White Fragility, Stereotype Threat, and How Not to Do DEI

In 1995, Claude Steele and Joshua Aronson published some research postulating that one contributing factor in racial achievement gaps was a thing they called stereotype threat. They theorized that performance can be affected by the expectation that you are being evaluated according to a stigmatized category that you belong to (or even that you are believed to belong to) that carries a stereotype of being less capable at the task you are doing. You then underperform, meaning you do worse than you would if the stereotype threat were not present. It is one of the most researched explanations of test score gaps along far more lines than race at this point. Ten years ago a survey article counted more than 300 studies done on the phenomenon. Not all of them have accepted it as a central explanation of achievement gaps, but the overwhelming consensus at this point is that stereotype threat is a real phenomenon, and it certainly can affect performance, In a field like social psychology where increasing attention has been given to problems with replicating results in further studies, stereotype threat is actually one of the most replicated results you will find. I have to say that I initially thought the idea was ridiculous when I first heard it about 20 years ago, but it seems pretty well confirmed at this point. There is no denying that the anxiety about being labeled as bad in some way can affect someone's performance in areas of complete competence. There is a real effect on executive function and cognitive processing that comes from such anxiety. (Incidentally, actual researchers looking into the achievement gap on the left and right have been converging toward accepting each other's explanations as partial explanations of the phenomenon. It's an interesting case of depolarization in academia while society gets more polarized.)

I taught Geoffrey Cohen's excellent book Belonging: The Science of Creating Connections and Bridging Divides this past semester. He gives a really interesting example of stereotype threat that flips all the expected categories. He describes an event in the life of Eminem that sure seems like an example of stereotype threat. Before Eminem was successful, he showed up for a competition of some sort for rappers, completely ready to perform. Some people then told him that because he's white he'll never be a good rapper. That triggered stereotype threat in him, which induced enough anxiety to affect his ability to do the thing that he has proved himself perfectly capable of doing in his subsequent career. He got up on that stage and just couldn't do it. Stereotype threat can occur when the marginalized or stereotyped identity is normally in the majority if in some smaller context that identity is in fact stereotyped and associated with lower performance. This is a really good case of that.

He then gives another example of stereotype threat among white people that blew my mind. He thinks the observations that Robin DiAngelo has made about white people having difficulty engaging with race-conscious people on issues of race, which she calls white fragility, provide an example of stereotype threat among white people. These interactions trigger the stereotype of white people being racists, and they then underperform in their ability to engage in conversation. They fail to see important distinctions. They think they are being accused of being racists when someone is merely pointing to unconscious behavior or systemic forces in society. They can't hear what's actually being said, and then they misrepresent it pretty badly. This is a phenomenon I have witnessed countless times in conversations online, so I was intrigued by the idea.

Now I think it's important to point out that DiAngelo's style of engagement does not help matters. She makes careful statements of how she is using terms in her books, and those who read them carefully can see that when she calls something racist she doesn't mean that the person is a racist in the sense that most people hear that term. Most people think of that word as meaning someone hates people of other races or has explicitly and conscious views of their inferiority or lesser worth. But DiAngelo is very clear that's not how she's using the word. Racism for her is structural and systemic. When those who have influence do things that have an effect (whether intended or not) that is disproportionately negative to a group that is more marginalized or out of power, then on her usage of the term it is racist, whether the people doing it are discriminating or not, whether they are operating from animosity or hatred or whether they are as well-meaning as can be. She uses terms like "discrimination" and "prejudice" for other things, things more commonly called racism by most people. But racism is reserved for something else, something that doesn't presume any negative motive, just a negative effect.

Given the stigmatizing element of the term "racism," this is at best a very unwise tactic, because it's entirely predictable that she will be putting people on the defensive by talking that way. They understand racism to be a negative thing, and racists are bad people. Being told that something you do is racist has the effect of being told you are a  bad person. And then you get people like Matt Walsh who will completely misrepresent her. On one level, it's partly her own doing, because she chose to use words in a way that predictably confuses people. But at the same time, anyone actually reading what she says carefully should know that she's not saying the thing they accuse her of saying.

She calls that pushback white fragility. And then she absolutizes it. She says that any white person questioning something a black person is saying is engaging in white fragility. I have seen her followers make such claims many times in online conversations. I've had it said against me multiple times, merely because I wanted to see if a claim that was being put forward had evidence to support it. Or I would point out that there are multiple explanations behind a certain statement or action, and assuming racist motives is not necessarily fair to the person in question. Some of this comes from standpoint theory, which in its core claim is merely noticing that someone who experiences something has understanding of that experience that others do not have. And it seems right that I shouldn't question what someone else's experience is if I haven't had that experience myself. But many absolutize standpoint theory in a way that makes the marginalized out to be infallible, so that it is always wrong for a white person to question a black person's false empirical claims, even when all public evidence points the other way.

And then part of what is at issue here is intent vs. impact. We can and should distinguish between intent and impact. Both matter morally, in different ways. Sometimes impact is enough reason not to do something. But we see the person differently if they intend harm than we do if they don't. This is how our homicide laws work, to take one clear example. Someone who premediates murder gets first-degree murder charges. Someone who gets angry and kills someone is going to face a lesser charge. But even accidental homicide can face manslaughter charges, even if no harm was intended. Why should it be any different for lesser moral offenses than killing? If I unintentionally insult someone, isn't that bad, even if it's not morally the same as hatefully insulting them? In the end, I'm not a fan of how DiAngelo categorizes things, but I am also not fond of how many have responded to her. To call it white fragility every time someone pushes back is not fair. But to act like the phenomenon she is drawing attention to is not real also seems to me to be incompatible with the realities I have experienced in having conversations about race with real people.

I have witnessed the phenomenon she describes many, many times. I have seen white people push back against some claim that something is racially problematic in a way that demonstrates they didn't even bother to hear what the person was saying. They assume their own experiences are the only ones that matter and refuse to hear someone who has experienced something unfamiliar to them. They refuse to accept interpretations of events other than their own. They ignore facts and well-documented history to try to confirm a narrative they have always been taught. And they try hard to avoid any feeling of complicity in anything that might be bad, which is a normal part of the human condition. It threatens their sense of moral righteousness to encounter someone who thinks something they are doing is wrong. This is something that happens in many areas of moral thinking. Why would it not happen with race issues? But what DiAngelo has noticed is that with race it can be much more hostile. Again, part of that might be her own fault. But I've seen it happen without the use of DiAngelo's tactics and her particularities of language. Why might that be?

I think Cohen might be right. The anxiety about being seen as a terrible person is especially high in our current moment. What many call wokeness can be motivated by wanting those on the margins to be more included and welcomed, by fighting against things that disproportionately harm them or keep them being treated and seen and talked about as lesser. But the way it's often done sends signals that white people, men, straight people, cis people, political conservatives, Christians, those with greater financial means, those with attractive bodies, those who are healthier or more able, etc. are somehow problematic for being so. Language calling such groups oppressor groups serves to trigger this response. And again, this use of that term does not remotely mean in the mouths of those who talk that way anything like how it is heard. They have a systemic concept of oppression, and it's a deliberate expansion of the concept to include forces hardly anyone has any control over.

Like the deliberate expansion of the concept of racism to include systemic, structural, and institutional harms and disparities that are not always intended to harm or might be the long-term effects of a long-gone racist intent from decades ago, the expansion of the concept of oppression is meant to capture how harmful something is rather than whether it is intended to be harmful. It is meant to capture bad effects without regard to how serious those effects are. So microaggressions, even if the amount of harm from any single microaggression is small, can contribute to a systemic pattern of being unwelcoming or seeing people as outsiders, creating a kind of lack of belonging that is real. Even if it's not as severe as what most people hear when they think oppression (e.g. locking people up, burning books, torturing or executing them because of who they are), they expand the concept to include less severe things and use a morally loaded term of severity to try to wake people up to the fact that such things really are bad, even if they are not as bad as some of the things traditionally falling under those concepts. It's well-motivated effort. Characterizing that as evil misses the genuinely moral motive behind the attempt to expand that language.

But using morally loaded concepts in such ways has another effect. It feels off to many people. It seems like an attempt to fit reality to a false narrative. And people immediately see that. Calling a black person the N-word is not like calling a white person a honkey. The two are not parallel. Nevertheless, it's also not like killing them or putting them in prison or refusing to hire them or excluding them from a hospital, hotel, or restaurant. There are moral differences between different things. And telling an American-born descendant of Chinese immigrants to go home to China is cruel. It's not okay to do. But it's not the severity level of how Chinese workers were treated when the railroads were being built or on the level of laws that prevented Chinese American citizens from owning a house or coming back into this country after visiting relatives in China. Conflating these levels of severity is not all that different from when Christians in the 1980s declared that there was a war on Christmas because people were saying "Happy Holidays" in public when they didn't know if the person they were talking to celebrated Christmas or when music programs in schools preferred to have more generic songs like Jingle Bells over songs about the baby Jesus at their holiday programs. Anyone who thinks that counts as oppression and persecution of Christians should go talk to people who were locked up by the Soviet Union, executed in Saudi Arabia for their Christian faith, or thrown to the lions in Roman arenas.

It feels to many people like the same mistake if you call it oppression because someone doesn't carry skin products designed for darker skin in a pharmacy store in a very white area. We can talk about whether we should stock those products as a matter of a genuine moral question. But calling it oppression not only puts it in the same category as more severe things but immediately means anyone making decisions of what to stock on the shelf is going to be seen as an oppressor, which has the connotations in many people's minds of slavery, unjustified imprisonment, and torture. And then people who are assigned the categories that are labeled oppressor categories will internalize that there is a stereotype about them of being a bad person. If I'm assigned to an oppressor category because I'm a white, straight, cis male with affiliations on the political right and identification as a Christian, then I am conscious that I'm being categorized in a way that triggers stereotypes of oppressiveness. I'm going to be anxious about being labeled an oppressor. I'm going to be facing identity threat when I know conversations about those topics are going to come up. So I will be primed going into a mandatory diversity training to face accusations of being a terrible person. I'm facing identity threat in a way that has some parallels to a black student who is told that black people are stupid and that school is for white kids having to take a standardized test, and then their anxiety leads them to perform less well on the test than they would if they had been told that it wasn't an evaluation of them but an evaluation of the test (and empirical studies show such different results do occur).

So what should we expect to happen when someone so primed then encounters a discussion about one of these social issues? One thing that is well-recognized in the literature on stereotypes is that people who regularly experience being stereotyped have their radar turned way up. This has consequences in two directions. One is that they are more sensitive toward seeing things that are there but that others don't see. This is one reason why standpoint theory, which says that marginalized groups have greater access to certain truths than those in the mainstream, has something undeniably right at its very foundations. A black man who has been treated as if he's a thug on a number of occasions, without any evidence, is more likely to notice others being treated the same way, and other people might not even see it. Women are more likely to notice that their ideas are not taken seriously when men are dominating a conversation. This isn't a lot different psychologically from how I started to notice pregnancy all around me when our first child was still in utero. My radar was simply more attuned to seeing something that was there all along. It wasn't that more people were pregnant, but I was noticing it more. And those not attuned to that are less likely to see it when it's there.

But another thing that is well-documented to happen when you have your sensitivity increased is that you are also more likely to think something is an example of what you have experienced even when it is not. Someone who has experienced racism is more likely to see something as racist even when something other than race is the explanation. Someone who has their radar attuned to spotting ableist practices in schools will see ableism even when something else is really what's going on. You are more likely to perceive bad treatment as coming from whatever identity you experience anxiety about, even when it's not coming from that identity. You might see a nasty attitude as being about race when it's actually coming from indigestion or just a generally nasty person. You might expect that your lack of being promoted is because of gender when someone else is actually more qualified. And those not attuned to that will see it as people looking for something that's not there. Their sense that accusations are mostly imagined or elevated in severity will be reinforced. So you have one group more easily able to spot genuine problems and also more likely to see them when they are not present, and then you have another group less able to see them when they are there but more likely to get it right and not see it when it is not there.

Cohen's Belonging book points to actual studies demonstrating aspects of this overall phenomenon. Both seeing discrimination it when it's there and seeing it when it's not there are increased by someone's being a member of the group being discriminated against. How does this apply to the case I'm discussing? White people who have seen the word "racism" applied to things that they know are explained by something other than race will see false accusations of racism all over the place, even when race really is what's going on. When you have Michael Brown being given as a case of a racist assassination of a kid who had his hands up, even though that is not what happened, you begin to form a narrative of race activists inventing fictional stories about race that are not the best examples of racism. Then you see George Zimmerman, who was undoubtedly a racist in his reasons for approaching Trayvon Martin, successfully defending himself with a self-defense justification for why he, in that moment, thought Martin was going to kill him, with forensic evidence to back up his claim that Martin was bashing his head against the pavement. And then the left continues to frame it as if Zimmerman simply killed Martin because he was a black kid in a neighborhood where he shouldn't have been. The case is more complex than that, and the right could see that. So we have yet another narrative that needed more nuance. The narrative of the left wanted to see everything as racism, and the narrative of the right wanted to see it as not racism. Neither was correct. But when the left is dishonest in both cases, and a movement begins to form that reinforces that dishonesty, the right begins to distrust anyone claiming racism as part of a pattern of inventing fictional stories about events that are more complex. And the distrust becomes animosity when the narrative starts treating all white people as part of a conspiracy to engage in oppression, with the particular building blocks being used coming from fairly minor things comparatively, such as microaggressions, or complex scenarios that are obviously being spun to fit the narrative, such as the cases of Martin  and Brown.

You then have people's sensitivities turned way up to look for false charges of racism all over the place. And this is furthered by the feeling that one's moral status is being threatened by the accusation of being racist. Cognitive processing is diminished when anxiety about how one is being evaluated kicks in. Executive function isn't what it should be when you have your hostility expectations turned way up expecting to be accused of all sorts of things. Just as Eminem couldn't perform when facing the anxiety of a marginalized identity within hip-hop (even if it's a majority identity within society at large), so too do white people underperform when showing up at mandatory diversity trainings expecting to be harangued for being racists. They don't listen charitably. They misinterpret what's being said. They can't hear the truths being talked about, because their ability to do so is hampered by what they have perceived as an identity threat. But they also spot ways that they are being thought of in a morally downward way when the people doing the trainings can't even see that their language implies that, at least when ordinarily understood in its usual meanings. They are getting something right, at least in terms of the ordinary meaning of those terms. But they are not getting the actual people's motives and views right. They are not hearing what's even being said.

So I think Cohen's thesis has plausibility. It fits with what I have directly witnessed many times. And it should give us pause in how we approach these issues, so as not to trigger this kind of response. It doesn't serve the goal of educating people about problematic racial realities if we create conditions where they simply stop listening. And this does happen among well-meaning people who value diversity and want everyone included. It's not just the people who don't want those who are different around who have this kind of response. Those in favor of DEI need to take that seriously and rethink how they do things. That's the reason behind the anti-DEI backlash. There is something there that people are responding to, and it needs to be reevaluated. I happen to think a lot of the anti-DEI response is also unhealthy and does not accomplish what it intends to do very well. It makes some of the same mistakes as those it is responding to are making, which is why it's becoming popular now to talk about the Woke Right as a way of engaging in the same ways as the Woke Left but with all the roles reversed. It's white, straight, men who are oppressed rather than black, gay, women, etc. I have a second post to follow up on this that will focus on the issue of what in DEI is worth preserving and how that can be done in a healthy and helpful way, but this is long enough as it is that I'll leave that for a second post.