In the aftermath of the 2016 election, there have been hundreds of reflections written on the behavior, attitudes, needs, and prospects of the “white working class,” a segment of the population that will prove vital to any progressive coalition that stands for both social and economic justice. However, within most analyses of the category, we find two stereotypes that seem neither true nor useful for building such a movement: either the working class is imagined as “white men,” or it is understood as “the poor,” identified as minorities and women.
Income and Education, Inexact Proxies for Class
Rather than speaking of the working class in terms that reduce it to only categories of race and gender, the emerging field of working-class studies is settling on definitions of class that are rooted in the power relations established in production, extending outward into politics and culture.  In this view, working-class people are in jobs where they have little or no control over the pace and content of their work—traditional “blue collar” goods producers in factories, mines, and construction sites; and also “white collar” service workers in call-centers, retail sales staff, most nurses and teachers, medical and lab technicians, and the like. Using occupation as the best indicator of class standing, I have found that the working class constitutes roughly 63 percent of the U.S. labor force. Crucially, it consists of both men and women and is multiracial and multiethnic.  White people are, of course, a big part of the working class, but if we settle on “the white working class” as a class in itself, and with the force of white supremacy, even a class for itself, we lose track of the role blacks, Native Americans, Hispanics, Asians, and other non-whites play in the working class. We also obscure the mutual interests that white workers have with their non-white coworkers.
Despite the importance of occupation in understanding the working class, voting surveys typically ignore it. Instead, to identify class, they sort respondents by income or education. Members of the working class are usually identified as people with low incomes (how low is arbitrary, but in most post-2016 election analysis, it is taken to be less than $50,000 a year), or without a college degree. Using income or education instead of power to establish the working class obscures the most important element class shares with race and gender: all are manifestations of power relations. This common denominator allows us to investigate and appreciate the ways in which class differences complicate race and gender experience and identity, just as differences in race and gender cross-cut and complicate class. Whether considering voting patterns or any other social phenomenon, the intersections among race, class, and gender need to be considered in the analysis, as well as the particularity of each.
Another reason income and education are poor proxies for class is that many working-class households, especially among skilled and unionized occupations, make more the $50,000 a year (although it is true that most households making less than $50,000 are working class). Also, many members of working-class households have college degrees to qualify them for their skilled jobs in education, medicine, science, and robotics. Using income or education to identify class because that is where the data are is like the guy who lost his keys in the alley at night but looks for them under the streetlight in front of the house because that is where the light is.
Trump’s Appeal to White Working-Class Voters
Despite these weaknesses in the categories used in exit poll surveys, it has been common to ascribe Trump’s victory to the reactionary character of the white working class. As Paul Krugman wrote some weeks after the election, “thanks to overwhelming support from white working-class voters,” Donald Trump was able to get close enough to victory that meddling by Vladimir Putin and FBI Director James Comey could put him into the White House.  This diagnosis leads to two opposite paths forward for progressives. One path essentially gives up on the white working class as hopelessly backward, urging instead a double-down on the mobilization of the various “identity” constituency groups that have organized themselves over the past fifty years and have become the current base of the Democratic Party. The other view challenges the Democratic Party to go back to its traditional working-class base and win it, meaning the “white working class,” back into the fold.
To get to the bottom of this complexity, it is important to examine voting patterns in those counties, especially in rural areas, that went for Barack Obama in 2012 but flipped last year to Trump. There are anecdotal reports of white workers in Ohio and other Rust Belt areas who flipped their votes.  Politico reported significant falloffs in union household support for the Democrats in swing states.
Obama won Ohio in 2012, besting Romney in those households by 23 percentage points. Clinton actually lost Ohio’s union households to Trump by 9 points. Exit polls show[ed] Clinton h[e]ld only a 13 percent advantage among union households [in Michigan, the day after the election, when Michigan had yet to be decided]. Obama beat Romney in union households [there] by . . . 33 percent in 2012. . . . [However] exit polls showed Trump’s share of union households only rose by 3 percent [nationally] compared with Romney in 2012. 
The significant erosion of union support for the Democratic Party in Michigan and Ohio and the modest gain in national union support for Trump reflect the close relation between Trump’s basic economic message and the AFL-CIO’s long-standing priorities: end the TransPacific Partnership (TPP), build new infrastructure, and create millions of good-paying jobs, especially in manufacturing. Clinton’s history of support for the TPP, in her pre-campaign words the “gold standard” of trade agreements, and her unfortunately expressed but truthful observation in a West Virginia forum that “we’re going to put a lot of coal companies and coal miners out of business,” made her a suspect champion on these issues. Her pledge in the same remarks to create alternative jobs in clean energy did not save her in coal country.
But changes in voter turnout rates were more important than vote flipping in explaining the change from Democrat to Republican majorities in rural counties. As reported in Slate,
compared with Republicans’ performance in 2012, the GOP in the Rust Belt 5 [Iowa, Michigan, Ohio, Pennsylvania, and Wisconsin] picked up 335,000 additional voters who earned less than $50,000 (+10.6 percent). . . . compared with the Democrats’ loss of 1.17 million voters (−21.7 percent) in the same income category. Likewise, Republicans picked up . . . 26,000 new voters in the $50-$100k bracket (+0.7 percent) but Democrats lost 379,000 voters in the same bracket (−11.7 percent).
Where data were available, similar patterns emerged for white as well as black and Hispanic voters.
Democratic support in the Rust Belt collapsed as a huge number of Democrats stayed home or (to a lesser degree) voted for third-party candidates. Trump did not really flip white working-class voters in the Rust Belt. Mostly the Democrats lost them.
Political strategists will argue over the weaknesses of the Clinton campaign for some time. Perhaps the decision to emphasize Trump’s negatives rather than highlighting Clinton’s positive platform contributed to her defeat. Clinton and her surrogates constantly pointed to Trump’s racism, misogyny, and general unfitness for office while directing people to her website to find her detailed program on dozens of issues. After her loss to Bernie Sanders in the April Wisconsin primary, candidate Clinton never returned to the state during the main campaign. Her trips to Michigan were limited to two stops in Flint to express support for clean public water. Incredibly, during the campaign, she spent more money advertising in Omaha than she spent in Michigan and Wisconsin combined.
Whatever the flaws in the Clinton campaign, it is certainly a serious concern that millions of white workers did vote for Trump. In total, 42 percent of people in union households (not the same as union members) voted for Trump. The American Federation of Teachers estimates that Clinton failed to get the votes of 20 percent of its members, while the National Education Association’s internal polls showed that only 65 percent of their members supported Clinton.
While these are better outcomes than for the population as a whole, union leaders and progressive activists need to confront the reality of Trump support among white people, and white workers in particular. Here, we can look at specific conditions white workers faced in the period leading up to Trump’s election, and observe their voting behavior in the historical arc dating from the Civil Rights and Voting Rights Acts of 1964 and 1965.
Job Loss and White Working-Class Trump Support
Two examples serve to represent the suffering recently experienced in white working-class communities, especially in rural areas. One of the most alarming signs is the increase in mortality rates and lower life expectancy for white middle-aged men, reflecting increased suicide rates and fatal opioid drug overdoses.10 These phenomena were centered in areas most hard hit by deindustrialization in the late 1980s and 1990s, but did not emerge until the period following the Great Recession of 2008-2009.
Journalist Eduardo Porter has presented a second circumstance that contributed to Trump’s support among rural whites. Matching the geography of job creation with the geography of Trump support, Porter reported that in the years following the Great Recession,
nearly all the gains from the economic recovery have passed them [white people] by. . . . Hispanics got more than half the net additional jobs. Blacks and Asians also gained more jobs than they lost. But whites . . . lost more than 700,000 net jobs over the nine years.
Because white people live disproportionately in rural areas while minorities are disproportionately urban, this difference in job experience suggests a basis for discontent, even anger, among whites in rural areas who, consequently, voted for Trump out of frustration and hope for something better. The frustration and pain of hard rural living with no relief in sight surely led many people to look for something different, some change, and to settle on Donald Trump. The fact that Hillary Clinton won nearly three million more votes than Trump but still lost the presidency is testament to the outsized power of relatively less populated rural areas in the American electoral college system.
Racism and the White Working-Class Vote
Still, the question remains: Why does the economic hardship of white voters lead so many to vote for Republicans, let alone a Donald Trump, while the demonstrably greater economic and social hardship in communities of color does not. Exit polls in 2016 indicate that 66 percent of (non-Hispanic) whites without a college degree voted for Trump while only 20 percent of nonwhites without a degree did. Trump won 57 percent of all white votes but only 21 percent of nonwhite votes. There must be something in being white that explains this. The long legacy of racism leads the list.
We know from the days of Richard Nixon’s law and order campaign and “Southern Strategy” in 1968, following the passage of the Civil Rights and Voting Rights Acts, that appeals to white racism have drawn traditional Democratic voters into the Republican camp, not only in the southern states of the old Confederacy, but also in the rest of the country. The 1980 presidential election brought forward “Reagan Democrats,” including the 45 percent of union households who voted for him—a stronger union showing than Trump’s 42 percent.
In 2016, Trump dominated among those concerned mainly with immigration and terrorism (31 percent of respondents), contemporary arenas where racism and xenophobia are playing powerful roles. And while the exit polls did not directly pose the question, we can be sure that Trump’s echo of Nixon’s cry for a pro-police “law and order” agenda and Trump’s open disdain for the Black Lives Matter movement also drew millions of white votes. We have no data focused solely on white workers in this regard, but anecdotal reports from unions across the country suggest that many of them responded positively to that call.
A Progressive Coalition Absent the White Working Class?
One reaction among progressives has been to write off the white working class, or whites more generally, viewing them as hopelessly reactionary. This path relies on strengthening the current Democratic Party electoral coalition. One example emerged in a November 21, 2016 Facebook post from faculty and students at the State University of New York at Stony Brook who declared themselves,
actively committed to protecting and supporting populations now targeted by the incoming federal administration. . . . We celebrate the vibrant diversity of the state of New York: African Americans, Latinos/as, Asian Americans and Pacific Islanders, Indigenous people, people of color, LGBTQ [lesbian, gay, bisexual, transsexual, queer] communities, Muslims and other religious minorities, immigrants, undocumented students, disabled people and women.
Notice that the “vibrant diversity” celebrated in this list omits white people. But the list does reflect the “demographic destiny” that is leading the United States to be a country in which white people are the minority. (For discussion of projected demographic changes in the U.S. electorate, see the exchange between Cristina Mora/Michael Rodríguez Muñiz and Richard Alba in this issue of New Labor Forum.) That future is an attractive political base for those who are frustrated with the “white working class,” who seemingly can be excluded from the progressive mix, except insofar as they support nonwhite constituencies.
The disdain for white workers has been astutely analyzed and sharply criticized by white commentators from working-class backgrounds. Sarah Smarsh, writing in The Guardian nearly a month before the election, pointed to the condescension with which media, political, and academic elites characterize white workers. “The faces journalists train the cameras on—hateful ones screaming sexist vitriol next to Confederate flags—must receive coverage,” she writes,
but do not speak for the communities I know well. . . . Countless images of working-class progressives . . . are thus rendered invisible by a ratings-fixated media. . . . This year  more Kansans caucused for Bernie Sanders than for Donald Trump . . . The two-fold myth about the white working class—that they are to blame for Trump’s rise and that those among them who support him for the worst reasons exemplify the rest—takes flight on the wings of moral superiority affluent Americans often pin upon themselves. . . . It wasn’t poor whites who criminalized blackness by way of marijuana laws and the “war on drugs.” Nor was it poor whites who conjured the specter of the black “welfare queen.”
The need for a united working class as an organized progressive political force cannot be overemphasized. Whatever “populist” aura the Trump campaign might have had, the billionaire corporate leaders he has brought into his cabinet are dedicated to undermining the very missions of the education, health, environmental, housing, labor, consumer, and financial regulation institutions they oversee. They are already promoting disastrous policies for working people of all races, genders, and nationalities, in the United States and across the globe. Ignoring or belittling white workers or excluding them from the progressive coalition required to resist effectively the reactionary onslaught coming our way is a recipe for failure, a capitulation to the divide-and-conquer strategy that dates all the way back to what Ted Allen called “the invention of the white race” as a counterpart to the imposition of racial slavery in colonial America.
While some liberals have given up on white workers, whom they see as unrepentant Archie Bunkers, others have been wrestling with the task of winning them back into the Democratic Party fold. Unfortunately, carrying forward much of America’s populist tradition, this has too often involved disregard for the interests of black people and others in “identity” movements. In the current moment, it is demonstrated (in implicit fashion as is so often the case) in a Politico report on the contest for chair of the Democratic National Committee taking shape at the end of 2016:
“We like Keith [Ellison],” one longtime Obama political ally, who was pushing [Labor Secretary Tom] Perez [for DNC Chair], told [the writer] in November. “But is he really the guy we need right now [an African-American Muslim] when we are trying to get all of those disaffected white working-class people to rally around our message of economic equality?”
Columbia University professor Mark Lilla offered advice in the same mistaken direction after Trump’s election:
In recent years American liberalism has slipped into a kind of moral panic about racial, gender, and sexual identity that has distorted liberalism’s message and prevented it from becoming a unifying force capable of governing. . . . We need a post-identity liberalism, and it should draw from the past successes of preidentity liberalism. 
Andrew Levison, author of The Working-Class Majority, writing to the Democratic National Committee in 1974, was an early proponent of this one-sided strategy. While acknowledging “deep currents of racism in working-class America,” he advised,
The “drift to the right” that the vote for [George] Wallace and Nixon supposedly demonstrated was in significant part the result of a basic error in liberal strategy. . . . It is worth considering how different the situation would be today if, in place of that long list of Democratic-sponsored social programs of the sixties, all of which antagonized workers, there had been instead . . . liberal proposals which championed workers’ real and unmet needs and indicated a genuine concern for their interests. 
Clearly, Levison here limits his meaning of “workers” to white workers “antagonized” by the civil rights and women’s movements.
Reducing the working class to white men carries forward the racist, nativist, and male chauvinist history of the labor movement itself. From their beginnings in the late nineteenth century until the last few decades, the practices and documents of almost all the craft unions comprising the American Federation of Labor systematically excluded blacks, women, and immigrants from their ranks, and their agendas of concern. The proud anti-racist history in many of the industrial unions of the CIO notwithstanding, in today’s popular formulations, minorities and women who work are. . .minorities and women, not workers.
The “pre-identity liberalism” that Lilla and others now promote would again systematically exclude black people from progress and marginalize women from social advances available to white men. Virtually every New Deal victory—including social security, minimum wage, fair-labor standards, union protections— was effectively denied to black workers by excluding agricultural and household labor to win over the Dixiecrats of the plantation South, who only pretended to represent the interests of their white working-class constituents.
We do not need a “new New Deal” that reproduces that history in modern attire. As we go further into the Trump years, we must refuse every attempt to subordinate the needs and movements of blacks and others in the name of unity, a process that has historically been the Achilles heel of all American populist uprisings.
The Trump agenda is a corporate agenda intensely hostile to working people. Responding to it with any hope of success will require a broadly unified working-class movement. But the working class has different sections with different experiences and different specific needs for justice, in addition to and related to their needs as workers.
We have to learn how to build class-wide movements that promote mutually reinforcing relations with the specific movements of blacks, women, and others resisting their own oppression. Recognizing the intersectionality of our movements, we must organize in each while respecting them all.
Bringing significantly larger numbers of white workers into these movements will be essential for our ultimate success. Few Trump voters were outright white supremacists or hardened misogynists. Yet millions of white men and women who voted for him were willing to overlook Trump’s racism and misogyny, voting for some personal gain they imagined he offered them. This softer expression of racism and male chauvinism in the acceptance of others’ oppression can only be overcome through direct challenge, best expressed by workers’ immediate coworkers or union leaders in the context of the requirements of successful organizing. Richard Trumka, president of the AFL-CIO, did this in his forceful speech to the United Steelworkers convention in 2008, when he called out the racism that was keeping many in the room from voting for Barack Obama.
As we try to overcome backward and divisive attitudes anywhere in our movements, it would be wise to follow the old adage, “Hate the sin but love the sinner.” In particular, as we seek to organize white workers, we cannot do so starting with disdain. We know from countless union organizing and contract fights that material gains often count for far less than the imperative to be treated with respect and dignity. This is equally so in politics. To forge deeply rooted mutually reinforcing social movements that simultaneously champion class-wide economic goals and specific community justice goals, the required education and transformation can only happen in the course of building those movements as they should be, from the bottom up with leadership across the board committed to the mutuality that is required to achieve social and economic justice for all.
1. Michael Zweig, The Working Class Majority: America’s Best Kept Secret, 2nd ed. (Ithaca, NY: Cornell University Press/ILR Press, 2012).
2. Ibid., 29-36.
3. Paul Krugman, “And the Trade War Came,” New York Times, December 26, 2016, available at https://www.nytimes.com/2016/12/26/opinion/and-the-trade-war-came.html?_r=0.
4. Nelson D. Schwartz, “Can Trump Save Their Jobs?” New York Times, November 13, 2016, available at https://www.nytimes.com/2016/11/13/ business/economy/can-trump-save-their-jobstheyre-counting-on-it.html?_r=0; Eric Russell, “Why This Maine Town Pivoted from Obama to Trump,” Portland Press Herald, November 13, 2016, available at http://www.pressherald .com/2016/11/13/why-one-town-turned-to-trump/.
5. Ted Hessen and Marianne LeVine, “Unions Investigate Their Poor Showing for Clinton,” Politico, November 10, 2016, available at http://www.politico.com/story/2016/11/laborunions-hillary-clinton-mobilization-231223.
6. Lauren Carroll, “In Context: Hillary Clinton’s Comments about Coal Jobs,” PolitiFact, May 10, 2016, available at http://www.politifact.com/trutho-meter/article/2016/may/10/context-hillary-clintons-comments-about-coal-jobs/.
7. Konstantin Kilibarda and Dana Roithmayr, “The Myth of the Rust Belt Revolt,” Slate, December 1, 2016, available at http://www.slate.com/ articles/news_and_politics/politics/2016/12/ the_myth_of_the_rust_belt_revolt.html.
8. Nebraska awards its Electoral College votes by Congressional district; Clinton was hoping to pick up one electoral vote from the Omaha district. Author conversation with Congressman Mark Pocan, February 4, 2017; and Ken Goldstein, John McCormick, and Andre Tartar, “Candidates Make Last Ditch Ad Spending Push Across 14-State Electoral Map,” available at https://www.bloomberg.com/politics/ graphics/2016-presidential-campaign-tv-ads/.
9. Greg Toppo, “Teacher Unions Smarting after Many Members Vote for Trump,” USA Today, November 23, 2016, available at http://www .usatoday.com/story/news/2016/11/23/electionunions-teachers-clinton-trump/94242722/.
10. Gina Kolata, “Death Rates Rising for MiddleAged White Americans, Study Finds,” New York Times, November 2, 2015, available at https://www.nytimes.com/2015/11/03/health/ death-rates-rising-for-middle-aged-whiteamericans-study-finds.html?_r=0. 11. Eduardo Porter, “Where Were Trump’s Votes? Where the Jobs Weren’t,” New York Times, December 13, 2016, available at https://www .nytimes.com/2016/12/13/business/economy/ jobs-economy-voters.html.
11. Eduardo Porter, “Where Were Trump’s Votes? Where the Jobs Weren’t,” New York Times, December 13, 2016, available at https://www .nytimes.com/2016/12/13/business/economy/ jobs-economy-voters.html.
12.Roper Center, “How Groups Voted: 1976- 2012,” available at http://ropercenter.cornell .edu/polls/us-elections/how-groups-voted/.13. Posted by Michael Schwartz.
14. Sarah Smarsh, “Dangerous Idiots: How the Liberal Media Elite Failed Working-Class Americans,” The Guardian, October 13, 2016, available at https://www.theguardian.com/media/2016/oct/13/liberal-media-bias-working-class-americans. See also Joan Williams, “What So Many People Don’t Get about
the U.S. Working Class,” Harvard Business Review, November 10, 2016, available at https://hbr.org/2016/11/what-so-many-peopledont-get-about-the-u-s-working-class.
15. Theodore W. Allen, The Invention of the White Race, 2 vols. (London: Verso Books, 2012); also Bill Fletcher Jr., “How Race Enters Class in the United States,” in What’s Class Got to Do with It? American Society in the Twenty-First Century, ed. Michael Zweig (Ithaca, NY: Cornell University Press/ILR Press, 2004), 35-44.
16. Glenn Thrush, “Keith Ellison’s One-Man March,” Politico, December 20, 2016, available at http://www.politico.com/story/2016/12/ keith-ellison-democrats-dnc-232838.
17. Mark Lilla, “The End of Identity Liberalism,”New York Times, November 20, 2016, available at https://www.nytimes.com/2016/11/20/opinion/sunday/the-end-of-identity-liberalism.html?_r=0.
18. Andrew Levison, The Working Class Majority (New York: Penguin, 1974), 153, 272-73.
19. Richard Trumka, address to the United Steelworkers, July 1, 2008, available at https://www.youtube.com/watch?v=7QIGJTHdH50.