HOW TRUMP EXPANDED HIS COALITION
Lessons for Liberals from the 2020 Election
“IT’S IMPORTANT TO ME that everyone has a place in this country . . . that anyone who’s willing to work hard can live here and feel like they belong” — those were the words of Abigail[1], a voter who I had the pleasure of interviewing weekly over the closing months of the 2020 presidential race. A lifelong Republican, Abigail voted for Trump in 2016 but had admitted to me that she felt regret. She managed a business in a wealthy, cosmopolitan suburb outside of Washington D.C.; she admired diversity, professed support for women’s reproductive rights, and as a mother of two, kindness was important to her. As the summer of 2020 unfolded with a bungled pandemic response and intensifying racial tensions, she felt increasing indisposition toward Trump; she always held umbrage with his rhetoric, but now she felt a burgeoning doubt in the president’s competence and ability to lead a divided nation.
Conventional wisdom in the pundit-sphere pegged voters like her — educated white suburban moms — as a critical constituency that would hand the election to President-Elect Joe Biden. That was the logic behind the campaigns, behind the millions poured into The Lincoln Project, even behind Joe Biden as a candidate. A coalition of turned white women, fired up minorities, and a sliver of the president’s own base of non-college educated white males would accomplish the rare feat of removing an American incumbent. The United States’ most illiberal president in modern history would be removed by its most diverse coalition — and then everyone on the bus would clap.[2]
Evidence seemed to support this narrative through the cycle: historically wide and stable polling leads for Biden, levels of civil unrest unseen since closing of the 1960’s, a broad-based social justice movement, a steady trickle of Republican defections and feel-good anecdotes of turned Trump voters. After two months of Abigail expressing persistent indignation at the president’s behavior, at his divisiveness to me on the phone, I had the sense that we were in store for a historic condemnation of the president. The voters had other plans.
On November 3rd, Abigail decided to cast her ballot for Donald John Trump last minute. Along with her, an increased share of Blacks and Hispanics cast ballots for a second Trump term. Now, it’s important to preface any discussion of Trump’s relative successes with the fact that he also experienced a historic loss for a sitting incumbent. Biden bucked historical trends and won the White House with the highest vote count ever recorded. However, we are comparing the results to the background of polling, punditry, takes, forecasting and campaigning that preceded it. Biden didn’t win a landslide electoral victory — with gains in the house and a decisive senate majority — fueled by a clear rebuke of the president. He reliably underperformed the polls; Democrats lost seats in the house and are in a dogfight in Georgia for a narrow senate majority. Biden won critical states with tight margins through gains in the suburbs and a massive swing in white men and senior citizens — not some progressive fantasy uprising of the marginalized.
But this essay is about Trump’s performance, not Biden’s. If exit polls and county level analysis is any indication, it seems that against all conventional wisdom Trump diversified his coalition from 2016. He traded a chunk of his support among white men for modest gains with minorities. While he did not win the largest share of non-white voters for a modern Republican[3] (that honor belongs to George W. Bush in 2004), Trump made vast improvements over both McCain and Romney. Additionally, whereas white women handed Democrats the house in 2018, they seemed to maintain their support for or even slightly warm to Trump in 2020. It is clear that even in loss, Trump managed to expand the Republican coalition.
This wasn’t what anyone expected. Liberals were counting on a stunning rebuke from women and “POCs”. Even Republicans like Sen. Lindsey Graham, before prostrating himself to the Commander in Chief, warned in 2015 that Trump’s brand of race-baiting xenophobia would do long-term damage to the party and diminish their prospects with Hispanics, a demographic that they’ve been courting in vain since the admonitions of Karl Rove. They were wrong. Donald Trump’s brand of politics put his party closer to Sen. Marco Rubio’s professed aim of a “multiethnic working-class coalition” than it has been in fifteen years. Once again Democrats, Republicans, pollsters, forecasters and other paid brainiacs fundamentally misunderstood the electorate. If the discourse following 2016 upset was centered around understanding the stunning “Whitelash,” the conversation following 2020 will be devoted to understanding why the marginalized groups that the intelligentsia pegged as the president’s victims ended up warming up to him.
TRUMP’S PERFORMANCE IN 2020. The nature of 2020 has made exit polling a particularly difficult science this election. In contrast to past years, the exit polls were a combination of in person exit polls taken at polling places and phone polls of voters who voted by mail or early-in person.[4] Considering the partisan divisions on who voted when, and the pollsters efforts to compensate for that, there is likely more “noise” in the results than past years. Nevertheless, there isn’t particular reason to mistrust the direction of the polling shifts; the shifts in support by race seem to comport with precinct analyses[5] and one wouldn’t expect gender to be particularly distorted by when people voted.
Voter Turnout. At over 160 million ballots cast, the 2020 election had the largest vote count in American history and the highest vote turnout in over a century. Because of the remarkable uptick in turnout from 2016, both Trump and Biden were able to expand their gross electorates, and both could increase their shares of the same key demographics by increasing turnout — what mattered is who increased turnout and their respective vote shares more. Biden won this contest, but it is clear that Trump was able to increase his turnout by shoring up support from his party, registering hundreds of thousands of new voters (beating Democrats at their own game), and making inroads with some women and minorities.
Exit Polls. Exit polls show that Trump was able to increase his support among blacks up to approximately 12%, a 4 point increase from 2016 and the highest Republican share of the black vote since 1996. Perhaps even more surprisingly, the president improved his performance with Hispanic (or what the Democrats, probably to their expense, call “Latinx”) voters by around 3 points from 2016 — just under a third of Hispanic voters favored a second Trump term. Additionally, Trump improved 7 points with Asian voters — a constituency the DNC has pinpointed as long-term investment. Finally, as a special affront to pollsters, who forecasted a historic gender gap of up to 31 points,[6] the actual gap sharply reverted to the mean of around 10 points.[7] White women’s support for the president held firm, and Trump made clear gains with women of color.
Precinct Analysis. Again, the racial shifts in the exit polls is largely reflected by the precinct analysis in crucial swing states. Hispanic Miami-Dade county was responsible for around 75% of Trump’s gain in net votes in Florida.[8] In North Carolina, majority black counties warmed to Trump, and Robeson County, home to the Lumbee Tribe, flipped solidly red.[9] Across Florida, George and North Carolina, Trump improved his showing in majority black precincts by 2 points and in majority Hispanic precincts by over 11 points.[10] Trump made similar gains with Hispanics in Texas and Arizona.
I will assert that I personally was not surprised by the racial shifts; I’m intimate enough with people from these backgrounds to know the lamestream media and Democratic establishment was missing a crucial story there. But being familiar with white women, having worked on a 2018 congressional race where they were critical for victory, and having the opportunity to interview one over the course of the race, their stable or slightly increased support for Trump was an extraordinary subversion of my expectations. However, it’s important not to draw the wrong story from the data — this isn’t a “Blexit” or a mass realignment (yet). In total, a vast majority of women and minorities supported the Democratic nominee; Biden got more net votes among these groups than Hillary Clinton, helping him win the presidency. However, the increased turnout — a perennial progressive goal — modestly improved Trump’s vote shares with constituencies that Democrats rely on. Given Trump’s character, and frankly, the project of the Republican Party, this is extremely alarming and reveals dangerous cracks in the Democratic Party’s desired coalition.
WHAT 2020 TAUGHT US. If this election cycle and the past four years have taught us anything, it’s that there’s a fundamental fragility in the liberal vision for this country; many Americans are susceptible to demagoguery and misinformation, and even for those who are not, partisanship can discipline them into dangerous complacency. Liberals cannot rely on them to have an “Aha!” moment. What needs to happen is an aggressive campaign that meets people where they are. To do that, we need to make sure 2020 taught us the right lessons; those lessons also happen to be the reason that Trump was able to expand his electorate. Through data, reporting, and personal anecdote, I hope to illuminate what they are.
Demographics is not destiny. Democrats will not be able to ride the demographic wave to comfortable electoral margins; rather than shift American public opinion, as non-whites grow in their relative share of the population, they will likely simply start falling into existing patterns of partisanship. Given their disproportionate support for Democrats, I expect a long run regression to the mean could only mean increasing relative support for Republicans in the future. We may have just seen the beginnings of this in 2020. As Hispanics become more embedded in the country’s social fabric; through assimilation intermarriage and cultural absorption — they’ll simply stop seeming and feeling so different. As Asians prevail as the highest-income group, I suspect that their posterior — second, third and fourth generation Americans — will trend towards the voting habits of high-income whites. Blacks pose a special case, but it’s very plausible that a combination of dissatisfaction with the status quo, exhaustion with a perceived pandering and condescension, religiosity, and the continued work of high-profile black republican celebrities and influencers (Kanye West, Diamond and Silk) will push more of them into the ranks of the Republican party. If liberals have any hope of demographic shifts handing them majorities into the future, they’re going to have to understand the people who comprise them.
Women aren’t fundamentally different than the electorate at large. I’m frankly a little embarrassed to have believed that women would be “more vulnerable” to campaign dynamics and that some magical maternal instincts would activate in response to Trump’s rhetoric to more than double a historically stable gender gap in favor of Biden. I’ve learned that this sort of wishful (possibly sexist) ideation should be discarded in favor of the reality that women are basically caught in the same patterns of reliable partisanship as the electorate at large. Any professed dismay at a candidate’s character or rhetoric would mostly succumb to the discipline of partisan identity. This is precisely what happened with Abigail. After months of exasperation with the president’s behavior and flirting with the prospect of casting a democratic ballot, she yielded to her more essential instincts: “I just couldn’t get myself to vote for Biden.”
People gave Trump credit for the pre-COVID economy. While Democrats privately rejoiced over the fact that Trump was presiding during the greatest economic contraction since the Great Depression, it turns out that assigning accountability for the COVID induced recession was more complicated. Until around late October, issue pollsters found that voters still slightly favored Trump to handle the economy; and even those polled were evenly split on who they trust on jobs.[11] Surprisingly, in the fall, around 56% of voters reported that they were better off than they were four years ago. Right before the pandemic, 61% of respondents said that they were better off than they were in the Obama years.[12] The lesson here is that the idea of the electorate mechanically punishing incumbent parties for economic downturns is dated; the narrative around accountability for the downturn matters. Trump was credibly able to blame a foreign virus and the opposition’s heavy-handed lockdown response. In many instances, this caused sections of the electorate to trust the opposition with the economy less. This was explicitly the case with Abigail, who when probed over why she voted for Trump, told me that “I have people on payroll, and I just can’t trust that Biden won’t do another lockdown.”
On balance, people may have primarily associated Trump with the slightly accelerated economic expansion before the pandemic, no matter the pains Democrats went through to attribute it to Obama. If the Trump campaign had any message discipline, it was in hammering down record low unemployment for blacks, Hispanics, and women. Progressives balked at this reductive characterization of the economic situation, where underemployment, stagnant wages, and high costs of living were harming the prospects of all of these groups — but in the end, the Trump narrative seemed to make a dent in their base. By rejecting the optimistic Trump narrative that Nobel-prize winning economist Robert Shiller suggested had a hand in improving consumer confidence and lengthening the expansion,[13] Democrats looked like “haters and naysayers.”[14] In 2020 and 2016, the narrative was more important than — and even in critical in affecting — the economic fundamentals.
For newer Americans, the politics in their country of origin matters. As an “Asian” American, Trump’s 7 point improvement with the demographic wasn’t surprising. For one thing, most Asian countries represented in the United States have a predominately negative view of China, and I know personally that Trump’s tough rhetoric on “the Communist Party of China” turned many of these voters on to him. Biden’s claims of xenophobia rang as tone deaf and weak. A particularly interesting anecdote is where one of my mother’s coworkers, a Uyghur woman on a green-card who had family members stuck in Xinjiang province, said she loved Trump because he was the only one who would stand up for her people. The fact that he had condoned the concentration camps according to his own national security advisor and that his administration is admitting a record low number of refugees did not register with her. Now she couldn’t vote, but on hearing on this Muslim immigrant woman of color’s love for Trump, I knew that Democrats had a severe messaging problem.
Many Indians also warmed to Trump (exit polling categories doesn’t separate them, but pre-election polling shows a modest shift),[15] as his brand of politics resonated with the nationalist (really, fascist) uprising back home. Prime Minister Narendra Modi is a historically powerful and popular prime minister, particularly with the kinds of upper-caste Hindus that immigrate to the United States. His rapport with Trump was well known and well covered in Indian media, this probably turned a good deal of Hindus on to him (astoundingly, my own family members, who were trying to bypass legal immigration procedures with phony asylum claims based on religious persecution, liked Trump!). Segments of America’s model minority are more sympathetic to reactionary politics and authoritarianism than we give them credit for.
Finally, as has been covered ad nauseam, Trump’s charges of socialism against the Democratic party likely had a very real impact on Hispanic voters.[16] Not obvious perhaps is that his comparisons of the party agenda with Venezuela-style socialism was probably effective with some not only because they don’t want to be like Venezuela, but because of many from the neighboring countries aren’t particularly fond of the Venezuelan refugees pouring in to their countries of origin (source: my Peruvian roommate).[17] In the end, Trump was able to shave off some of the Democrat’s minority support because these people aren’t the down-trodden little guys who Democrats want them to be, they’re full-fledged people with their own histories, cultures and even deep-seated bigotries — how very American.
We don’t hear and see the same things. The most important lesson — in fact the macro lesson that encompasses all the others — is that Americans are not hearing and seeing the same things. This may sound like a banal point, anyone paying attention knows how bifurcated our media has become. But the point I’m making is more sophisticated, and more important than this. Yes, media polarization and social media means in some instances we’re seeing different content; what’s more interesting though is how we’re responding so differently to the same things. President Trump isn’t really micro-targeting his messaging; his campaign might do this in the digital realm but if anything, the Democrats have been doing this even better. No, his rallies and his twitter feed go to a mass audience; his voters themselves are segregating his message by projecting their own hopes and anxieties onto his heavily stylized language. His style allows people to pick where he’s being sincere, and where he’s being a blowhard — different people make different selections.
Rather than targeting different images to different people, the president has presented a singular gestalt image — where what you see depends on where you look. Like the famous internet memes “The Dress”[18] or “yanny laurel,” once someone’s seen or heard one version its near impossible to convince them of the alternative perception. Trying to convince a Trump fan to see an incompetent bigot is more like trying to convince someone that “The Dress” is blue when they see gold than a matter of intellectual debate.
This is why my mom’s muslim coworker could see a humanitarian where my family members see kindred spirit to Modi’s anti-muslim Hindu nationalism. Additionally, where white liberals heard attacks on Hispanics, many Hispanics themselves didn’t hear it as an assault on themselves.[19] No, when Trump was attacking drug dealers, criminals and rapists he wasn’t attacking them, he was attacking those townies back home, the bad ones.
The difference between 2016 and 2020 Trump is clear when you compare his Inauguration speech to his 2019 speech. American carnage gave way to unbridled American optimism, a narrow message began to accommodate a larger audience. I would argue that in 2016, the president won because of a disciplined singular narrative: “you’ve been cheated by a corrupt status quo designed to help outsiders, I will fight to help you.” In 2016, it was easy to tell Trump was addressing a white anxiety. Three years into his presidency however, his image splintered — he messaged on broad based inclusive economic growth, had appointed conservative justices, and made very direct appeals to minorities. Trump the nativist multiplied into Trump the jobs creator, Trump the bulwark against socialism, and even Trump the humanitarian. In contrast to Trump the insurgent, Trump the president offered a gestalt image that broadened his appeal.
CONCLUSION. In the days and weeks following the 2020 election, I could tell that many Democrats, particularly on the far left, learned the wrong lessons. Many saw Trump’s expanded appeal as evidence that systemic racism and sexism were even more pervasive than we thought. Hispanic and Black men were attracted to Trump because of an innate machismo, white women were continuing to uphold a racist patriarchy that benefitted them. To these people, what was needed was more education, more activism. I doubt that’s the answer.
I don’t doubt that some voters were motivated by latent racial anxieties or sexism; but for most of these people it’s unconscious, and they don’t want a sensitivity training. Abigail didn’t vote for Trump because of racial anxiety — and I looked for it. She didn’t care about school zoning in our county bussing kids from the Hispanic neighborhoods, or high-density housing being built near her home (“I think that’s good. I want my kids to know not everyone is like them.”) She weirdly liked Kamala Harris. Trump’s racial rhetoric offended her; she voted for him because she didn’t trust Biden. When asked about the president’s corruption, his authoritarian instincts, she retorted “well that’s what checks and balances are for, right?” While many voted for Trump out of anger and racial resent, I suspect that Republicans on the margins voted for him because of a basic faith and optimism that our institutions could withstand the ramblings of a very rude man. They felt it was safe enough to fall back into their partisan identities.
It would be easy to write off the votes of white women as irresponsible complacency towards a racial hierarchy that benefits them if its purported victims — blacks and Hispanics — didn’t also warm to Trump. No, these people didn’t vote for Trump because they thought he would uphold a racial hierarchy.
We know this: Democrats’ attempt to convince more voters that Trump was a racist (or worse, that they were secretly racist, but they just didn’t know it) didn’t really work. To them The Dress is firmly gold, not blue. Trump’s gains among minorities after four years of shouting “racism!” should be a hard lesson for liberals, and likely a source of vindication for Republicans for years to come (“If Trump was a racist then why . . . “). Instead of doubling down on that strategy, as many progressives are inclined to, we have to try something different. Instead of trying in vain to get people to see the flip side of the gestalt image, give them something prettier to look at instead. I believe this is more or less what happened this election. The pundits have pegged the 2020 election as a referendum on character. That’s probably the case, but I don’t think that Biden swung the white vote because he convinced new people that Trump was a racist misogynist, I think that many voters were just ready to change the channel. It’s important that we know what actually worked.
To sum up, Trump’s ability to expand his coalition with the very people who were supposed to rebuke him teaches us that Americans are more complicated, and America more interesting than any modeler can capture. People won’t behave how we want or expect them to, and some probably delight in doing precisely the opposite.[20] There’s a natural resistance to being pegged as a perennial victim; people desire a sense of agency. There is probably a perverse satisfaction in voting for someone you’re not “supposed to” — I would argue that there’s a characteristically American rebellious spirit behind it. Trump earned more support among these groups not because they wanted to ingratiate themselves to a racial hierarchy, but rather because they are becoming ever more American, warts and all.
[1] Alias is used to protect the interviewee’s identity.
[2] This is a reference to an internet meme. https://knowyourmeme.com/memes/and-then-the-whole-bus-clapped
[3] A Republican after the realignment of 1964
[4] https://fivethirtyeight.com/features/exit-polls-can-be-misleading-especially-this-year/
[5] Precinct shift maps show shifts to Trump in predominately black and Hispanic counties; most notable is Miami-Dade county which may have solidified Florida as a Republican state.
[6] https://www.washingtonpost.com/politics/2020/09/29/daily-202-trump-faces-historic-gender-gap-heading-into-first-debate/
[7] https://apnews.com/article/election-2020-joe-biden-donald-trump-voting-rights-elections-84ef3db79532c0029894ff25a316370b
[8] https://www.wsj.com/articles/trumps-florida-election-win-hinged-on-big-gains-in-miami-dade-11604518272
[9] https://www.fayobserver.com/story/news/2020/11/12/nc-vote-between-trump-and-biden-shows-gop-dem-shifts-2020-election/6227924002/
[10] https://www.nytimes.com/interactive/2020/11/03/us/elections/forecast-president.html
[11] https://www.usnews.com/news/elections/articles/2020-10-21/poll-biden-pulls-ahead-of-trump-on-handling-of-economy
[12] https://news.gallup.com/opinion/gallup/321650/gallup-election-2020-coverage.aspx
[13] https://www.cnbc.com/2019/10/20/shiller-recession-likely-years-away-due-to-bullish-trump-effect.html
[14] This is my favorite turn of phrase from Boris Johnson.
[15] https://timesofindia.indiatimes.com/world/us/despite-small-shift-toward-donald-trump-2/3rd-of-indian-americans-back-joe-biden/articleshow/78139899.cms
[16] https://www.nbcnews.com/news/latino/caudillo-or-crusader-against-socialism-florida-s-latinos-debate-trump-n1100036
[17] https://www.amnesty.org/en/latest/news/2020/02/peru-rechazo-ilicito-de-personas-venezolanas-que-buscan-proteccion/
[18] https://www.wired.com/2015/02/science-one-agrees-color-dress/
[19] This is a great Ezra Klein podcast on the issue https://open.spotify.com/episode/4V2GXkWuh0IxsYWvbfDmzV?si=dxNS7MSTR3K3ZhypDt81QQ
[20] See, Kanye West
