Academic exaggeration of conservative-side racism can contribute to our divides
Updated: Oct 8
Below is an excerpt from my book Defusing American Anger. The more I've thought about it, the more important this specific section seems to me, so I wanted to share it. While there are some people, including academics, who have written and done research on this, it is almost entirely undiscussed in the mainstream: I myself only learned about these things because I got curious about the question, "We hear a lot about how prevalent racism is in America, but what does the research actually show?"
So this excerpt examines something I think is a core problem in our American divides: the fact that some highly influential liberal-leaning academic work seems to greatly exaggerate the amount of racism in America, especially the amount of conservative-side racism. The primary way this seems to happen is that politically conservative views are equated with racism. To be clear: me examining these things is not to say that racism doesn't exist, or that racism isn't a problem (one of my first politics-related works was about a white nationalist I personally know, which is just to say I'm aware of that problem): this is about the degree to which racism is a problem, and about how biased, worst-case interpretations of the data may be amplifying our divides.
One way this presents an obstacle to reducing American divides is that, for many liberal-leaning people who work in polarization-related areas, there can be a belief that the goal of depolarization-aimed work is defeating the widespread racism/xenophobia on the right. And I, and many others, don't see the problem that way: I'd argue that liberals who have this view are being influenced by the basic nature of conflict and how it results in many of us having distorted, overly pessimistic perceptions of the "other side." Animosity and fear influence some liberals to interpret research and data in the most pessimistic ways possible. For example, they'll be more likely to interpret conservative-associated beliefs as representing racism, even thought those views can be held, and are held, by politically conservative black people and other conservative racial minorities.
And these things get in the way of what I see as the essential part of depolarization-aimed work: getting more people on both sides to see their fellow citizens more clearly—for example, getting liberals to see some of the more rational and understandable reasons for why people have voted for Trump (which I examine in detail in my book), and not see all Trump voters as a monolithic mass, all as bad as the worst, most hateful people in that group—and getting liberals to be willing to examine their contributions to our divides. We should want to avoid liberal-side bias being an obstacle to effective peacebuilding.
I recently learned about a great paper by Musa Al Gharbi on this topic titled Race and the race for the White House. His paper is likely more convincing than my work is on this topic, as he has studied it much more than I have.
Here's the excerpt from my book:
Views of America as largely racist
In some liberal-side narratives, America is a very racist country, maybe one of the worst in the world. But there are many studies that show that the U.S. is one of the least racist countries in the world.
A 2020 study by the World Values Survey studied how racially intolerant various countries are. Those who said they wouldn't want people of a different race as neighbors were deemed racially intolerant. It found that, to quote an Upworthy.com story about the study: “Americans are among the most tolerant people in the world,” and “The map shows that Eastern Europe and Asia are less tolerant than America, with The Middle East and North Africa ranking among the least tolerant areas of the globe.”
A 2019 study was titled Do Some Countries Discriminate More than Others? Evidence from 97 Field Experiments of Racial Discrimination in Hiring. It studied rates of employment callbacks depending on whether the applicant’s name was indicative of being from a racial minority or from the racial majority. From that study’s summary:
We find significant discrimination against nonwhite natives in all countries in our analysis; discrimination against white immigrants is present but low. However, discrimination rates vary strongly by country: In high-discrimination countries, white natives receive nearly twice the callbacks of nonwhites; in low-discrimination countries, white natives receive about 25 percent more. France has the highest discrimination rates, followed by Sweden. We find smaller differences among Great Britain, Canada, Belgium, the Netherlands, Norway, the United States, and Germany. These findings challenge several conventional macro-level theories of discrimination.
Such studies are important because they help us understand why people can find liberal-side narratives about American racism unconvincing and exaggerated.
And, again, one can criticize liberal-side stances on race while believing that racism still exists and is still a problem. That’s something most of us agree on. Remember that in the More in Common study, 80% of conservatives said that they believed racism still exists in America.
It would be impossible to argue that racism doesn’t exist in America, or in any country. There are, unsurprisingly, many studies that have found convincing evidence of racism in America. To mention one example: a 2020 study by Ravi Shroff that found that in some regions, after sunset, there can be a 5-10% decrease in the number of black drivers pulled over, which implies that during the day police are using race as a factor in who to pull over.
People objecting to liberal-side stances are not usually objecting to the belief that racism exists in America: they’re simply objecting to the most pessimistic and highly certain depictions of America as hugely racist.
Narratives that most Americans are racist
If you’re someone who reads a good amount of news, you’ve likely stumbled across various stories about the large number of racist Americans, or the large number of racist conservatives.
To take one example, a USA Today article from 2012 was titled AP Poll: U.S. Majority Have Prejudice Against Blacks. Here’s an excerpt from that piece:
Racial attitudes have not improved in the four years since the United States elected its first black president, an Associated Press poll finds, as a slight majority of Americans now express prejudice toward blacks whether they recognize those feelings or not.
When measured by an implicit racial attitudes test, the number of Americans with anti-black sentiments jumped to 56%, up from 49% during the last presidential election. In both tests, the share of Americans expressing pro-black attitudes fell.
"As much as we'd hope the impact of race would decline over time… it appears the impact of anti-black sentiment on voting is about the same as it was four years ago," said Jon Krosnick, a Stanford University professor who worked with AP to develop the survey.
To give another example: in 2016, Brian Schaffner and colleagues surveyed white Trump voters to try to find factors that influenced their voting. In a Vox.com piece, Schaffner wrote:
We are left with competing narratives, with some reports suggesting that economic insecurity was the decisive factor in this election, and others highlighting the role of racism or sexism in driving voters toward Trump. The truth, however, is that there is no single cause of Trump’s success among whites. All three factors played an important role.
These surveys, and others like them, are often used to make the case that a large percentage of Americans are racist, or that a large percentage of conservatives are racist. For example, Ezra Klein used this kind of research in his Why We’re Polarized book to s
upport the narrative that Trump support was largely about xenophobia and racism.
But the problem is: When you start digging into the research behind these framings, you’ll find that these interpretations are quite debatable. These framings can be criticized for taking the most pessimistic view possible of what people are saying. Some of the interpretations categorize standard conservative views, including views held by black conservatives, as racist.
And this is important because these are the kinds of studies used by many liberal people to make the case that America is a hugely racist country, or that conservatives are largely racist. And this can present a major obstacle to doing effective, persuasive depolarization-related work. If we care about the goal of decreasing toxic polarization, it’s important that we question some of our certain and righteous narratives, and be willing to examine nuance in these areas.
I’m going to delve into some of this now. This analysis clearly won’t be exhaustive, but my goal is to make some points for why we should be skeptical of a lot of these interpretations.
One way to measure racism is by measuring what is called explicit racism, which is racism that’s fairly overt and obvious. This would be someone expressing views likely to be considered as racist by almost everyone, like being against interracial marriage.
Explicit racism is, not surprisingly, still present in America. But surveys measuring explicit racism show a very positive picture of racism being on a continually decreasing downward slope. For example, Gallup surveys show that approval of marriage between “whites and nonwhites” has continually gone up since the 1960s. In 1961, approval of interracial marriage was about 5%. In 2021, it was at 94%.
This jibes with other measurements of explicit racism. In 2008, a New York Times survey found that roughly 5% of Americans said they would not vote for a black presidential candidate.
But some people thought that measuring explicit racism wasn’t telling the whole story. It was thought that some racist people would want to hide their racism and wouldn’t be willing to admit to such beliefs on surveys. Some people thought that in order to measure America’s racial divides, you’d have to use more indirect methods.
The new, more indirect ways of trying to measure racism used terms like new racism and symbolic racism and racial resentment. Another way of attempting to indirectly measure racism was with implicit bias or unconscious bias research: these tests attempt to find subtle differences in how people react to race-related words or images.
The problem with some of these approaches is that the more indirect your methods become for measuring racism, the more it becomes likely that you’re measuring something else entirely, perhaps something not relevant at all. The more indirect your methods, the more noise you’ll have in the data. And that can lead to bad interpretations of the data, especially if that data is being interpreted by biased people.
Another criticism can be: the more subtle the racism a study claims to find, the more debatable it becomes that such things play any role in the real world.
The survey questions aimed at detecting this “new racism” are often variations on questions like these:
“Do you think racism is still a problem in America?”
“Do you think black people are getting too much assistance?”
“Do you think bad results of black people as a group are more due to discrimination or personal motivation?”
The thinking is that if someone says that racism isn’t much of a problem in America, or thinks that black people are getting too much help, they’re demonstrating racial resentment, which is then likely to be interpreted, either by the researchers or the media, as an indication of racism.
The previous survey mentioned, the one by Brian Schaffner that claimed to find that white support for Trump was “driven by racism,” relied on a single question that asked if people agreed or disagreed with the statement that “white people in the US have certain advantages because of the color of their skin.” If someone answered “No” to that, they were deemed to be motivated by racism.
But hopefully, if you’ve read this section on race so far, you can see how someone might answer these questions in different ways without being racist. It’s possible to see how non-racist people, including some black people, might not believe racism is a big problem in America. And yet people who say that, even if black, would be categorized as racist based on the common interpretations used in these studies.
It’s possible to see how someone might think that black people are getting too much assistance. That would go along with believing that racism isn’t a big problem in America. And it’s possible to see how conservatives would be more likely to believe that, because conservatives are generally against government assistance. And it’s possible to see how some black people might believe that. And yet people who say that, even if they’re black, would be categorized by some researchers as racist for having that belief.
Let’s look closer at one specific study. A study from 2018 was titled Understanding White Polarization in the 2016 Vote for President: The Sobering Role of Racism and Sexism. This work was done by Brian Schaffner and colleagues. That study gets referenced a good amount to support the idea that Trump support was largely about racism.
For the segment of that survey that dealt with race, they asked people to agree or disagree with these statements:
1. White people in the U.S. have certain advantages because of the color of their skin.
2. Racial problems in the U.S. are rare, isolated situations.
3. I am angry that racism exists.
Let’s imagine how one might be able to answer these questions in ways that might strike some people as bigoted but that aren’t actually bigoted:
For questions #1 and #2: Even if you yourself believe racism still plays a big role in America, it’s possible many people simply don’t see the evidence for that. For example, let’s say you live in a town you think is racially harmonious. If the media you consume portrays a world where racism hardly plays a role, it would be easy for you to think that racism is not much of a problem in America.
Also, for the statement “Racial problems in the U.S. are rare, isolated situations,” the phrase “racial problems” is ambiguous. What exactly constitutes a racial problem? Isn’t that a subjective thing? Also, the phrase “rare, isolated situations” is also ambiguous: what exactly qualifies as “rare” in a nation of 330 million people?.
For question #3 (“I am angry that racism exists.”): It’s possible for someone to believe that racism exists and is a problem, while not being made angry by it, in the same way that one may realize all sorts of bad ideas and bad people exist in the world and yet not be made angry by them. It can be seen as similar to asking: “Are you angry that murder exists?” In other words, it’s possible to take a stoic view of things that seem so embedded in human nature. (I’m not sure I myself am angry that racism exists, as much as I am simply aware that racism and other forms of harmful stereotyping are common and unfortunate human tendencies that we need to work against.)
And another factor for that question: the more one feels that racism is not a major problem, the less likely one is to be made angry by it.
And also a possible factor here: the more one perceives liberal stances on race and racism to be unreasonable, the more reasons one might have to answer No to this question as a way to vent some of that frustration. The more polarized and angry we are, the more we’ll be likely to vent our anger in surveys in various ways.
Even in this simple set of three questions, we can see a lot of ambiguity. And yet the results of this survey have been used to make confident pronouncements that many Americans are racist. If you’re someone who cares about healing America, you should be willing to examine the idea that biased, irresponsible framings on these topics may be contributing to our divides.
There are plenty of criticisms of these “new racism” studies from academics.
A paper by Stanley Feldman and Leonie Huddy was called On Assessing the Political Effects of Racial Prejudice. This paper made the point that some of the studies on new racism might not actually be measuring racial prejudice. (I interviewed Huddy for the podcast about this.)
To quote the opening sentence of their paper: “There is still no broad consensus on the extent to which racial prejudice influences white Americans’ political attitudes, in part because of an ongoing dispute over the nature and measurement of racial prejudice.”
Here’s another quote from that paper: “Despite considerable effort by numerous researchers over several decades, there is still no widely accepted answer as to whether or not prejudice against blacks remains a potent factor within American politics.”
One of the criticisms of “racial resentment” measurements discussed in the Huddy paper and in other places, is that some of this work can be seen as studying political and philosophical views just as much as, or more than, it’s studying race- or racism-related views. To quote from Feldman and Huddy’s paper:
Consider an item in the racial resentment scale, one of the most common measures of the new racism within political science, which states that “if blacks would only try harder they could be just as well off as whites” (Kinder & Sanders 1996). A strong individualist would agree with this statement. But the individualist would also agree with any statement that referred to the positive effects of hard work, regardless of the target person’s race, gender, or other characteristics.
Kinder & Sanders (1996) believe that individualism has become entwined with racism, so that agreement with the notion that blacks are unwilling to work hard is a form of racism. But this leaves no room for racial policy opposition grounded in general, non-racist individualism.
In other words, it can be hard to say what’s due to political views and what’s due to “racial resentment.” If a conservative believes in small government and believes people should largely be responsible for themselves, they would score high on many racial resentment surveys.
Here’s an excerpt from a 2011 paper titled On the Meaning, Measurement, and Implications of Racial Resentment by E.G. Carmines:
Howard Schuman has suggested, however, that the index used to measure racial resentment may be fundamentally flawed because it may be conflated with the measurement of attitudes toward racial policies. The authors’ analysis supports Schuman’s suggestion. They conclude that racial resentment is not a valid measure of racism, which raises questions about the extent to which a new racism now dominates the thinking of white Americans.
It’s also been found that people categorized as “racist” for answering survey questions in these ways often have the same feelings about white immigrant groups, not just racial minorities. Here’s an excerpt from a 2015 paper titled Conservatism and Fairness in Contemporary Politics by Carney and Enos:
To investigate the attitudes underlying modern racism scales, we offer a novel test: we deploy surveys using the exact questions that constitute standard modern racism scales, but we substitute Blacks for other target groups that are not commonly associated with known stereotypes or overt prejudice in the United States. For example, we substitute Blacks for Lithuanians and then measure the difference between mean resentment toward Lithuanians and Blacks.
Across multiple groups and multiple samples on different survey platforms, we find a strong and consistent pattern: the results obtained using groups other than Blacks are substantively indistinguishable from those measured when Blacks are the target group. Decomposing this measure further, we find that political conservatives express only minor differences in resentment across target groups. [emphasis added]
It’s also been found that black conservatives answer these kinds of questions in very similar ways as white conservatives. In a paper titled Racial Resentment and Public Opinion across the Racial Divide, the researchers took some questions that are often asked to white Americans as a way of determining racial resentment, and asked those questions to black Americans. To quote from that paper: “Among both blacks and whites, the relationship between partisanship and racial resentment is nearly identical, with strong Democrats scoring lower on racial resentment than strong Republicans.”
In other words, if we apply the same interpretation often used to describe white Americans and white conservatives as racist, black conservatives would also largely be considered to be racist.
I could keep going, but you likely get the idea. It’s hard to say if the views often pointed to as evidence of racism are actually measuring racism. Considering the ambiguity in these areas, and considering how incendiary it can be to claim that many Americans or many conservatives are racist, making such confident claims can be seen as irresponsible and as contributing to our divides.
The more that liberals label conservatives as racist, the more angry and defensive conservatives will become. And the more angry and defensive conservatives become, the more likely they’ll be to answer survey questions in ways that vent their frustration towards liberals, in ways that’ll be interpreted by some liberals as racial resentment and racism. There’s likely a feedback loop at work here, as there can be for many polarization dynamics.
Let’s imagine that in the year 2070, almost all black Americans think that America is largely fair and just and unprejudiced. If those black Americans were to take some of today’s racial resentment surveys, they’d be scored as being high in racial resentment towards black people. This begs the question: how non-racist must America become for it to be understandable that people can answer such questions in very different ways?
Put another way: At what point along that hard-to-define spectrum of American progress on issues of race and equality might we cut people some slack for viewing America and Americans as being mostly fair and unprejudiced?
If we’re able to see how there can be understandable differences in perception in these areas, we’ll lower our anger. We’ll also be able to see why it is that highly pessimistic interpretations of this kind of data bothers people.
Criticisms of implicit bias tests
Another way to study “new racism” is by measuring so-called implicit bias. This is largely aimed at measuring more unconscious forms of prejudice.
A common way implicit bias is measured is by studying very subtle reactions in how someone reacts to pictures of black and white people, or even just reactions to words that have “black” or “white” in them. Here’s one description of how this can work, from a paper by Leony and Feldman:
In one of the most common tasks, evaluative association, subjects are briefly exposed to the target prime (e.g., the word black) and then make a judgment about the positive or negative valence of a subsequent target word. The time taken to make this judgment is used to assess the valence of one's attitude toward the prime. If the prime and target word share similar valence, the valence decision is made more rapidly. Thus a subject who decides more quickly that “bad” is negative than that “good” is positive after being primed with the word “black” holds a negative implicit attitude toward blacks.
These kinds of tests find a lot of evidence of unconscious racial prejudice amongst people, and not only white people. It’s been reported that these tests find that more than 50% of black people have an unconscious preference for white people over black people.
These tests have gotten a lot of attention in the liberal-leaning mainstream media. You can find many articles and op-eds that speak as if it’s straightforward and proven that these tests show that many people are prejudiced.
To take one of many examples: a 2014 Washington Post article was titled Across America, whites are biased and they don't even know it. According to that article:
[The implicit bias surveys tell us] something pretty crucial to our understanding of racial bias: It is everywhere, from north to south, from Maine to California. It is present among liberals and conservatives, men and women, young and old. We have a huge amount of work to do.
You can find these ideas referenced in many workplace trainings related to diversity and racial sensitivity, as a means for making the case of how embedded our racial biases are. You can find this work often referenced in academic papers and workshops.
Malcolm Gladwell’s book Blink was one way these studies became more widely known in the mainstream. In that book, Gladwell, who's half Jamaican, took an implicit bias test, which found he had a bias against black people.
The problem is that, as with a lot of the new racism research—and more broadly with a lot of psychology research in general—there’s a lot of ambiguity here. Some academics believe it is wrong to make confident pronouncements about what these tests are finding, or about how significant those findings are. I’ll just go through a few reasons to be skeptical about the significance of this research.
When it comes to how a person scores on such tests, it can vary greatly not just from person to person, but from moment to moment. Vox, a progressive-leaning news site, had a piece about this titled For years, this popular test measured anyone’s racial bias. But it might not work after all. The subtitle of that piece was: People took the implicit association test to gauge their subconscious racism. Now the researchers behind the test admit it can’t always do that.
In that piece, the writer, German Lopez, wrote about his experiences getting a wide variety of conflicting results from the most well known implicit bias test (www.implicit.harvard.edu). A co-creator of that test, Tony Greenwald, admitted that the IAT is only “good for predicting individual behavior in the aggregate, and the correlations are small.”
To quote from that piece:
This isn’t how the test was sold in books like Malcolm Gladwell’s Blink or the pages of news organizations like the New York Times. The assumption seemed to be that you could take the test once and come away with a clear picture of your bias. And that served a real-world purpose: In a society that no longer tolerates explicit racism nearly as much as it used to, uncovering people’s subconscious implicit biases seemed like the way to show people that they really can be and are still racist.
The IAT, however, does little to make its flaws clear to test takers. When you finish the test, you get a big message on your screen proclaiming you either have an “automatic preference,” with different measures like “slight” or “strong” to gauge the level of your preference, or “no automatic preference.” There’s no clear disclaimer about how one shot at the test is likely meaningless for predicting an individual’s bias and behavior.
So maybe these tests aren’t so great for individuals, but what about in aggregate, measuring larger societal levels of bias? To quote from Lopez’s piece:
[Here’s what Hart] Blanton told Singal: “If you’re not willing to say what the positive [IAT score] means at the individual level, you have no idea what it means at the aggregate level. … If I’m willing to give 100 kids an IQ test, and not willing to say what an individual kid’s score means, how can I then say 75 percent of them are geniuses, or are learning disabled?”
In other words, if the test can’t predict individual behavior after one session, how can we be so sure that it can really tell us anything through an aggregate of those individual tests?
Similar to the racist resentment studies, we can see how liberal-leaning people and organizations might have a tendency to promote ideas that align with their preferred narratives, no matter the complexity and ambiguity present in the science. (And to be clear, in case it needs stating: this requires no conspiracy. It’s an entirely understandable outcome of people having various biases and blind spots.)
Even if there is actual racial bias being found in these studies, it doesn’t logically follow that that bias would play a role in the real world. For example, it wouldn’t be surprising that people demonstrate some small preference for people who have traits like their own. We are, after all, tribal creatures, and some attitudes that can be called racist can be seen to be due to these primitive but understandable group psychology instincts.
Or, to take another example, it’s possible some of us might be influenced by our environment to form stereotypes about some racial groups. For example, let’s say you live in an area where black people commit most of the crime, and for that reason you may be more likely to associate black people with negative traits, even as you consciously operate on a principle of treating everyone equally. And it’s possible even that black people might have such a bias, if they lived in an area where black people committed most of the crime.
But these biases, even if found, wouldn’t necessarily mean anything in a practical sense. Such bias might be meaningless when compared to all the other factors governing our choices and behavior. Let’s say you’re black and have a slight bias against white people, as a group: Your rational way of navigating the world might completely eclipse that bias, and that bias might never play any real role. So we can see that even when we’re sure we’ve found bias, we can be unsure how meaningful that is.
What we can say for sure is that there’s a lot of ambiguity and uncertainty in implicit bias research and interpretations. And if we think depolarization is a worthy goal, we should see it as important to avoid confident assertions about what such findings tell us about the people around us.
END OF EXCERPT
If you were reading this thinking, "But clearly Trump is racist; to support him, his voters must be, too," you might appreciate this excerpt about our hugely divergent, polarized perceptions of Trump.
If you enjoyed this excerpt, you might enjoy my podcast. One relevant episode is this talk with Guy Burgess about how liberal-side bias can be an obstacle to conflict resolution.