Every election year is accompanied by countless analyses of why Americans voted the way they did. The 2020 election is no different.
Liberal and Democratic Americans who were hoping for an overwhelming victory for Joe Biden were disappointed, although the Democratic candidate did defeat a sitting president by nearly 6 million votes. This is itself noteworthy considering how infrequently incumbent presidential candidates lose reelection in the post-World War II era.
The questions that arise now are simple enough: what caused Trump to lose the election, and to what extent did Americans vote for Biden, as opposed to voting against Trump? Answers to these questions emerge from a careful examination of pre-election polling and Edison’s national exit polling, the latter of which surveyed both in-person and mail-in/absentee voters to collect information on voters’ demographic backgrounds and their political and economic motivations. What we find is that mass anger at racial injustice, the disaster of Covid-19, and the depressed economy that accompanied it, were all chief difference makers in the election outcome.
Continuity with Previous Elections
First things first, I should point out that it’s perhaps surprising that so many on the left thought or hoped Biden would win by massive margins over Trump. U.S. presidential elections in the last two decades have almost always been highly competitive, with five of the last six elections (including 2020) having outcomes in which the candidates finished within less than 5 percentage points of each other in the popular vote, as we see below:
2000: Al Gore, popular vote winner by +0.5%
2004: George Bush, +2.4%
2008: Barack Obama, +7.2%
2012: Barack Obama, +3.9%
2016: Hillary Clinton, +2.1%
2020: Joe Biden, +3.8%
It is an exceptional circumstance, as we saw in 2008, when one presidential candidate wins by more than 5 percent. With Barack Obama’s win over John McCain, it took the dual forces of a collapsing economy and mass anger over the war in Iraq to turn the tide so heavily against the Republican Party. Looking at the election results above, it is obvious that the United States is strongly divided when it comes to electoral politics, and this division is unlikely to subside in the foreseeable future.
In some ways, this election was simply more-of-the-same, with the breakdown of voter support for Biden and Trump in 2020 echoing the divisions seen between Clinton and Trump in 2016. This is apparent after reviewing the Edison exit polling from both elections. Partisan and ideological Americans overwhelmingly cut in favor of their preferred candidates in both 2016 and 2020, with 90 percent or more of liberals and Democrats preferring Clinton (2016) and Biden (2020), and with similar numbers of conservatives and Republicans favoring Trump in both election years. Racially, patterns observed in voting from 2016 have persisted into 2020. Black Americans voted 87-12 percent for Biden over Trump in 2020, compared to an 88-8 split favoring Clinton over Trump in 2016. Similarly, LatinX voters preferred Biden over Trump 65-32, compared to favoring Clinton over Trump by 65-29. Whites continued to favor Trump over Biden by 58-41 in 2020, compared to by 58-37 supporting Trump over Clinton in 2016. The gender split in voting remained much the same as well. Men voted for Trump over Biden by 53-45 in 2020, while favoring Trump over Clinton by 53-41 in 2016. Women preferred Biden to Trump by 57-42 in 2020, compared to their preference for Clinton over Trump by 54-42 in 2016. Education and income remained consistent markers of division in voting for 2016 and 2020. Individuals with a college degree favored Biden over Trump in 2020 by 55-43, while individuals with a four-year degree favored Clinton over Trump by 49-45, and those with a post-grad degree favored Clinton over Trump by 58-37. Poorer Americans continued to prefer the Democratic candidate in 2020. Those earning less than $30,000 a year voted Biden over Trump by 54-46 in 2020, compared to favoring Clinton over Trump by 53-41 in 2016. Similarly, those earning between $30,000 to $50,000 preferred Biden to Trump by 56-43 in 2020, compared to favoring Clinton over Trump by 51-42 in 2016.
As related to income, the above findings undermine the myth that Trump’s support is heavily dependent on “working class” voters. His support was always strongest among more affluent Americans, and that support grew from 2016 to 2020. But the biggest gap in support for Republicans over Democrats has never been among the wealthiest Americans. Those earning $200,000 to $249,999 a year were evenly split between Clinton and Trump in 2016, 48-49, while those earning more than $250,000 a year were also divided between Clinton and Trump by 46-48. Similarly in 2020, those making more than $200,000 a year were split 44-44 between Biden and Trump. Rather, the biggest gain for Trump between the two elections was among middle-upper class voters, with those earning between $100,000 to less than $200,000, who were evenly split 47-48 between Clinton and Trump in 2016, but with this same group cutting in favor of Trump over Biden by 58-41 in 2020. These findings undermine the notion, promulgated endlessly over the last 4 years, that Trump speaks for the “common man” who has fallen behind in the neoliberal era due to rising financial insecurity, and as a function of outsourcing, downsizing, and the rise of the contingent “gig economy.”
What Changed This Time Around?
Despite all of the continuity documented above, Edison exit polling reveals that one big shift between 2016 and 2020 was due to the national trauma inflicted by the Covid-19 crisis, which represents the worst pandemic in a century, and which has taken to date about 250,000 American lives, while deeply depressing the U.S. economy. Those most concerned with Covid-19, and those who have been the most harmed by negative economic developments over the last year, cut steadily – sometimes heavily – against Trump and in favor of Biden. On the Covid-19 front, and for individuals who said that the pandemic “mattered most in deciding how you voted for president,” the overwhelming majority voted for Biden over Trump, by 81-15 percent, with these voters representing a significant 17 percent of the voting public. Similarly, individuals who agreed the “U.S. efforts to contain the coronavirus pandemic” was going “somewhat badly” or “very badly” preferred Biden to Trump by an overwhelming 81-18.
As Covid-19 has imposed severe economic consequences on the public, those struggles also informed the vote. Polling data from as early as March 2020 suggested that disapproval of Trump was running about 15 percentage points higher in parts of the country that experienced the most “severe” economic impacts of Covid-19. And this early anxiety continued through election day, as Edison’s exit polling revealed. Individuals who said their “family’s financial situation” was “worse today” (following the election) compared to “four years ago” cut toward Biden by 77-22. Similarly, individuals who said that “the coronavirus pandemic caused you” “a severe financial hardship favored Biden over Trump by 69-29, while those who said they faced “moderate financial hardship” because of the virus voted Biden over Trump by 59-39. More generally, those who said the U.S. economy was doing “not so good” or “poor” favored Biden to Trump by 80-17.
On the issue of racism in America, large numbers of voters cast a vote against the Trump administration, in opposition to its tone-deaf approach to protests of police brutality disproportionately directed against people of color. Voters who cited “racial inequality” as the issue that “mattered most in deciding how you voted for president” overwhelmingly favored Biden over Trump by 92-7, with these voters representing a sizable 20 percent of the electorate. Combining the above findings on racial inequality and Covid-19, we can confidently conclude that a solid third of the electorate – 32 percent – were primarily motivated to vote against Trump either because of their support for anti-racism, or due to their anger with the Coronavirus pandemic – the latter of which has been greatly worsened by the Trump administration’s non-response to the crisis.
What 2020 Wasn’t About
The exit polling data should be enough to hammer the final nail in the coffin for the “Trump is a working-class hero” narrative that first emerged in 2016, and which has persisted over the last four years. This narrative depicts Trump’s support, and reactionary political values more generally, as springing from mass rage over the economic losses associated with outsourcing, downsizing, and mass financial insecurity. This narrative has always suffered from an almost total lack of empirical evidence, as I’ve exhaustively documented through a mountain of studies of available survey data, and which you can find here, here, here, here, and here, and in my last two books, Rebellion in America, and Unequal America.
The Trump “working-class” narrative – to the extent that it has been embraced by many on the left and beyond, represents a blatant form of plutocratic false consciousness. The myth that financially insecure whites support Trump originated from Republican Party propaganda, right-wing media, and from Trump himself. It was then amplified by journalists in the mainstream media who have systematically failed or refused to challenge the myth. As far as I can tell, it stems almost entirely from the finding that white voters with lower levels of education heavily support Trump over Biden. But there is a lot of diversity among less educated whites – especially those with two-year college degrees or vocational post-high school training – in terms of their incomes. And it’s a mistake to simply assume that because an individual has only a high school diploma or a two-year college or vocational degree that they are financially insecure, or that economic insecurity is driving them to support Trump. Most-all of the Gen Xers and Baby Boomers I know who have only a high school degree or a two-year/vocational degree earned well above the national median household income of $68,703 in 2019. I expect that many Americans would say the same thing about many of the older, less educated Americans that they know.
Tangible economic metrics reinforce the notion that we should avoid blanket references to “working-class” Americans as financially insecure. As I document in Rebellion in America, self-identified “working class” whites are not significantly more likely, statistically speaking, to identify as Trump supporters. And “working class” whites who hold factory jobs, while more likely to support Trump, are not more likely to struggle financially, when looking at their incomes. They are significantly more likely than other households to earn above the national median income.
Contrary to popular myth, support for Trump is not significantly associated with “deaths of despair” among economically insecure whites, and as I’ve carefully documented in previous research. While white Trump supporters are more likely to say they struggle with depression and alcohol abuse (although not illicit drug abuse), these struggles are not linked to financial insecurity (lower incomes), and such drug problems are common across the income spectrum, among lower, middle, and upper income earners. And even prominent scholars such as Angus Deaton and Anne Case, who have traced the rising frequency of suicides and drug abuse back to white blue-collar Americans, do not suggest that this group’s anxieties stem primarily from financial insecurity. They have also avoided the unfounded claim that white blue-collar workers suffering from these afflictions are more likely to support Trump. As I review in a detailed meta-study of the available scholarly research on sources of support for Trumpism, there is very little evidence that support for Trump is seriously linked to financial insecurity, among Americans in general or among whites.
I document in Rebellion in America how occupationally stressed, xenophobic white Americans – those who report working overtime or a second job, and who agree that “immigrants exert a harmful impact on society because they take American jobs, housing, and health care” – are significantly more likely to support Trump. This finding speaks to the interplay among Trump supporters of occupational anxiety and racist scapegoating, with Americans who struggle in the neoliberal era to “get ahead” in their jobs falling victim to Trump’s race-baiting demonization of immigrants. But saying that white Trumpeters are angry because they are struggling to get ahead in their jobs is very different than saying these individuals are the same people who are being left disproportionately behind in an era of rising inequality and worker insecurity. The latter does not appear to be the case. Rather, Trump supporters are facing the same occupational stresses that most Americans face in an era in which the average worker finds him/herself working harder and more efficiently for stagnating economic returns. What separates Trump supporters from other Americans, who are also facing occupational stresses is the former’s willingness to blame immigrants for their struggles. Most Trump supporters, as I’ve statistically documented in previous writing, would rather embrace xenophobic attacks on poor immigrants of color than take a hard look at Trump’s own plutocratic economic policies, which have failed to address the decline of well-paying manufacturing (or other) jobs in America, while favoring the rich over the middle-class, working-class, and poor.
Many Americans uncritically repeat Trumpian claims that his presidency represents an uprising against corporate free trade and intensifying economic insecurity among the “white working class.” If one hasn’t been disavowed of such propaganda after 4 years of studies and a mountain of social science data, another study probably won’t make much of a difference. Still, I took another look at the “Trump-white working class” narrative, as related to voting in 2020. NORC’s mid-to-late October (10/14-10/21) national survey provides an opportunity to reassess these claims, as related to the racial and economic complexion of Trump voters.
I looked specifically in the survey at white Americans from households earning less than $50,000 a year, and who said that they were planning on voting for Donald Trump. Utilizing statistical regression analysis, I controlled for respondents’ gender, age, education, political party affiliation, and self-declared ideology (conservative, moderate, liberal), to measure whether being white and working-class (earning less than $50,000 a year) was associated with an increased likelihood of voting for Trump. Only 41 percent of white working-class voters indicated they were planning on voting for Trump, with another 42 percent planning on voting for Biden. This group was equally split in terms of their electoral preferences; they do not cut heavily, or even meaningfully in favor of Trump. As my regression analysis shows, being white and earning a lower income is not significantly associated with an increased statistical likelihood of voting for Trump, after controlling for the other demographic variables that were discussed above.
Furthermore, the evidence from the NORC survey suggests that the “white working-class” narrative is radically oversold in relation to Trump’s electoral fortunes. Only eight percent of American adults can be classified as white Trump voters who have never attended college or vocational/technical school (“high school-only educated”). And white, high school-only educated Trump supporters make up just a quarter of all Trump voters (24 percent). Additionally, white high school-only educated Trump supporters make up only 39 percent of all whites who are high school-only educated. Most whites who are high school-only educated – 61 percent – did not vote for Trump in 2020. There is little here to suggest that less educated whites are uniformly (or even primarily) pro-Trump, even if a sizable minority of this group voted for him in 2016 and 2020.
A Democratic Mandate in 2021?
A serious engagement with the available evidence reveals that economic and racial anxieties in recent years, and particularly in 2020, were strong factors pushing voters away from Donald Trump. But we shouldn’t read these statistics as suggesting that the 2020 election represents a strong mandate for Democratic governance. The Democratic Party remains firmly in the hands of establishment neoliberal figures, including Joe Biden, Chuck Schumer, and Nancy Pelosi. Bernie Sanders’ insurgent 2020 primary campaign made significant strides in capturing grassroots support from within the party, although Sanders failed to capture more than a third of the total delegate count before conceding to Biden. Concerning Biden himself, I don’t want to suggest that his victory represents an embrace of neoliberal politics. Seldom did I encounter voters in 2020 who had any illusions about Biden being a progressive, or who were committed to the notion that, if left to his own devices, he would pursue a progressive political-left agenda.
Most Democratic voters chose Biden without any starry-eyed illusions. Sixty-eight percent of Biden voters indicated in Edison’s 2020 exit polling that their vote was more of a vote “against his opponent” than a vote “for” Biden. Similarly, pre-election polling from the Pew Research Center found that 56 percent of those who leaned toward or planned on voting for Biden said that the “main reason” they supported him was because “he is not Trump.” These statistics suggest that the Democratic Party remains in a precarious position moving into 2021, with most of Biden’s “supporters” retaining a weak commitment to his presidency.
Donald Trump’s defeat means the beating back of reactionary forces across America. My hope is that it will be accompanied by successful efforts to roll back militant contempt for medicine and science and by an undermining of the rampant paranoia and conspiratorialism that took hold under Trump’s presidency. His defeat should help to challenge the hateful reactionary socio-cultural beliefs that drive Trumpism, although bigotry will no doubt persist in American political culture. Trump’s defeat also represents a decisive rejection of the pathological lack of concern or empathy for the hundreds of thousands who have needlessly died due to the government’s refusal to manage the Covid-19 crisis. These deaths fall not only on Trump’s doorstep and that of Republican governors, who elevated the “freedom” to spread a deadly virus over the right to not be infected. These deaths also fall at the feet of the 71 million Americans who were happy to allow Trump’s Covid-madness to continue for a second term. This is something we shouldn’t forget as Trump fades into the background of American politics and political discourse.
Despite the benefits that will come with a return to minimal sanity in post-Trump America, we should be reluctant to celebrate Biden’s victory. As has always been the case, progressive mass movements, not the Democratic Party, will need to be the primary engine in achieving democratic political change. As the saying goes, democracy is not a spectator sport, and if we pretend that our jobs are done after November 3rd, there’s little reason to think that the erosion of mass pressure will result in the dramatic political changes necessary to address the climate crisis, to combat racial injustice, to reduce record economic inequality, or to mount an effective campaign against plutocratic corporate power in Washington.