Understanding the Role of Racism in Contemporary US Public Opinion

In the contemporary context, it is inescapable that racism is a factor in US public opinion. When scholars take stock of the way we typically measure and conceptualize racism, we find reason to reconceptualize the racial resentment scale as a measure of perceptions of the reasons for political inequality. We also see reason to move beyond thinking of racism as an attitude, toward conceptualizing it as a perspective. In addition, we see reason to pay closer attention to the role of elites in creating and perpetuating a role for racism in the way people think about public affairs. The study of racism is evolving in parallel with the broader public discussion: toward a recognition of the complex and fundamental ways it is woven into US culture and political life.


INTRODUCTION
Racism is a factor in US public opinion. To some, that is obvious; to others, it is controversial and a matter of definition. But the contemporary political context has made it inescapable that racism plays a role, when racism is defined as a particular form of prejudice in which antipathy is expressed toward racial groups. The advent of cell phones and their instantaneous presentation of evidence of racist acts, white supremacist support for President Donald Trump, the attention to racial justice insisted upon by professional football players kneeling during the national anthem, and other recent events have made it impossible for the US public and scholars alike to treat racism as a minor issue.
It is time to take stock of the way political scientists measure and conceptualize racism in the study of US public opinion. When we examine its most common measurement, the racial resentment scale, in light of contemporary events such as the elections of Barack Obama and Trump to the presidency, we find reason to reconceptualize this scale as a measure of perceptions of the reasons for political inequality. We also see reason to move beyond thinking of racism as an attitude, toward conceptualizing it as a perspective. In addition, we see reason to pay closer attention to the role of elites in creating and perpetuating a role for racism in the manner in which members of the public think about public affairs. In these ways, we can see the study of racism evolving in parallel with the broader public discussion: toward a recognition of the complex and fundamental ways it is woven into the culture and political life of the United States.

THE NATURE AND ORIGINS OF THE RACIAL RESENTMENT SCALE
Much of the modern study of the role of racism in public opinion has its roots in Allport's (1954) The Nature of Prejudice. In that book, Allport gave to social science a comprehensive review and synthesis of what was known to date about ethnic prejudice, which he defined as "an antipathy based upon a faulty and inflexible generalization. It may be felt or expressed. It may be directed toward a group as a whole, or toward an individual because he is a member of that group" (Allport 1954, p. 10). Allport looked across existing studies on prejudice and provided a detailed argument about the factors that appeared to be necessary to reduce prejudice through intergroup contact, typically referred to as the contact hypothesis. That approach catalyzed a dominant understanding of racism as prejudice, in other words as a matter of psychology. It resulted in measures of racism as stereotypes against out-groups, interracial hostility, and beliefs in the innate inferiority of certain racial groups.
As the civil rights movement progressed and cultural norms of appropriateness changed, many scholars began to view racism as less blatant. Rather than defining racism as behaviors rooted in beliefs about the innate inferiority of others, scholars working in the Allport tradition argued that racism was more typically expressed as a perception that certain racial groups did not abide by norms of hard work and patriotism (see Hutchings & Wong 2014, p. 421). This newer, symbolic form of racism has been conceptualized as a set of attitudes acquired through socialization, as perceptions that people teach to each other through interpersonal interaction and learn through education, mass media, religious institutions, and other important sources of communication (Henry & Sears 2002).
In the past several decades, the dominant measure of symbolic racism in political science has been the racial resentment scale, developed for the American National Election Study (ANES) in the mid-1980s by Kinder & Sanders (1996). The standard scale consists of the following four items (survey respondents are asked whether they agree or disagree and how strongly they do so): 1. Irish, Italian, Jewish, and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors.
toward government, or attitudes toward redistributive policies (Rabinowitz et al. 2009). It is possible that proponents of these different approaches are talking past one another and each offering valid measures of racism among subsets of the population (Neblo 2009a,b). And yet, there is a wealth of evidence that the traditional racial resentment scale exhibits a relationship to many political preferences. These include evaluations of and support for political figures, including Obama (Tesler & Sears 2010, Tesler 2016; see also Berinsky et al. 2011) but also Trump (Tien 2017Setzler & Yanus 2018;Sides et al. 2018, ch. 5), especially when racial resentment is interacted with emotions toward these individuals (Redlawsk et al. 2014, Tolbert et al. 2018. There is also substantial evidence that racial resentment is related to a variety of policy stances such as those in the realm of criminal justice [e.g., support for the death penalty and beliefs about harshness of penalties given by courts (Brown & Socia 2017), use of force by police (Carter & Corra 2016), privatization of carceral institutions (Enns & Ramirez 2018) 2 , perception of crime events (Goidel et al. 2011), and juvenile rehabilitation (Pickett et al. 2014)]; gun control opinions among whites (Filindra & Kaplan 2016, O'Brien et al. 2013) and blacks (Filindra & Kaplan 2017); and even opinions on a policy not related to federal spending, paying college athletes (Wallsten et al. 2017).
Various studies have also found evidence that racial resentment is related to whites' political behavior, including activism with the Tea Party Movement (Tope et al. 2015), willingness to write to one's member of Congress about a policy that benefits racial or ethnic minorities (Hassel & Visalvanich 2015), and turnout in congressional elections (Petrow 2010).

WHAT ABOUT EXPLICIT RACISM?
When Kinder & Sanders (1996) proposed the concept and measurement of racial resentment, they argued that racial resentment was at that time the most common expression of prejudice. Recounting the history of racism in the United States, they wrote, "A new form of prejudice has come to prominence, one that is preoccupied with matters of moral character, informed by the virtues associated with the traditions of individualism. At the center are the contentions that Blacks do not try hard enough to overcome the difficulties they face and that they take what they have not earned. Today, we say, prejudice is expressed in the language of American individualism" (Kinder & Sanders 1996, pp. 105-6).
These authors were writing at a time when explicit racism had become less socially acceptable and less blatantly manipulated by political elites (Tesler 2012a, pp. 110-11), although not necessarily less expressed through various forms of discrimination (Schuman et al. 1997, pp. 39-53). The late twentieth century was also a time when political science was beginning to consider the role of implicit bias in cognition (Devine 1989). Mendelberg's (2001) award-winning book The Race Card demonstrated that because of the common role of implicit bias in human cognition, and the presence of strong norms against overt expressions of racism, political campaigns could and did subtly invoke racist imagery to influence political choices. Some have challenged this work, arguing that implicit images are no more effective than explicit images in priming racial resentment (Huber & Lapinski 2006). Methods of measuring implicit racism developed by social psychologists, such as the Implicit Association Test (Dovidio et al. 2002) and the Affect Misattribution Procedure (Payne et al. 2005), have made it clear that "[i]mplicit bias against African Americans is pervasive in American society," but, as Tesler (2016, p. 23) notes, "[t]he impact of that implicit bias on white Americans' political preferences is much less conclusive." It may be that the present context is one in which implicit bias is a weaker force in the political realm. Studies of the effect of implicit racial bias in the 2008 election using the best available data suggest little to no impact on whites' votes (Tesler 2016, p. 23; see also Ditonto et al. 2013, Kalmoe & Piston 2013. Ditonto et al. (2013) show that explicit measures of racism were far more powerful for predicting votes, with the implicit measures predicting votes only among Latinos (see also Pasek et al. 2014 for evidence of a role for explicit but not implicit racism in electoral choices in 2008 and 2012). However, other work using the Affect Misattribution Procedure finds that implicit racial bias was an important influence on individuals' perceptions of the state of the nation during the Obama presidency (Lundberg et al. 2017) and had an impact on turnout in the 2008 presidential election (Pasek et al. 2009).
Subtle and implicit racism continue to be forces in public opinion-but explicit racism continues as well. Obama's presidency did not reduce levels of prejudice, despite some early evidence to that effect (Goldman & Mutz 2014;Tesler 2016, p. 27). Much as rhetoric heightens the role of racial resentment in attitudes toward policies like welfare and social security (Winter 2006(Winter , 2008, the presence of an African-American candidate for president heightened the role of explicit racism in public opinion. Tesler (2012a) argues that "old-fashioned racism" (OFR) became an increasing force in American politics with the rise of Obama to the presidency. He conceptualizes OFR as "(1) desire for social distance between the races, (2) beliefs in the biological inferiority of Blacks, and (3) support [for] public policies insuring racial segregation and formalized discrimination" (Tesler 2012a, p. 114). His analyses of panel studies and experiments show that Obama's election increased the influence of OFR on contemporary political attitudes such as opposition to Obama, partisanship, and votes for congressional candidates (see also Knuckey & Kim 2015).
In an extension of that work, Tesler (2016) demonstrates the increased impact of racism on a wide range of opinions in the age of Obama, including evaluations of Obama's ideology, the employment rate, other political actors, partisan identity, congressional elections, and even Obama's Portuguese water dog. That work demonstrates that racism operated in the age of Obama at a variety of levels, from the subtle (measured as racial resentment) to the more explicit (measured as OFR, racial stereotypes, and antiblack affect; see also Piston 2010).
The fact that Obama's presence elevated the role of racism in evaluations of even his dog underscores what Tesler (2016) calls the spillover of racialization (see also Parker 2016). This theory holds that "cues that connect racialized public figures to specific issues and political evaluations are expected to activate racial considerations in mass opinion much the way that code words and other subtle racial cues have linked African Americans with political evaluations in prior research" (Tesler 2016, p. 32). According to the spillover theory, public opinion on policies or things connected to Obama should have become more polarized, which is what Tesler finds. The presence of an African-American president appears to have especially heightened the role of racism in public opinion in the realm of health care, Obama's signature policy initiative (Henderson & Hillygus 2011;Tesler 2012b;Lanford & Quadagno 2016;Tesler 2016, ch. 5;Grogan & Park 2017). Scholars have also observed the spillover of the effect of racial resentment onto policy attitudes with respect to the federal response to Hurricane Sandy (Sheagley et al. 2017) and climate change (Benegal 2018).
One might call the spillover of racialization a subtle effect of racism, but we might also describe it as a symptom of the pervasiveness of racism in US political culture. In this historical moment, it seems that we would do well to maintain a focus on racism in all its guises-in its implicit and subtle as well as explicit forms.
Wilson and colleagues have argued that our measure of racial resentment needs to be finetuned in the age of Obama to be more explicitly about resentment (Wilson & Davis 2011). In other words, they suggest that even this measure of subtle racism needs to recognize the blatant nature of resentment. "[W]e conceptualize racial resentment toward African Americans as an irritation over the use of race as a justification for prerogatives, rather than the more controversial conceptualization emphasizing an explicit blend of anti-Black affect and deeply held values" (Wilson & King-Meadows 2016, p. 37; see also Wilson & Brewer 2013, Wilson et al. 2015, Wilson & Davis 2018. In a review of racism studies, Feldman & Huddy (2009) argue that the attention to symbolic and implicit racism may have drawn attention away from explicit or blatant racism too soon. That warning appears to be prescient in our current context. Today, racism is often not at the low simmer of resentment but at the rolling boil of anger. Especially among conservative whites, attitudes of racial resentment are often rooted in anger in the contemporary context (Banks & Valentino 2012). In other words, the changing landscape of American politics and scholarship on American political behavior suggests that the main question is no longer if racism matters, but how it matters.

THE EVOLVING MEANING OF THE RACIAL RESENTMENT SCALE
The ongoing debate in the field of political science over whether the racial resentment measure taps racism or instead individualism has tugged scholars toward a recognition that these items are best conceptualized as measures of whether people blame individual or system failings for racial inequality. Kam & Burge (2018) review open-ended questions asked of white and black survey respondents about each item on the racial resentment scale to examine whether it is "prejudice or principle" that people are considering when responding to each item. They conclude that with respect to responses on the racial resentment scale, "there is no evidence here that race is divorced from principles" (Kam & Burge 2018, p. 318). In other words, they find no reason to conclude that racially resentful responses are driven by adherence to individualism alone. Instead, they argue that the most racially resentful people blame racial inequality on "disposition, motivation, or individual choice, cling fiercely to the promise of individualism, and deny or diminish the existence of discrimination in the country" (Kam & Burge 2018, pp. 318-19). On the other end of the scale, the whites and blacks who are least racially resentful "interpret differences across blacks and whites less by pointing to the character of black Americans and more by identifying structural features of discrimination that undercut the promise of individualism" (Kam & Burge 2018, p. 319). The people scoring on opposite ends of the racial resentment scale appear to be blaming racial inequality on different things. The authors therefore argue that we should reconceptualize the scale and rename it the Structural versus Individual Attributions for Black Americans' Economic and Social Status scale.
We can see a similar move in the extension of Tesler's (2012aTesler's ( ,b, 2016 work to the 2016 presidential campaign (Sides et al. 2018). He and his coauthors pay close attention to the role of racism in the 2016 election, using the racial resentment scale as well as other measures. To be understood by a general audience, they label the standard four-item ANES battery "views of racial inequality." Following Kam & Burge (2018), they refer to this as a measure of whether people attribute blame to a lack of individual initiative by blacks or to systematic discrimination (Sides et al. 2018, p. 254, n. 41). Their examination of the role of racial resentment and white racial identity in the 2016 election demonstrates that attitudes about the deservingness of blacks and the attribution of blame for racial inequality to lack of effort as opposed to discrimination were strongly related to votes for Trump.
This reconceptualization is a positive advance that answers scholars' call for more attention to the nature of individuals' explanations for inequality (e.g., Peffley et al. 2017). It also reflects growing attention within political science to the institutionalized legacies of slavery and their impacts on contemporary political behavior through such policy realms as criminal justice (including incarceration and felon disenfranchisement), welfare, and housing (Acharya et al. 2018; see also Soss et al. 2011, Burch 2013, Lerman & Weaver 2014, Enos 2017.
Much of this work has looked at feedback effects among black Americans, but racism studies could build on this heightened attention to institutionalized discrimination and its effects on the political realm to scrutinize the roots of prejudice among whites and members of other racial groups, and the effects on attitudes and behaviors of policy changes that attempt to reduce racial inequality.

TREATING RACISM AS A PERSPECTIVE
If the key question is how racism is at work in political behavior, as opposed to whether it is at work, then political science needs to continue to consider both implicit and explicit racism. However, it also needs to look beyond racism as an attitude, to how it functions as a lens or perspective through which people make sense of the world, including the political world.
One step in this direction is a revival and refinement of the concept of ethnocentrism within the field of political behavior by Kinder & Kam (2009). Their book Us Against Them examines ethnocentrism as a "mental habit" and "general predisposition" (Kinder & Kam 2009, p. 8), following the classic work on the topic, The Authoritarian Personality (Adorno et al. 1950). Rather than treat ethnocentrism as another consideration or ingredient in attitudes about particular policies or candidates, the authors try to understand the extent and the manner in which some people view politics generally with a tendency toward ethnocentrism.
This attention to the mental habits of ethnocentrism, which encompasses the attitude of racism, begs the question of how this general predisposition influences the way people interpret politics. I suggest that we deepen our consideration of racism by examining how it works as a perspective. To do so, we ought to supplement multivariate analyses with interpretive approaches. In the typical multivariate approach, we attempt to explain dependent variables such as political attitudes and behaviors as a function or effect of independent variables such as ethnocentrism and other controls such as race, income, partisanship, and ideology. That approach assumes that we can isolate, or hold constant, the effect of ethnocentrism from these other characteristics of an individual; it assumes that racism operates independently of other considerations. But if we are conceptualizing racism as a perspective that people construct as they absorb racist images and messages in their environments, there is reason to not treat racism or ethnocentrism as a variable that is independent from their characteristics. We might instead put the focus on understanding how people are interpreting the political world and their understanding of racism, and how they construct these meanings by drawing on multiple aspects of themselves simultaneously. Racial prejudice is intertwined with fundamental predispositions, such as attitudes toward social change (Grossman & Thaler 2018) and religion (Hall et al. 2010). Investigating how people construct understandings that weave these things together would illuminate the role of racism.
One example of what it might look like to investigate racism as a perspective, and the insight that could result, comes to us from discussions of the 2016 presidential election. An ongoing debate in the aftermath of that election is whether it is economic anxiety or cultural anxiety (i.e., racism) that drives votes for populist candidates such as Trump. When we think about this in a multivariate context, and include separate indicators for economic anxiety and racial resentment, it appears that for whites, racial attitudes have the more substantial effect (Sides et al. 2018, pp. 239-41).
However, when studied in a manner that allows us to hear people drawing on their perceptions of the local and broader economy, their sense of deservingness, and perceptions of injustices faced by different racial groups, economics and race do not appear separable. For example, in my own work, inviting myself into the conversations of white rural residents in the state of Wisconsin from 2007 to 2012, many people I talked with conveyed what I call a rural consciousness, which is a perspective of identity as a rural resident combined with a sense of distributive injustice. They talked about feeling that they were not getting their fair share of political power and attention, resources such as taxpayer dollars, and cultural respect. They resented working hard and playing by the rules and yet not getting what they felt that they deserved. They perceived that others, such as urban elites, public employees, and people of color, were unfairly advantaged (Cramer Walsh 2012, Cramer 2016.
I observed people meeting with others whom they normally spend time with, in venues such as gas stations and diners in which they usually visit. In these conversations, people rarely mentioned race and seldom made racial or racist statements. However, the notion of hard-working or good Americans played a central role in their conceptions of who in the population is deserving. These notions stood in contrast with conceptions of people they perceived as not working hard and therefore being less deserving.
When people talked about their taxpayer dollars going elsewhere, to undeserving people in the cities, they were at times referring to people of color. Wisconsin is significantly racially segregated, such that less urban areas of the state are overwhelmingly white. Sometimes the urban undeserving that people in rural places were referencing were white public workers and white elites. However, the significant segregation suggests that comments about "those people in the cities" were references at least in part to people of color.
Through hearing these conversations, one could understand that the perspective of resentment is a powerful lens that politicians can invoke and capitalize on. Rural consciousness contains not only resentment toward cities and city people but also resentment toward public employees, urban elites, wealthy people, left-leaning people, Democrats, and people of color. By tapping into animosity toward one of these entities, a political actor has the potential to take advantage of resentment toward the others. Understanding how this works requires recognizing that this resentment contains multiple intertwined strands, which should not be seen as separate forces or attitudes (Cramer Walsh 2012, Cramer 2016. Racism studies have fruitfully drawn from the fields of psychology and social psychology, although perhaps not enough (Huddy & Feldman 2009). As we consider conceptualizing racism as a perspective, we might benefit from turning toward sociology for examples. Feagin (2013, p. 6) conceptualizes racism as part of "the racialized 'water' in which [we] metaphorically [swim], the water of a sophisticated white framing of Western societies." He writes about a "white racial frame" or worldview that "encompasses a broad and persisting set of racial stereotypes, prejudices, ideologies, images, interpretations and narratives, emotions, and reactions to language accents, as well as racialized inclinations to discriminate" (Feagin 2013, p. 3, emphasis in the original). This "socially inherited racial frame is a comprehensive orienting structure, a 'tool kit' that whites and others have long used to understand, interpret, and act in social settings. The important aspects of the frame listed above become taken-for-granted 'common sense' for those who hold to them, and most holders often use these tools in automatic or half-conscious ways" (Feagin 2013, p. 12). His argument is not that all whites view the world through the same frame but that the concept "is an 'ideal type,' a composite whole with a large array of elements that in everyday practice are drawn on selectively by white individuals acting to impose or maintain racial identity, privilege, and dominance vis-à-vis people of color in recurring interactions" (Feagin 2013, p. 14).
A turn toward sociology has the potential to deepen our understanding of the way racism functions as a perspective and the implications of that function for political behavior. However, I am not alone in the field of political science in advocating for attention to perspectives. Hetherington & Weiler's (2018) book on worldviews, titled Prius or Pickup, argues that in general, Americans view politics through either a fixed or fluid worldview. Fixed types perceive the world as more dangerous and believe our priority should be to protect ourselves, while fluid types believe that "it's a big, beautiful world, mostly full of good people, and we must find a way to embrace each other and not allow ourselves to become isolated" (Hetherington & Weiler 2018, p. xi). In the latter view, embracing people of different racial and ethnic backgrounds is a priority, but in the former, it is a source of danger. The authors show that these broad worldviews have become particularly correlated with views about race. They find that individuals with more fluid worldviews displayed less racial resentment than those with fixed worldviews between 2000 and 2016, but between 2012 and 2016 they displayed a considerable reduction in resentment (Hetherington & Weiler 2018, p. 166). In other words, in the present context, views about race are particularly important for the perspectives or worldviews with which people in the United States are interpreting the world.
Meeting the challenge of conceptualizing racism as a perspective would be assisted by the use of methods that allow us to examine the language that people use to make sense of their lives. One approach is to use in-depth interviews to illuminate the meaning and place of racism in the lives of white people in the United States (Bonilla-Silva 2017). Another is to make use of open-ended questions in surveys, as various scholars have done to great effect in recent years. Jardina's (2019) open-ended questions have uncovered how people make meaning with respect to racial identity and validated the concept of white identity. A third option is to use stop-and-think probes, such as Kam & Burge (2018) use to investigate how respondents understand the racial resentment scale items. Yet another is the analysis of views about policing among black residents of highly policed neighborhoods, expressed during conversations among people who are brought together through the creative use of technology (Weaver et al. 2019).
We should also turn to methods that enable us to examine the language that people use while they are interacting with other people in order to better understand the role of racism. I have two reasons for this recommendation. First, the process of making sense of the political world often involves making sense of communication from afar (Lippmann 1922), but people do not perform this sense-making in isolation. If we would open up the field of political behavior to ethnographic methods, we could examine how people teach racism to each other, and how they imbibe, resist, or transform messages about race from others, whether those others be acquaintances on social media or elected officials. This work could include ethnography of people interacting face to face (Cramer 2016) and ethnography of online interactions.
My second reason for advocating studies of social interaction is that when people make sense of current events, they often reference their own lives and their specific geographic community (Cramer Walsh 2004, 2007Cramer 2016). We have many reasons to think that place is important for the way race enters into political understanding. Returning to the example of the relative power of economic anxiety versus racism in explaining votes in the 2016 election, when racism is conceptualized as an attitude and investigated with survey data at the level of individuals, the evidence shows that racism is a stronger driver than economics. But when we take the geographic context into account, we see suggestions that there is more to the story. The economy and general well-being of the context in which voters live matter for votes. Counties that experienced a stagnation or decline in life expectancy were more likely to vote for Trump in 2016 (Bor 2017). In the presidential elections from 1992 through 2012, people living in counties with slower economic growth were less likely to vote for the incumbent (Healy & Lenz 2017). Also, the economic health of a place has been shown to be intertwined with attitudes about restrictionist immigration policy (Filindra & Pearson-Merkowitz 2013).
Layered onto these context effects is evidence that levels of racial resentment are not constant across states. There are more racially resentful people in places receiving more federal money, and racial resentment is four times more powerful than income in explaining opposition to federal spending in those places (Krimmel & Rader 2017).
What is going on to produce these results? How is it that people are making sense of their environments? How does the context of a tough economy in one's county translate into vote choices in such a way that racial attitudes at the individual level are a stronger driver than economic considerations? How are people translating local distress into racial anxiety and vice versa? Finding out requires listening to people, preferably in a way that allows us to observe the role of place in their interpretations.
Place matters for public opinion, perhaps especially so for attitudes related to race. How people make sense of the boundaries of their community and which racial and ethnic groups they perceive to be members of their community influence political attitudes such as their support for interracial marriage and interethnic political cooperation (Wong 2010). The geography within which people live seems to influence the racial attitudes they express. People who move in and out of urban areas appear to adjust their levels of racial resentment to be in line with those in their new context (Carter & Carter 2014). Conservatives living outside the South tend to blame racial inequality on low individual initiative more than do Southerners (Carter et al. 2014). We have also learned from Enos (2017) that the distribution of people of different racial and ethnic backgrounds across geographic space influences their political attitudes and behavior. In ongoing work by Rogers (2018), we are learning that the suburbanization of the black population is having a profound influence on the black public sphere and likely enabling a diversification of attitudes among blacks. This is an important project from which we will be learning more about the role of place in the formation of public opinion in coming years.

GETTING BEYOND THE FOCUS ON WHITES
The vast majority of the literature on racism focuses on the attitudes and behaviors of whites, but it is a welcome advance that recent scholarship includes attention to prejudice among other racial and ethnic groups and is moving beyond treating race as a dichotomy of white/black or white/nonwhite (Tesler & Sears 2010, Ditonto et al. 2013, Hutchings & Wong 2014, Filindra 2017, Kam & Burge 2018. These studies show that the way prejudice works varies across racial and ethnic groups. Ditonto et al. (2013) find that traditional measures of prejudice predict a range of political attitudes for whites, African Americans, and Latinos in the 2008 ANES, but a new measure of implicit racism holds explanatory power only among Latinos. Filindra (2017) shows that racial resentment and OFR are negatively related to support for gun control among whites, but racial resentment is positively related to support for gun control among blacks.
Increasing recognition of the multiracial nature of the United States has shown up in racism studies in another form: growing attention to effects of the multiracial context on how racism operates. In particular, there is heightened attention to attitudes toward immigrants and about immigration policy. Recent work has produced evidence that the issue of immigration has driven whites from the Democratic to the Republican Party and has demonstrated that racial resentment toward blacks is positively related to anti-immigrant sentiment (Abrajano & Hajnal 2015, especially p. 70;Ostfeld 2019). The relationship between attitudes toward different racial groups and immigration stances is complex, however. Baek & Lee (2012) argue that whites with the strongest anti-immigrant stances are those who hold positive stereotypes of blacks but negative stereotypes of Hispanics and Asians.
While much of the work examining the intertwining of racism and attitudes on immigration is focused on whites' attitudes toward Latino or Hispanic immigrants, important exceptions exist, such as Nteta's (2013) analyses of blacks' attitudes on immigration policy. Even studies that do focus on racism among whites recognize the exceptional nature of whites' attitudes toward blacks in the United States context (Harrel et al. 2016).
Work on racism using the racial resentment model has had to stretch a bit to apply a measure that was intended to measure resentment among whites toward blacks to other racial and ethnic groups and their attitudes toward each other. 3 It should be noted, though, that the group position model lends itself more easily to considering prejudice and variation in how it operates across groups in a multiracial context, as demonstrated by Bobo & Hutchings (1996) and Hutchings et al. (2011).
At the same time that racism studies are broadening to consider the views of people of a range of backgrounds as well as attitudes toward a range of groups, the understanding of white public opinion is more sophisticated than ever. In particular, attention to white racial identity (Bonilla-Silva 2017, Sides et al. 2018, Jardina 2019) is pushing us to take into account whites' perceptions of their position in US society and the meaning that they are making of its increasingly multiracial context.

THE ROLE OF ELITES
The issue of white identity brings us to what is perhaps the most fruitful avenue of racism studiesthe role of elites. The relevance of any social identity, including racial identity, to politics is not innate. It takes work for people to notice the relevance of their group attachments for their assessment of particular issues or candidacies. When Wong & Cho (2005) examined white identity using the 1972-2000 ANES data, the connection they observed between white identity and politics was not very strong. But in recent years, things have changed. The presence of a black presidential candidate appears to have heightened the effect of white identity on whites' vote choices ( Jardina 2019), as we have observed in the past at lower levels of office (Petrow et al. 2018). One might argue this is the result of human cognition: The presence of a member of an out-group simply raises the salience of in-group identity.
However, the Trump campaign suggests that enterprising political actors play a role in heightening the role of white identity in the realm of politics. People with strong white identities were more likely to favor Trump over the other Republicans running in that party's 2016 primary ( Jardina 2019, p. 238). Sides et al. (2018, pp. 87-90) argue that this is the result of Trump appealing to white identity. They recount the multiple ways in which Trump invited this link: He retweeted tweets from white supremacists and did not distance himself from white nationalist supporters, for example (Sides et al. 2018, p. 88).
There is more to understand about how, at the individual level, white identity is distinct from racial resentment and blatant racism and how voters are weaving these things together ( Jardina 2019). However, the fact that white identity appears to be heightened in some contexts and by certain candidates suggests that to advance our understanding of the role of racism in politics we need to include a focus on the behavior of elites. What frameworks for understanding do candidates and other political actors provide? How do politicians exploit implicit bias, and how do they weave racial stereotypes into their campaigns (McIlwain & Caliendo 2011)? In what ways are people encouraged to attribute blame for racial inequality to individual failings versus structural causes? How does a president's discussion of racial policy invoke or not invoke racial resentment (Gillion 2017)?

CONCLUSION
The contemporary context gives us good reason to consider the explicit expression of racism, but the study of implicit racism remains important, particularly for its attention to elite-driven messages. The current study of public opinion is grounded in an awareness that what political elites put into the information environment matters for the nature of public attitudes (Zaller 1992). Even when it comes to social identities, tools for interpreting the world that individuals give meaning to in the course of their daily life (Cramer Walsh 2004), the field is more aware than ever that elites play an integral role in connecting these things to political perceptions (Achen & Bartels 2016). Understanding the role of racism in public opinion requires more attention to the elite behaviors that perpetuate its influence.
The importance of elites in the role of racism in US public opinion is another reason to apply sociological approaches. The prevailing approach to the study of racism in public opinion has been to use the racial resentment scale or a variation of it, but perhaps it is time to give renewed consideration to the racial threat model. This model moves us from the realm of individual information processing to the realm of human beings creating a power context through their interactions. We could use the framework of the racial threat model to consider how elites are creating perceptions of threat as well as how individuals are perceiving the position of their group in the contemporary context.
The rise of racism in presidential politics suggests a perception of threat to the white racial order by a significant segment of the public. This may be a product of individual racial resentment, but these sentiments arise from broader social forces in which political elites play a significant role in shaping the public's sense of the distribution of justice across groups. When we consider the ongoing debate about whether it is economic or cultural anxiety that drives support for populist candidates such as Trump, the racial threat model offers the opportunity to consider both of these drivers as threats to the group position of whites (see Mutz 2018 for a related conceptualization of support for Trump as a matter of status threat). The reconceptualization of the racial resentment scale as a matter of attributions of blame for racial inequality (Kam & Burge 2018) also suggests that turning toward sociology would be fruitful. We need to know how people process information in their environment (i.e., we need to draw on research in psychology), but we also need to know how the interactions of people create that environment (i.e., we need to turn to sociology as well).
At the same time that more attention to elite messages is warranted, so too is more attention to the grassroots work of people teaching each other racism and its place in politics, through their daily interactions. Elites exert enormous influence on the shape of public opinion, but people are not blank slates. When racist appeals gain traction, they do so by resonating. It takes both elites and members of the public for racism to maintain its grip on US politics.

DISCLOSURE STATEMENT
The author is not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review.