy Americans, voters without such economic privilege have supported the party for its policies that preserve other privileges, especially along the lines of race, gender, and religion. Party leaders have held this tenuous coalition together by arguing that conservative policies benefit all Americans by promoting economic growth. These promises wore thin, especially in the aftermath of a severe recession and a selective recovery. Trump’s campaign excited those Republicans, however, who believed he was offering something new to address their circumstances—a brand of white nationalism tailor-made to reinstate their economic status. Much like the Klan of the 1920s, Trump intertwined his appeals to economic grievances with appeals to privileged identities based on race, gender, and religion.

  In this chapter we examined election results for the primaries, caucuses, and general election of 2016. We saw how Trump’s appeal, when competing for the Republican nomination, was particularly attractive in locations where relatively few residents held a college degree. The size of the college-educated population not only predicted the vote for Trump but also determined how other community attributes—like income, unemployment, percent of women who work, marriage rates, and religious identity—correlated with Trump support. While Trump secured the nomination by whipping up a core group of supporters in counties where Republican candidates typically fared poorly, it was the support of traditional Republicans in the general elec
tion, combined with Trump’s base, that put him in the White House.

  This deep divide among those who voted for him, however, will challenge effective governance. Trump energized Republican voters who did not buy into traditional Republican economic goals that favored the wealthy. Through his brand of protectionism and nationalism, he made it clear to voters that he planned to direct his efforts toward shoring up the economies of white communities. But he was still beholden to the economic elite, and would have to negotiate an alliance between coalitions now that his candidacy brought these disparate agendas out into the open.

  6

  POLITICS AND WHITE NATIONALISM

  The political sociologist Barrington Moore called the American Civil War the last bourgeois revolution.1 Moore believed there was no inherent conflict between an agrarian slave economy in the South and a rapidly industrializing, wage-based economy in the North. The problem, he argued, was the unresolvable stalemate of housing these two economies under the same political roof. “The fundamental issue,” he wrote, “became more and more whether the machinery of the federal government should be used to support one society or the other.”2 The national government could not simultaneously meet the demands of both, and the bloody civil war that followed not only ended slavery but made it possible to integrate the Northern and Southern economies.

 

‹ Prev