The Five Dogmas of DEI
And how to avoid the totally unnecessary confusion they create.
The efforts of the Trump administration to purge DEI (“diversity, equity and inclusion”) programs from the federal government bureaucracy have revealed an enormous amount of confusion about what DEI actually is (or was). In part this is due to the way it was sold by its practitioners, who often presented it as a straightforward extension of the ideas and principles that animated the civil rights movement of the 1960s. In reality, many of their claims were a lot more controversial. If there is any hope for resisting the full-fledged assault that is taking place, it should begin with a re-examination of these claims, with an eye to producing a more defensible body of doctrine.
Like most people who work in large bureaucratic organizations, I have attended several diversity seminars and training sessions over the past decade. I also chose to observe a similar session in my children’s high school. In each case, I had a rather foreboding sense that the ideas being presented about racial justice were not just controversial but unnecessarily controversial. Specifically, I felt that the instructors were making claims and picking fights over issues that didn’t really need to be settled for the purposes of the program.
Most obviously, DEI proponents have encouraged an extremely controversial approach to affirmative action, that in many cases abandoned any concern over procedural fairness in order to achieve targeted outcomes. Many Republicans who complain about DEI use the term as a synonym for this style of affirmative action. What I would like to focus on instead are the controversial theoretical and philosophical claims that played such an important role in the ubiquitous training seminars (e.g. implicit bias, structural racism, white privilege, etc).
I trust that these claims will be recognizable to anyone who has attended such a seminar (if not then congratulations, you’ve been attending better seminars than I have). I want to pick out five issues that I think need to be re-examined. The first three constitute factual claims that, despite having been widely accepted in the progressive discourse around race in the United States are either false or else constitute extreme simplifications of complex states of affairs. The final two constitute claims about justice that are either incorrect or hugely controversial.
I do not want to suggest that the current backlash could have been avoided, if only proponents of DEI had been more cautious in their claims. It does however seem to me that if one is going to address such an important set of civic issues, and especially if the training is going to be mandatory, then the curriculum needs to be absolutely bullet-proof. Most obviously, it should not contain false, speculative, or unnecessarily controversial claims—otherwise you are just handing ammunition to your critics. This is, unfortunately, exactly how the standard DEI curriculum was handled. Although it is difficult to generalize, many people will easily recognize the following claims, which I tend to think of as “the five dogmas of DEI”:
1. Race Is A Social Construct
It is important for people to know that “race” is no longer considered a useful scientific concept. It is also important for them to know that the major physiological differences that can be observed between human population groups from different geographic regions, such as differences in skin color, hair type, or eyelid fold, are quite minor from a biological point of view. So even though human beings may look quite different from one another in our external appearance, these differences are superficial; from a biological perspective we are all very closely related, and quite similar with respect to our major features.
Many anti-racism educators, however, have felt the need to go beyond teaching these widely accepted facts, to make the more extreme, complex, and in certain respects false claim that race has no underlying biological “reality,” but that it is instead a “social construct.” This claim is expressed in various ways: race is not real, race is not biological, race has no genetic basis, or race is a biological myth. The idea was featured perhaps most influentially in the instructional video, Race: The Power of an Illusion, originally aired on PBS. (In the teaching materials issued to accompany the video, the claim is expressed as follows: “Race has no genetic basis. Not one characteristic, trait, or even one gene distinguishes all the members of one so-called race from all the members of another so-called race.”)
These claims are at best confusing (in part because they invoke very difficult philosophical concepts, e.g. what does it mean for something to be “real”?). But there is also at least one sense in which the claim is false. In every canonical use of the term, race is determined by ancestry, and ancestry is a straightforward biological concept. (Determination here means that if two individuals have the same ancestry, e.g. the same parents, then they must be of the same race.)
The claim made in The Power of an Illusion is actually a confused interpretation of an argument made a long time ago by Richard Lewontin, who calculated that approximately 85% of the genetic variation found in humans occurred between individuals, and 15% between geographically dispersed population groups. He concluded, on this basis, that race had no taxonomic significance in human biology. This statement is often interpreted as the claim that race has no “genetic basis,” or that it is “not real.”
The problem with Lewontin’s argument is that, even though a point-to-point comparison of each genetic locus will not allow one to distinguish members of one race from another, examining the correlations that exist within all of this variation can be used to determine an individual’s ancestry, which in turn determines their race. This is why, as everyone knows, you can take a saliva sample, send it off to a company that does DNA testing, and find out what your racial background is. It does nothing to advance the cause of racial justice to deny these well-known facts.
If educators really need to use philosophical terminology, the correct thing to say would be that race is not a “natural kind.” Race is a folk concept, similar to when we say that someone has “caught a cold.” There is, in a sense, no such thing as a “cold” virus, because the term refers to a heterogeneous collection of viruses that don’t have much in common. Scientific journals therefore do not refer to cold viruses, they use much more specific terminology (rhinovirus, adenovirus, etc). But it would be highly misleading to claim, on this basis, that colds are “an illusion,” or that they are a “social construct.”
2. Stereotypes Are False
Consider a claim about groups, such as “men are stronger than women.” This can be interpreted in different ways. If it is understood as a claim about averages (“on average men are stronger than women”) or probabilities (“any randomly selected man is likely to be stronger than a randomly selected woman”) then it is true. But if it is taken as a universal generalization (“each and every man is stronger than each and every woman”) then it is false. It follows that universal instantiation based on such a claim (“he is a man, so he must be stronger than her”) is invalid.
Social psychologists, when speaking precisely, strive to avoid this ambiguity by referring to claims about groups that are true in the average sense but not in the universal sense as “accurate.” (Statements about groups are “inaccurate” when they are false both as claims about averages and as universal generalizations.) A very common form of unjust discrimination occurs when people move from making an accurate claim about a group to making a false claim about an individual member of that group, via unwarranted universal generalization and instantiation.
This reasoning is not difficult to explain or understand, and the intuition about fairness is very widely shared. The takeaway lesson is that we need to be very cautious when moving from making claims about groups to making claims about individuals. Unfortunately, the way that many anti-racism educators have tried to encourage this caution is by telling students that these beliefs about groups are stereotypes, and that stereotypes are simply false, not just as universal generalizations, but even as claims about averages, probabilities, or frequencies. The easiest shortcut to making this argument is to claim (fallaciously) that the statement expressing the stereotype is inaccurate because it is false as a universal generalization.
This has resulted in enormous confusion among Americans about what is and is not racist. For example, expressing the view that “Asians work hard in school” would be widely derided as the expression of a racist stereotype. Yet even though this claim is not true universally, it is nevertheless accurate—on average Asian students in the United States spend much more time doing schoolwork than white or Black students. Not only is the statement accurate, it expresses an important fact about the world, since hard work is the predominant factor explaining Asian-American educational success. As a result, educators may put themselves in the unenviable position of either denying, or condemning as racist, what seem to most people obvious statements of fact.
The idea that all stereotypes are false, and therefore inaccurate, is further exacerbated in discussions of “implicit bias” or “unconscious bias,” since many of these biases arise from stereotypes (which generate the implicit associations, etc). The problem is that the term “bias” is intrinsically pejorative. It sounds contradictory to speak of an “accurate bias,” and yet many of the examples given in anti-racism exercises of implicit or unconscious biases are based on stereotypes that are accurate. Strictly speaking then, they should not be called biases, but to the extent that they are, it reinforces the mistaken idea that all stereotypes are inaccurate. This again commits the educator to denying plain facts about the world.
3. Racism Is Not Innate, But Learned
The desire to believe in the innocence of children is powerful, but it should not be allowed to influence the study of human nature. Many anti-racism educators have unfortunately chosen to promote the idea that “racism is learned” (e.g. “We were each born without prejudice into a world that has systematically taught us to accept an oppressive system”). In some cases this is probably meant as a bromide rather than a statement of fact, to encourage optimism about the potential for progressive social change. While obviously we are capable of suppressing innate dispositions, it is true that racism would be a great deal easier to eliminate if it was entirely learned behavior—we could simply stop teaching it to children or modeling it for them.
Contrary to this optimistic claim, the current scientific outlook on the psychological foundations of racism is nuanced. The basic tendency to classify the social world into in-group/out-group, and to act more cooperatively toward members of the in-group than the out-group, is clearly innate. The basis of demarcation, however, does not appear to be innately specified, and so those who fixate on race as the primary basis of in-group/out-group differentiation have most likely picked this up from the social environment.
Unfortunately, because the groupish mentality is innate, and because the physical differences underlying racial classification are fairly salient, children are perfectly capable of inventing racism for themselves without any instruction (or on the basis of extremely subtle cues). Furthermore, they need not witness any intergroup antagonism or prejudice. Even observing patterns of mutual support may cause them to demarcate the social world along racial lines, which may combine with their own exophobic tendencies to generate racism. If one is dogmatic in the insistence that racism must have been learned from someone, it can easily produce a witch-hunt mentality, aimed at locating the sources of the racism that the supposedly innocent children are suddenly manifesting. This can lead teachers to suspect that parents are racist, parents to suspect that schools are racist, and so forth. (The difficulty that everyone has locating all of this racism probably contributes to the belief in widespread subconscious racism or implicit bias.)
The correct thing to say is that, although we are not hardwired to be racist, humans are by nature groupish and exophobic, dispositions that can easily lead to racism. Furthermore, the social emotions implicated in this complex lead many people to find racism psychologically gratifying. We are not all innocent until corrupted by society; the seeds of racism lie within our own minds. Our job must be to ensure that they are not allowed to grow, which requires that we be vigilant with respect to our own psychological tendencies.
4. Majority Privilege Is Unjust
Perhaps the single most influential piece of writing in the DEI canon is Peggy McIntosh’s “White Privilege: Unpacking the Invisible Knapsack.” This article introduced the idea of a “privilege checklist,” which white students are asked to go through in order to reflect upon the advantages they have enjoyed in life by virtue of their race. McIntosh’s original checklist has attracted hundreds of imitators and has proven to be a lightning-rod for controversy.
Many people have objected to these checklists on the grounds that they are divisive, but there is a more glaring problem with them, which is that they (almost invariably) fail to draw an important distinction between the forms of privilege that they identify. McIntosh mixes up racial privilege, which whites enjoy by virtue of their membership in their racial group, with majority privilege, which whites enjoy, not by virtue of their race per se, but simply because their racial group forms the demographic majority in American society.
This confusion has become ubiquitous in DEI instruction. For example, one privilege checklist widely used with U.S. secondary students includes the question, “When I go to the store, people believe that I am trustworthy and I will not steal something,” but also, “The majority of the staff at my school look like me.” These are obviously quite different. A white person studying in China might still enjoy the former privilege but obviously would not enjoy the second. That’s because the second question is not really diagnosing racial privilege, it is merely identifying consequences of being in the majority demographic group.
The problem with failing to draw this distinction is that, while the advantages that stem from racial privilege are clearly unjust (everyone should be able to walk around a store without attracting suspicion merely because of their race), the advantages that stem from majority privilege are not necessarily unjust (should everyone be able to attend a school at which the majority of teachers are of their own race?). In many cases, the benefits are just a consequence of having mainstream tastes and preferences.
For example, McIntosh lists the greater ease with which white Americans are able to find a competent hairdresser as a form of white privilege. But she then goes on to claim that white privilege creates a system of racial dominance. This invalid inference leads to all sorts of problematic conclusions. Most obviously, the inclusion of majority privilege on these checklists is tacitly anti-integrationist, insofar as it suggests that members of racial minority groups are dominated merely by being in the minority within a society, or in a group interaction. This conclusion has been amplified by the tendency to use morally charged terms, such as “white supremacy,” to describe any institution in which whites form the numerical majority.
In the past, the term “white supremacy” invoked images of the KKK, and so the demand to “end white supremacy in America” would have received overwhelming endorsement. Through the confusion of majority and racial privilege, however, the demand to end white supremacy has now become a great deal more controversial, because of the implication that white supremacy can only be ended by making it so that whites no longer form a demographic majority. Indeed, this claim lends support to the “great replacement theory,” which is often dismissed as paranoid, but has received unexpected support from the use of these privilege checklists with their clear implication that ending white privilege will require drastic reduction in the white share of the population.
Again, it is important to emphasize how unnecessary all of this is. It is not difficult to come up with a checklist that focuses on ways in which whites in America are exempt from forms of manifestly unjust treatment that are experienced by minority groups. Drawing their attention to this can be a useful pedagogical exercise. There has been some debate over whether trying to induce feelings of guilt in whites constitutes an effective political strategy for promoting racial justice. But without even settling this question, it is easy to see that trying to make whites feel guilty about aspects of American society that are not even unjust is going to be ineffective at best, and more likely counterproductive.
5. Racial Disparities Are Unjust (Per Se)
Another major concept popularized by DEI initiatives is “structural racism” or “systemic racism.” These terms are intended to describe circumstances in which larger social systems reproduce objectionable racial inequalities, without any specific social actor necessarily engaging in discriminatory behavior. For example, many facially neutral job requirements can have the effect of limiting access for members of certain racial groups. It is therefore important to examine any requirement that has this sort of disparate impact in order to determine whether it is justifiable.
This set of ideas has unfortunately encouraged another shortcut, in which DEI educators treat the mere existence of racial disparity as sufficient evidence of “systemic racism.” This generates the view that Matt Yglesias refers to as “disparityism,” which basically collapses the distinction between structural racism and racial disparities in outcome. Instead of treating unequal outcomes as grounds for suspicion that some unjust discrimination is occurring, which must be investigated further in order to render a verdict, disparityism jumps directly to the conclusion that the outcome, being unequal, must be unjust. (This is most prominent in the work of Ibram X. Kendi, who defines racism as the existence of racial disparity in outcome.)
The problem with this view is that practically every disparity is the product of multiple causes, some of which are innocent, others of which are not. For example, Kendi cites the difference in homeownership rates between white and Black Americans as a canonical example of racism as he understands it. Similarly, almost everyone has heard the statistics on differences in household wealth between white and Black Americans. And yet in order to pick out what is objectionable in these states of affairs, it is necessary to “control for” various factors that influence them without implying injustice. For example, the median age of whites in the U.S. is 10 years older than that of Blacks. This obviously needs to be factored out if one wants to assess racial inequality in the distribution of goods like housing (which tend to be owned by older people). Making judgments of equality and inequality between groups is incredibly difficult for this reason.
Mentioning these mitigating factors, however, almost always gives rise to an incredibly unproductive debate, in which those who are more sympathetic to the accusation of racism point to some other, unjust factor that may have contributed to the disparity. This will often be true as well. The point that gets overlooked is that the disparity, taken as such, proves nothing about the justice or injustice of the outcome. This is why social scientists who study these questions are generally cautious when it comes to leveling charges of racism. While there may be perfectly good grounds for suspecting that there are unjust inequalities at work in many of these disparities, demonstrating this can be painstaking, because it is necessary first to factor out all the innocent causes. A great deal of DEI thinking, however, simply blows past all of these difficulties, using the phrase “systemic racism” as a shortcut to denounce as racist practically every institution in American society, without actually doing the work necessary to establish such a claim.
Lessons To Be Learned
As I mentioned at the beginning of the discussion, none of these five dogmas is entirely false. The problem is that they are not entirely true either, as a result of which their widespread promulgation generates completely unnecessary suspicion and controversy. Furthermore, none of the claims are important to the basic pedagogical objectives of the anti-racism curriculum. In each case, claims that are both true and uncontroversial can easily be substituted for them, without any loss of integrity.
To arrange things in a style that will be familiar to DEI supporters, I would recommend the following:
Given the unpleasant character of many of the attacks that have been made against DEI education, anti-racism seminars, and the teaching of “critical race theory,” one would like to be able to offer a wholehearted defense of the curriculum. In order for this to be possible, however, it would have to be the case that all empirical claims were based on clear scientific evidence, and every normative statement was well thought-out.
In reality, DEI has contributed to a great deal of confusion about race and racism in American society. I understand that some things are difficult to explain, and that the claims on the right-hand side of the table above are more complicated than the ones on the left. But they are also a great deal more accurate. While I sympathize with the motives of those who make the claims on the left, ultimately they cut too many corners, and try to make the argument a bit too easy. Proponents of DEI need to come back with something stronger, with a body of doctrine that one can reasonably expect the entire population to accept.
Joseph Heath is a Professor in the Department of Philosophy at the University of Toronto and writes the Substack In Due Course.
Follow Persuasion on X, LinkedIn, and YouTube to keep up with our latest articles, podcasts, and events, as well as updates from excellent writers across our network.
And, to receive pieces like this in your inbox and support our work, subscribe below:
 | A guest post by
|
Exactly. DEI as it's been practiced begins with grains of truth that they exaggerate into lies. We need to reject the latter without denying the former.
Useful article but straining things a LOT to suggest innate in-group bias is a particularly strong or easy pathway to racism: "The correct thing to say is that, although we are not hardwired to be racist, humans are by nature groupish and exophobic, dispositions that can easily lead to racism."
In-group bias can occur along entirely arbitrary lines and has formed naturally among kids wearing red versus blue t-shirts, or being allocated into groups by adults based on their alleged fondness for one painter over another. Should we say that being groupish can easily lead to T-shirtism?
It's also not clear that in-group bias makes us innately exophobic. In fact, studies suggest people are mostly neutral / indifferent to "innocent" strangers. Exophobia kicks in when our ingroup feels in some way threatened - the defense mechanism is innate but needs to be triggered by a real or perceived threat, whether e.g. from predators or a rival human group.
While race itself may no longer be the barrier to success that it once was, as evidenced by the academic achievements of African immigrants in the UK and US, the insidious impact of sustained intergenerational poverty and underachievement has created a different form of social capital.
Institutional racism WAS a social construct in US history. It's a little-known fact that in the early days of the Virginia colony Black people were free to own property, serve in the defense of the colony and take cases to court - including, in the case of "Anthony Johnson. Negro" to secure the return of his slave. When the colony of Georgia (which included much of modern-day Alabama and Mississippi) was founded it had an absolute ban on slavery which lasted for decades, until soaring demand for cotton made planters "Stark mad after Negroes" in the 1750s.
The US wasn't directly responsible for most of the souls brought over from Africa on the Middle Passage, but the country did create an ideology around the righteousness of racial slavery in the US (unlike, bizarrely, the international Slave Trade). US breeding programs were framed as a moral good grounded in spurious claims of racial superiority, or more euphemistically, racial difference. JC Calhoun favorably compared the care of elderly retired slaves to the Dickensian conditions of the urban poor in slums - like these are my only options?
Attitudes to slavery and race tracked very closely to the mix of crops grown in each region. Sugar in Louisiana, "Carolina Gold" rice and cotton in Georgia all profited from the domestic breeding of slaves sold down-river from the Upper South. None of these regions even had a high rate of manumission before the Civil War, unlike Virginia, Maryland or Delaware, where wheat replaced tobacco as the main crop. All had Slave Codes that prevented free Black people from living there, and sumptuary laws that prevented Black people from dressing like white people.
That heritage of slavery and racism persisted through the twentieth century in Jim Crow laws. Black and white people were not allowed to marry in Virginia until 1967. DEI programs may have made inaccurate or controversial claims, but the legacy of racial prejudice and discrimination remains.