At nearly the last moment, one final event fed Trump’s momentum. On October 28, FBI Director James Comey announced that the investigation into Hillary Clinton’s unauthorized email use as secretary of state would be reopened. The route by which this occurred was circuitous and sordid. Clinton’s top personal aide was Huma Abedin, whose estranged husband was Anthony Weiner, former Democratic congressman from New York. Weiner had been forced to resign in 2011 after he admitted to sexting with multiple women. In 2016, a new round of Weiner sexting, this time directed at a fifteen-year-old, led to an investigation. In the process of examining the former congressman’s laptop, the FBI came across a new batch of emails from Clinton to Abedin. Comey, who had clearly wanted to split the difference in July, was forced to decide in October. If he reopened the investigation before Election Day, he would be accused of tilting toward Trump. If he kept the 650,000 emails on Weiner’s laptop under wraps until after Election Day, he would be accused of covering up for Clinton and keeping important information away from the voters. After receiving considerable heat in recent weeks as details emerged that seemed to show the original FBI investigation might have been corrupted to let Hillary off the hook, Comey decided to go the other way.34 Democrats, who had praised Comey in July, were apoplectic. What Comey giveth, Comey could taketh away.
The FBI worked overtime to process the emails, developing new software to try to identify which emails had not yet been seen in previous reviews. Two days before Election Day, Comey sent yet another letter to Congress, indicating that the reopened investigation was, again, closed. But damage was done. Clinton’s seven-percentage-point lead of October 17 had already fallen to 4.5 percent on October 27, and it narrowed further to 1.6 percent by November 4. Correspondingly, her brief moment of glory atop 270 electoral votes at the end of October ended in a freefall. By November 8, RealClear-Politics estimated Clinton’s electoral votes at only 203, with Trump’s count at 164 and a decisive 171 too close to call. Although national polls tended to show a slight recovery for Clinton at the end, it was clear that the last phase of the campaign belonged to Trump. No longer were Texas and Utah tossups; rather, the last weekend saw Clinton rushing to Michigan, long thought safe for the Democrats, and Pennsylvania, which they thought had been locked up a few weeks before. Clinton spent November 7 campaigning with Barack and Michelle Obama, Bon Jovi, and Bruce Springsteen at a Philadelphia rally that drew 30,000.35 In a whirlwind befitting his entire campaign, Trump traveled to Florida, North Carolina, Pennsylvania, New Hampshire, and Michigan, promising to “drain the swamp” and end the “betrayal” of America by the political class.36 The experts averted their gaze, putting their faith in national polls that continued to show Clinton with a narrow lead. Prediction markets gave her a four-in-five chance of winning.
For their part, the third-party candidates had mostly faded, falling victim to the longstanding tendency of the electorate to move toward the major party choices as Election Day draws near in order to avoid “wasting” a vote on a sure loser.37 Despite having a golden opportunity to draw the disaffected from both major parties, Gary Johnson’s Libertarian effort had peaked above 9 percent in the RCP average in mid-September but suffered steady decay since then, abetted by the candidate’s failure to identify “Aleppo” (the town in Syria that had been the focus of a widely publicized humanitarian crisis for months) or any foreign leader he admired and by his running mate’s frequent paeans to Hillary Clinton. Johnson fell short of the 15 percent national poll average required for inclusion in the debates, complained, and got nowhere. Jill Stein had peaked at nearly 5 percent in late June but had been on a downward slope since then. Evan McMullin’s national appeal was blunted by his obscurity and limited ballot access, and even his drive to win Utah was apparently stalled, as the last polls showed Trump pulling away.
Now all that was left was to vote, and wait.
RESULTS
Not since Harry Truman beat Thomas Dewey in 1948 did the final results of a presidential election surprise so many participants and observers. Perhaps none were more surprised than the Democrats who turned out to Hillary Clinton’s “victory party.” In 1948, one correspondent reported that “Republicans who stayed up all night at national headquarters to celebrate looked at dawn today like haggard brides left waiting at the church.”38 Democrats in 2016 knew the feeling, as they turned from confidence to cautious optimism to concern to despair. Donald Trump took an early lead in popular votes that held up from the first East Coast results to the border of California. More important, he also took an early electoral vote lead and never gave it up.
As results came in, Trump won the key big eastern swing states: Florida, North Carolina, and, surprisingly easily, Ohio. He pushed Clinton to the limit in Virginia before losing it in the end, another early warning sign for Democrats. It was when reports headed west, though, that panic began to set in at Clinton headquarters. Unexpectedly—except perhaps to Trump strategists, who had long seen this as their man’s path to the White House—Trump jumped out to leads in Wisconsin and, more narrowly, in Michigan. He never relinquished them. Then, decisively, after the Philadelphia vote had delivered a large (but not large enough) lead to Clinton, the rest of Pennsylvania came in. In the middle of the night, returns from rural Pennsylvania put Trump on top, where he stayed. Pennsylvania, Michigan, and Wisconsin together were enough to give Trump well above the electoral vote majority he needed. (Though little noticed in the hoopla, he also came within 1.5 percentage points of taking Minnesota, which had not gone Republican since 1972.) He had brought down the vaunted “blue wall” of Midwestern states that Democrats had long counted on. Clinton won in the swing states of Colorado, Nevada, and New Hampshire, the last of which was not called for several days, but it did not matter. Clinton declined to concede publicly during the night, perhaps too shocked or emotionally spent to face the cameras and the crowd. The next morning, she spoke to the nation, acknowledged her loss, and withdrew to nurse her wounds.
Altogether, Trump won thirty states worth 306 electoral votes, while Clinton prevailed in twenty states plus the District of Columbia with 232 electoral votes. He flipped a total of six states from blue in 2012 to red in 2016 (Florida, Iowa, Michigan, Ohio, Pennsylvania, and Wisconsin). Trump’s electoral vote total was the most for a Republican candidate for president since George H. W. Bush was elected in 1988. As commentators noted, one could drive from the tip of Florida to the border crossing with Canada in Idaho without ever touching a state that voted for Hillary Clinton.
Yet, as absentee and mail-ballot votes continued coming in from California, Washington State, and New York City, Clinton gained and then expanded a lead in the nationally aggregated popular vote. By the time states were done counting their votes, Clinton had won 48 percent to Trump’s 46 percent. Her lead reached about 3 million votes nationally, which was more than accounted for by her 4.3 million vote advantage in California, one of the few states where she won a higher percentage than Barack Obama had in 2008 or 2012. So, as Republicans celebrated a presidential win, Democrats noted that 2016 represented the sixth election of the last seven in which the Democratic candidate had won more votes nationally than the Republican. Indeed, Trump received a smaller percentage of the total vote than Mitt Romney in 2012 (47.2 percent) and only slightly higher than John McCain in 2008 (45.6 percent). However, as a share of the two-party vote, Trump bested Romney by nearly a point and a half nationwide, improving the Republican presidential share in forty states. Though leading Trump, Clinton had suffered a fall-off of nearly three percentage points of the total vote since Obama had won 51 percent four years earlier. Although Democrats complained bitterly about losing in the Electoral College, their defeat was the result of the hyper-concentration of Democratic votes in coastal urban centers—or, to put it another way, the lack of Democratic appeal in a broad enough swath of the country. Clinton “ran up the score” in places she already had well in hand, while Trump’s votes were spread more efficiently. And the Democrats’ problem
was getting worse: In 2000, Al Gore led George W. Bush by 500,000 votes while barely losing the Electoral College. Clinton led by 3 million and still soundly lost the electoral vote. Overall, looking at county-level results, the 2016 election was the most geographically polarized of the last six elections, with a higher proportion of votes coming from counties giving more than two-thirds of the vote to the winning candidate.
The third-party and independent candidates limped across the finish line. Gary Johnson finished with 3 percent of the national vote, Jill Stein with 1 percent, and Evan McMullin with half of a percent. Johnson had pockets of greater strength, including his home state of New Mexico, where he polled 9 percent, and a number of prairie and Pacific Northwest states where he won 5–6 percent. Stein had a universally dismal night, breaking 2 percent in only the hippie strongholds of Hawaii, Oregon, and Vermont. McMullin ended with 21 percent in Utah, 7 percent in Idaho, and a smattering everywhere else people could vote for him. Overall, third-party candidates gained almost 6 percent of the national vote, three times the rate in 2012 but less than one-third what Ross Perot polled in 1992, in the last full-scale outbreak of outsiderism. Maybe the third parties did not do as well in 2016 because many voters saw Trump as a third-party candidate himself, operating under a Republican label.
So how did Trump win?
The answer fixated on by analysts was that Trump had broken through, as he had throughout the Republican primaries, with white working-class voters. Given how frequently this group was referenced in 2016, it might be surprising to note how difficult it is to define and identify it. One common way is by education; here, white voters without college degrees—fully one third of the electorate—gave Trump 66 percent of their votes.39 In the crucial states that he took away from the Democratic column, these voters represented an even larger share of the electorate. In Iowa, they were half of the electorate, in Wisconsin nearly half, and in Michigan and Pennsylvania two in five. Another is by income. Trump’s best income groupings were among those who made between $50,000 and $200,000, the great middle of the American income distribution. In comparison with 2012 results, however, Trump did better than Romney in the below-$50,000 group and worse than Romney above it. As usual, the Republican candidate did better in the top one-third of the income distribution, the Democrat in the bottom one-third—but the gap between the two was much smaller than usual.40 Results in some key working-class and rural counties demonstrated the general tendency: from 2012 to 2016, Juneau County, Wisconsin, went from a seven-percentage-point Obama win to a twenty-six-point Trump win; Macomb County, Michigan, shifted from a four-point Obama edge to an eleven-point Trump advantage; Lackawanna County, Pennsylvania, saw Obama’s twenty-seven-point win turned into a four-point Clinton win; counties like Michigan’s Huron and Tuscola, which had given about 55 percent of their votes to Romney, gave two-thirds to Trump. Moreover, working-class and rural counties and counties with a larger proportion of voters aged fifty-five and older saw increased turnout, up 1.4 million in four years.41 In 2012, Romney enthusiasts had pointed to the missing white working-class vote in Ohio and elsewhere as the key ingredient in Romney’s loss, and 2016 may well have proven their point. Conversely, Clinton expanded on Obama’s margins in places dominated by affluent professionals such as Montgomery County, Maryland, and Fairfax and Loudon counties in Virginia, as well as college towns like Boulder, Colorado, and Charlottesville, Virginia.
However, Trump’s strength among working-class whites is far from the only factor behind his win, as the exit polls yielded several other crucial (and sometimes surprising) results outside of Trump’s domination of the white working class. Along with Trump’s strength where it was widely expected, at least two things did not happen that were expected by analysts.
The women’s vote did not turn against Trump with sufficient strength to give the election to Clinton. To be sure, there was a gender gap—Trump defeated Clinton 53–41 percent among men and lost to Clinton 54–42 percent among women—but it was basically a wash. In contrast, Obama had beaten Romney by 11 percent among women in 2012, while only losing by 7 percent among men. While Clinton did a single percentage point better than Obama among white women, she still lost them badly overall. Among white women without college degrees, Trump won a whopping 61 percent; among married women, he almost broke even. Clinton also did slightly worse than Obama among black women and significantly worse than Obama among Latinas. Nor was there a mass rebellion of Republican women against their party’s nominee: Trump won 90 percent of the Republican men and 89 percent of the Republican women. Columnist Naomi Schaefer Riley contended in the aftermath of the election, “If the left learns nothing else from this election, perhaps they should understand that there’s no such thing as female solidarity—not, at least, as they envision it . . . [women’s] experiences in life and their views on policy issues tend to be shaped more by whether they went to college, whether they live in a city or suburb or rural area, whether they live in a blue state or a red state, how much money they make and how they identify themselves racially and ethnically.”42
Another earthquake that did not happen was the widely expected outpouring of minority mobilization against Trump. In particular, it was hypothesized early in the campaign that Latinos, outraged by Trump’s characterizations of Mexican immigrants as rapists and his dismissal of Judge Curiel, would rise up en masse to block him from the White House. In actuality, Trump did slightly better among racial minorities than Mitt Romney had four years before. Hillary Clinton got the support of 66 percent of Latinos in 2016 compared with Obama’s 71 percent in 2012, and 89 percent of blacks versus Obama’s 93 percent in 2012. Moreover, while the Hispanic share of the electorate edged up slightly, the African American share fell by the same percentage. There were some states where Latinos did mobilize—for example, adding to Clinton’s big California margin by giving her 71 percent there. But, though there were some snap assumptions to the contrary, upon examination it turned out that Clinton’s southwestern swing-state wins were not the result of an unusual outpouring of Hispanic support, which did not happen. In Nevada, Clinton only garnered 60 percent of the Latino vote, and in New Mexico only 54 percent. Red states for which Democrats had briefly entertained hopes fell through with similar results, with only 61 percent Latino support for Clinton in Arizona and Texas. Clearly, Latino voters, like women voters, were not a monolithic bloc.43
A third pillar of Barack Obama’s coalition had been the young and the first-time voters. Clinton’s struggles with younger voters were well documented during the Democratic primaries, when Bernie Sanders swept millennials, and continued in the general election. In 2012, Obama won eighteen-to twenty-nine-year-olds by a 60–37 percent margin; in 2008 he won by even more, 66–32 percent. In 2016, Clinton won among those eighteen to twenty-nine years of age by a 55–36 percent spread. Again, Trump had run in place, but Clinton had lost ground for the Democrats. When Obama won his change election in 2008, he swept first-time voters, winning nearly seven in ten. Clinton won them, too, but by a much reduced margin of 57–38 percent. In the twenty-five states for which exit poll data was available showing the eighteen- to twenty-nine-year-old vote, Clinton only won fourteen states decisively, with more than 50 percent. In four states Trump won the eighteen-to twenty-nine-year-old vote; in another two states Trump tied Clinton for the lead; and in the remaining five states Clinton won narrowly with less than 50 percent of the vote. A reprise of 2008 it was not.
Overall, a key question of the Obama years was whether his winning coalition was a Democratic coalition, transferrable to any other Democratic candidate, or just an Obama coalition. In 2016, the outlines held, but not the same outsized levels of support. There was a regression to the mean that exposed the chief deficiency of the “coalition of the ascendant”: in order to succeed, Democrats had to win unnaturally high proportions of the vote in their target groups. Hillary Clinton was simply not as appealing as Obama to minorities or the young, and women were not swayed as a bloc by her
historic status.
A supplementary question during the Obama era was whether his campaigns’ technological edge and superior ground games would persist and bear fruit for the Democrats when he was no longer on the ticket. However, by the 2014 midterm elections it was apparent that Republicans were catching up on the technological front, and Clinton’s vaunted ground game did not come through in critical places including Philadelphia, Pittsburgh, Detroit, and Milwaukee. Indeed, some published reports indicated that the Clinton campaign might have inadvertently turned out Trump voters. According to Becky Bond and Zach Exley, “Volunteers for the Clinton campaign in Pennsylvania, Ohio and North Carolina have reported that when reminding people to vote, they encountered a significant number of Trump voters. Anecdotal evidence points to anywhere from five to 25 percent of contacts [that] were inadvertently targeted to Trump supporters.”44
Trump did not just win because Clinton failed to re-create the Obama coalition in its full might. He also performed well among other groups that are typically in the Republican coalition but were thought for a time to be wobbly in 2016. Most notably, a majority of college-educated whites wound up voting for Trump by a 48–45 percent margin, and white evangelicals voted overwhelmingly for Trump, despite his failings. At 80 percent support, Trump did better among evangelicals than Mitt Romney (78 percent), John McCain (74 percent), or George W. Bush (78 percent in 2004). Evangelicals were also a larger share of the electorate than in 2004, when their influence on the election alarmed secular liberals across the country. Other groups normally in the Republican coalition remained strong, as Trump won big among those who attend religious services weekly, among veterans, and in rural communities. Clinton held on tight to highly secular voters, single women, those with postgraduate degrees, and voters in big cities—the “coastal elites.” But Trump narrowly won some critical swing groups, including suburbanites (49–45 percent) and Catholics (50–46 percent), and improved the Republican vote share among the unmarried.
Defying the Odds Page 18