More than a year after the 2016 election, progressive analysis and strategy continue to be limited by the ping and pong of class-not-race and race-not-class accounts, and recriminations they provoke. Understanding what happened and charting a way forward require an alternative, a thoroughly intersectional analysis of race and class. On such a view, taking race seriously is necessary to understand how class works, not to diminish its importance.
“Intersectionality” risks depletion with its rise as a buzzword, but I mean to invoke specific insights animating the pathbreaking work of Kimberlé Crenshaw and other feminist scholars of color. In particular, they argued that understanding race and racial oppression requires an analysis of how race is gendered and gender is racialized. As Sarah Haley argues in a recent tour de force in this tradition, “gender is constructed by and through race.” So, too, we cannot understand and respond to the racism on display in the 2016 election and since without understanding its intersection with class, and how class is constructed by and through race.
Unfortunately, since soon after the votes were tallied, a legion of commentators have offered a class-not-race analysis of the election and how progressives should respond. The general thrust is that the white working class has been “left behind” by neoliberalism, and that this, as opposed to racism (or, more subtly, sublimated into racism), explains its shift in allegiance toward Trump. This thinking was widely seen as underlying Congressional Democrats’ unveiling this July of an economically-focused platform under the slogan “A Better Deal: Better Jobs, Better Wages, Better Future.”
Unfortunately, the structural flaw in such class-not-race analyses is often obscured, or even reproduced, in critiques that counter with race-not-class. Take, for instance, the headline “Yep, Race Really Did Trump Economics: Still Think Economic Anxiety and Trade Policy Were the Big Factors In Trump’s Win? Data Suggests Otherwise.” This makes sense only if “race” and “economics” are separate, competing systems. In this familiar vein, “class” is understood as “economic,” and “the economy” is treated as a freestanding, self-contained system that sits apart from racial politics, from the family, from so many things deemed “noneconomic” in their underlying structure. Within such a framework, class and race are separate and competing systems of social organization, and so the way to reject class-not-race becomes the assertion of race-not-class: “It isn’t the economy. It’s the racism, stupid.” Proof that race matters, a lot, gets interpreted as proof that class does not.
This separate spheres framework is terribly, dangerously wrong here, as it is in so many different contexts in which “the economy” is set apart from other domains of social life. We must learn to reject the entire framework, not just its specific manifestations in class-not-race or race-not-class.
As always when discussing racial capitalism in the United States, slavery provides a useful compass. Slavery was (among other things) a system of brutal economic exploitation and accumulation, one organized by racism. The thoroughness of its racism neither contradicts its economic character nor is reducible to it.
Indeed, separating race and the economy itself does racist work. It functions to insulate the so-called market economy from racial critique, and to confine matters of racial equality to other spheres. Vice versa, it obscures economic injuries of deprivation, of exploitation, of dispossession when they can only be seen clearly with a racial analysis.
Consider, for instance, insights from Michelle Alexander, Bruce Western, and others showing how racialized mass incarceration produces racial stratification in labor and housing markets, via criminal records exclusions and otherwise. These are not merely effects on economic domains. Rather, as Donna Murch’s analysis of criminal justice debt makes clear, these involve the construction of racial inequality with economic tools themselves justified with racism.
Vice versa, scholars and advocates have been showing how economic inequality yields racially specific vulnerability. For instance, people of color get racially profiled, stopped by the police, and then fined. The fines cannot be paid because of lack of income, assets, and credit. At the next stop, they now are vulnerable to arrest because of an outstanding warrant, and, if the encounter deteriorates, to getting shot and possibly killed, again mediated by racialized attributions of uncooperativeness and threat. These kinds of race/class interactions have been ubiquitous among high profile police killings of working class Black men such as Eric Garner, Walter Scott, and Alton Sterling.
Returning to the election post-mortem itself, how can we look through an intersectional lens at the data driving the race/class conversation? A good starting point is the strong evidence of variation among white people in levels of support for Trump versus Clinton. This is true both by gender and by class, with all the attendant complexities about the use of income, educational attainment, or other proxies that cannot be resolved here. Here are some of the key exit poll results that drove the immediate post-election debate:
The usual conceit here is that looking at differences among white people somehow factors out race to reveal the independent influence of class or gender. So we see class (not race) operating in the large swing between white college graduates (48% for Trump) versus whites with no college degree (66% for Trump). Intersectionality, though, teaches us that class, gender, or race often are not “independent” influences in this way. For example, if class were independent of race, then class differences among people of color should operate like class differences among whites. But these data show that working class people of color turned out overwhelmingly for Clinton. Indeed, they supported Clinton at slightly higher rates (76%) than those with college degrees (72%), a “class” effect in the opposite direction than for whites. This simple analytical point shows that class did not operate independently of race. It recalls Stuart Hall’s insight that “race is the modality in which class is lived.”
Thus, drawing conclusions about class from data about white people ignores working class people of color and consigns them to pure “race” voters. It reproduces a racially unmarked but implicitly white concept of the working class, one that is especially problematic against the backdrop of racially stratified stereotypes of who is and is not “hard working.” Using, as has become common, the racially qualified term “white working class” is an improvement, but too often it goes without an accompanying analysis of how whiteness shapes class among white people nor how class matters to people of color.
Limiting class analysis to white people erases issues tremendously important to the material circumstances of people of color. Generations of research detail the economic devastation inflicted on Black communities from the decline in manufacturing jobs. Likewise, the more recent assault on public employment and public employee unions has dislodged critical footholds in middle-class jobs, especially among women of color. Treating these as “economic” issues, and therefore outside questions of racial justice, dangerously circumscribes racial critique. It also insulates those perpetuating such neoliberal policies from accountability in racial justice terms.
Identifying the importance of race in these ways provides no reason to ignore class. Instead, we should analyze class differently. Substituting a race-not-class approach requires homogenizing race in ways that simply ignore the data showing substantial differences among groups of white people. These differences show that race is not operating independently of and dominant over all other factors. To be sure, Trump commanded majority or near-majority support across all white sub-groups, so the point cannot be to let any of them off the hook. Nonetheless, no thorough analysis of what drives political response can afford to ignore differences as large as the swing of more than 50 points between white, female college graduates (who favored Clinton by 7%) and white, male nongraduates (who favored Trump by 48%).
Those doubting the relevance of class often counter this education-level data by pointing out that voters making under $50,000 favored Clinton, seemingly contradicting the idea of Trump’s distinctive (white) working class appeal. That conclusion is misleading, however, because women and people of color are vastly overrepresented among low-income households. If, however, one conducts the intra-race comparison for income analogous to the comparison for education level above, the data on voting by income bracket among whites largely tracks the pattern for education. That is true with regard both to 2016 itself and to the changes over time discussed below.
A historical perspective both reinforces and complicates this call to consider intra-group differences. Part of the challenge is to understand what was distinctive about Trump’s election relative to past Republican candidates, or at least how it intensified rather than simply reproduced earlier patterns. In 2016, the astounding 35% gap between white college graduates and nongraduates was three times larger than the analogous gap in the George W. Bush and Obama elections, a shift also consistent with analyses of county-level voting patterns. Furthermore, as recently as Bill Clinton’s two elections, whites without college degrees were slightly more supportive of Democratic candidates than were college educated whites. Since then, working class whites have migrated steadily toward Republican candidates while college-educated whites have remained at best evenly divided, with occasional swings further right but never further left. Similarly, 2016 marked the first time since measurement began in 1948 that the highest income whites tilted Democratic relative to the electorate as a whole, and the first time since 1968 the bottom third of the white income distribution tilted Republican. In notable contrast, the 2016 gender gap, while substantial, was much closer to what it had been in 2012 and as far back as 1980.
Crucially, these class differences should not be conceptualized as independent of race merely because they arise among white people. To the contrary, an essential insight of intersectionality is that intra-group differences themselves can have racial content. This point has been best developed with regard to differences among people of color, but it applies equally to white people. Intersectional analysis of race/gender emphasizes that femininity and masculinity are racialized, such that gender relations and differences between women and men of color are constituted by rather than independent of racial specificity. Dorothy Roberts has long argued this point with regard to Black motherhood. Similarly, feminist scholarship long has explored notions of multiple masculinities and femininities, including those differentiated by class, such that differences among women may reflect differences in, and sometimes conflict over, how to be a woman. Among white women, such differences typically are left racially unmarked, and so Devon Carbado has brought all these strands together through his analysis of “colorblind intersectionality” with regard to the performance of (white) femininity.
How might Carbado’s critique of colorblind intersectionality extend to the racially intersectional character of class difference among white people? With regard to the white working class, Cheryl Harris provided the template over twenty years ago in her classic “Whiteness as Property.” The baseline against which working class whites have “fallen behind” is one that normalizes white working class racial privilege and renders invisible its racial content. What has been lost, in other words, is in part a racial entitlement to economic status, and to a broader social status articulated as the citizenship claim of the “hard-working” white patriarch: the material and social “wages of whiteness” in the tradition of DuBois. The New York Times perfectly, but unwittingly, illustrated this point with an article lamenting that whites have lagged behind other racial groups in recent job gains. No mention went to the facts that, despite this, white people continue to earn higher wages and have lower unemployment rates than Blacks and Latinos. This is the racist logic by which equalization—especially through levelling down—can spawn chants of “you will not replace us.”
Importantly, this racial understanding of white working class resentment can enrich rather than displace recognition of class grievance. The point is that class grievance may itself be racialized, albeit with that racial component often obscured by colorblind ideology. For instance, white working class resentment of white elites may reflect the accusation that those elites have expropriated or redistributed the wages of whiteness that working class whites expect as their due, a betrayal of white racial solidarity articulated through nationalism.
In this way, the racially intersectional character of class does not disappear as we travel up the class hierarchy. It changes form while remaining in relation to working class whites. To be sure, more class-privileged whites were less supportive of Trump. More revealingly, they were less supportive of Trump than of Romney and Bush II. Obviously, a class-not-race analysis cannot explain why whites outside the working class supported Trump overall, even if not to the same degree as working class whites. At a minimum, Trump’s overt appeals to white supremacy were no deal-breaker for white elites.
But the failure of class-not-race is no warrant for race-not-class instead. Superficially, that stance seems to put all white people equally on the hook. It does have the virtue of highlighting the racism that flourishes among white elites, as exemplified by Richard Spencer and others. But it can also have the opposite effect. This occurs when reconciling race-not-class analysis with the evidence of intra-white class variation in intensity of support for Trump. If white working class people were more supportive of Trump, but because of racism and not economic grievances, then that implies that racism, while widespread, is less stubbornly ingrained in college-educated whites and is more concentrated among working class whites.
Class, in other words, returns, not as the fuel for racism but as its (partial) antidote. This is a particularly dangerous conclusion because it fuses nominally anti-racist politics with class snobbery and self-congratulation among the college-educated whites who dominate media and academic production. A race-not-class analysis simultaneously absolves white elites of responsibility for class inequality—which is ignored or trivialized—and renders opposition to Trump a decisive anti-racist credential. So (white) Wall Street Democrats get to be progressive heroes while (white) working-class labor leaders seem suspect.
If we approach class and race intersectionally, however, then the class position of college-educated whites may have specifically racial content, just as it does for working class whites. Rather than being less racist as a race/class, they might simply have perceived Clinton as not particularly threatening to the specific forms of white privilege that accrue higher up the class hierarchy, or Trump as not particularly protective of them. In other words, the “wages of whiteness” might have been no less important to college-educated whites, but merely less imperiled. Just as we can recognize how both coal miners and bankers are masculinized (and male dominated) occupations, yet in different and often competing ways, the same is true for their whiteness, and a fully intersectional account would integrate race, class, and gender here.
This theoretical possibility coheres with how whites higher up the class hierarchy have had more to gain from neoliberalism’s regressive tax and transfer policy, deregulation, and “great risk shift” onto individuals and away from employers and the state. These are the people declared to be winners in the supposedly meritocratic game of free markets that reward the accumulation of “human capital.” Indeed, part of the meritocratic ideology associated with markets is precisely the supposed irrelevance of race and other forms of difference.
Even while wrapping themselves in the mantle of colorblindness, however, these systems are parasitic on and productive of racism. Part of neoliberalism’s justificatory structure, and its method of enforcement, rests on the racial disparagement of people of color. For instance, racial attributions of mendacity, laziness, and sexual irresponsibility have provided essential tools for delegitimizing the welfare state. So, too, have they supplied racial distinction to those who (often falsely) imagine themselves as self-sufficient. Similarly, racial tropes of criminality and dangerousness have been deployed to rationalize the social disorder produced by neoliberalism as something other than evidence of its failings. These are phenomena to which mainstream Democratic politics have often contributed, even if not with the same gusto as those even further to the right.
Keeping neoliberalism on the hook is what is at stake in rejecting race-not-class. But to do so without invoking class-not-race requires caution. In particular, it requires avoiding the facile notion that a politics attentive to class provides a simple platform for cross-racial solidarity in the name of common economic interests. That vision of social democracy almost always is cast in the colorblind language of universalism—good jobs for all, good schools for all, a secure retirement for all, and so on. And such calls to “place economics at the center” are inevitably coupled with warnings against being “obsessed with cultural issues” sounding in race, gender, and sexuality. Two domains of politics, separate and unequal.
Dropping racial critique in favor of a colorblind universalism necessarily fails to recognize and attack the racialized manifestations of class for people of color. The stubborn truth is that even when nominally universal systems are formally racially inclusive, they are vulnerable to being undone by those seemingly separate “noneconomic” systems. For all the breathless attention to “precarious work” and economic insecurity in the “gig economy,” what is the full range of threats to secure work? A predictable schedule and protection from layoff don’t go very far if every drive to work risks missing my shift—and getting fired—if I get stopped by the police and arrested or worse. Sandra Bland never got to start the job that had brought her to Prairie View, Texas. The pinnacle of universalism, a guaranteed basic income, starts to lose its sheen when some people’s grant gets taxed back through racially targeted criminal justice debt.
Worse yet, there are reasons to doubt that genuinely universal economic security policies can secure durable and broad-based white working class support. Although universal systems often have benefitted people of color relative to overt systems of white supremacy like Jim Crow, their political success frequently has been mortgaged to a racialized narrative of white deservingness. That foundation intrinsically limits universalist aspirations even while masking those failures in assertions of colorblind eligibility rules. A perfect example is Social Security, the usual paradigm of politically successful universalism. Although a great improvement on nothing at all, it nonetheless is deeply structured by privileges afforded to marriage and formal employment, both of which simultaneously constitute the system’s appeal to moral deservingness and inscribe race and gender hierarchy. So, too, with the minimum wage and labor law, once one takes seriously their history and patterns of partial coverage, from the New Deal agricultural and domestic worker exclusions to today’s conflicts over home care workers, young workers, and underenforcement in the informal sector, let alone labor coerced through incarceration or its threat.
This analysis has an ineradicably pessimistic strand relative to a class-not-race program that imagines the 99% rising up and overwhelming the 1% in a triumph of majoritarianism. Taking racism seriously means tempering our confidence in democracy. But taking class seriously offers a ray of hope, especially in a world where fairly modest shifts in voting patterns or demographics could produce sharply different electoral outcomes.
Our best hope lies neither with colorblind universalism nor with multicultural neoliberalism but instead with approaches that fuse race consciousness with broad reach. Jesse Jackson’s success with the Rust Belt’s white working class in the 1980s offers one model, as does the synergy in recent decades between movements for immigrants and low-wage workers, a synergy built on recognizing how employers exploit immigration enforcement. Most ambitiously, the Vision for Black Lives provides an inspirational and comprehensive roadmap animated by an intersectional spirit, one reflected in both its comprehensive scope and its innovative proposals like integrating race-conscious reparations with a universal basic income. Hopefully, we will get another chance, or rather, we will create one.