When, in the spring of 1999, the call came from Skip Gates inviting me to present the prestigious Du Bois Lectures at Harvard’s African and African American Studies Department the following April, I was delighted. I fancied myself a proverbial prodigal son, as the event marked my triumphant return to Harvard nearly a decade after having moved across the Charles River to Boston University, seeking greener pastures in the midst of a midlife crisis. In retrospect, I was determined—cravenly so, it is now clear—to prove how stridently at odds I had come to be with American conservatism on racial issues. Indeed, it pains me now, some two decades on, to realize how badly I needed acceptance and affirmation from Harvard’s “Negro cognoscenti.” Presenting those three lectures, “The Economics and Ethics of Racial Classification,” in April 2000, and seeing them published by Harvard University Press as The Anatomy of Racial Inequality in 2002, confirmed that I had, at long last, “come home.” I can still feel the warm glow of satisfaction that washed over me while being introduced before each talk—in flattering terms, to overflow crowds—by the eminent black scholars William Julius Wilson, Lawrence Bobo, and Kwame Anthony Appiah. Here, finally, was my chance to deploy all of my social scientific and expository powers to inveigh against the evils of racial inequality and to defend the dignity of “our people.” I was determined to atone for my apostasy and, so it now appears upon rereading the text, I succeeded admirably! Such was the emotional impetus driving me to write this book.
But the arguments on offer here sprang from my head as much as from my heart. Again, my time at Harvard in the 1980s was crucial. For that is where I encountered Thomas Schelling, the Nobel Prize winning economist, who passed away in 2016. We worked together for nearly a decade at the Kennedy School of Government, and I learned much from him. We created and co-taught for several years a course called Public Policy in Divided Societies. There I encountered and read deeply writers such as Amartya Sen (“Rational Fools”); Albert Hirschman (Exit, Voice, and Loyalty); Erving Goffman (The Presentation of Self in Everyday Life); David Lewis (Convention); Leo Strauss (Persecution and the Art of Writing); Kenneth Arrow (Social Choice and Individual Values); Robert K. Merton (“The Self-Fulfilling Prophecy”); Howard Raiffa (Negotiation Analysis); Mancur Olson (The Logic of Collective Action); Michael Spence (“Job Market Signaling”); Harold Isaacs (Idols of the Tribe); Jon Elster (Sour Grapes); and Michael Walzer (Spheres of Justice). This experience of teaching with Tom more than three decades ago laid the intellectual foundations for this book. It broadened my horizons beyond the narrow confines of technical economic theory. It showed me the power of strategic analysis to illuminate questions about identity, inequality, social ethics, and imperfect rationality. It inspired me to study in comparative spirit a wide range of real-world phenomena.
With Tom’s encouragement, our students investigated topics such as the plight of Roma in Europe and the indigenous in Central America; caste in India and slave maroon communities in the Caribbean; skin color hierarchies in nineteenth-century cities like New Orleans and Charleston; debates among the Deaf community over sign language versus lipreading; changing one’s name or accent to disguise ethnic/regional origins; collective punishment; group-based feelings of pride and shame; the problems any group faces to maintain a public reputation; inequality-promoting implications of endogamy and assortative mating; racial profiling; complementarity between sexual divisions of labor at home and in the workplace, and so much more.
Beyond this, Tom Schelling helped me see how an open-minded economist can theorize rigorously about such ubiquitous social phenomena as rumors; seduction; plausible deniability; dog-whistle politics; riots; signaling; “passing for white”; usage of expressive imprecision; group think; code words; discursive taboos; naked emperors; one’s knowledge of the other’s state of knowledge; the subtle differences between promises, threats, and bluffs, and so on. Such was the intellectual feast I enjoyed in the 1980s, which empowered me years later to write these lectures.
As for the book itself, The Anatomy of Racial Inequality sought to do three things: outline a theory of “race” applicable to social and historical circumstances in the United States; sketch an account of the stubborn persistence of racial inequality in America; and offer a conceptual framework for social criticism on race-related issues, so as to influence political and intellectual elites on behalf of a program of reform. Today I stand by my analysis of the first two themes, but I have come to have considerable misgivings concerning what I had to say about the third one.
Looking back, with two decades of hindsight, I remain proud of what I accomplished in this book, notwithstanding its emotional origins in what was a jumble of guilt, regret, self-doubt, and virtue signaling. It is rigorously argued. It explores deep problems in economics, sociology, politics, and social ethics. And, though there is not a single equation in the text, it illustrates how much light an economist’s formal framing can shed on one of the central moral questions of our time. Although I would not argue in exactly the same way today, for reasons discussed below, I have decided to resist the temptation to rewrite history. So, the original text of my 2002 book is presented here, unamended. Racial stereotypes, racial stigma, and racial justice—these were the themes of my lectures. Permit me to introduce this newly released edition with a brief discussion of each.
Any theory of “race” must reckon with the fact that people take note of and assign significance to superficial markings on the bodies of other human beings—their skin color, hair texture, facial bone structures, and so forth. This universal human practice appears to have a deep neurological foundation. This is the point of departure for my analysis. I refer to a society as being “raced” when its members routinely partition the field of human subjects whom they encounter in that society into groups, where this sorting convention is based on the subjects’ possession of some cluster of observable bodily marks. This leads to my claim that, at bottom, “race” is all about “embodied social signification.” Let us call this the social-cognitive approach to thinking about “race.” It can be usefully contrasted with an approach derived from the science of biological taxonomy, in which one classifies human beings on the basis of the natural variation in genetic endowments across geographically isolated sub-populations. Such isolation was a feature of the human condition until quite recently (on an evolutionary time scale), permitting there to be some independent biological development within sub-populations that might be thought to have led to the emergence of distinct “races.”
Of course, this idea of “biological race” is controversial. When philosophers deny the reality of “race,” they have in mind this biological-taxonomic notion, and what they deny is that meaningful distinctions among contemporary human sub-groups can be derived in this way. Whether they are right or not is a scientific question that lies beyond the scope of this book. But, whatever the merits of that dispute, I am at pains to stress that the validity of racial classification as an exercise in biological taxonomy is conceptually distinct from the validity (and relevance) of my concern with racial categorization as an exercise in social cognition. The social convention of thinking about other people (or oneself, for that matter) as belonging to different “races” is a deeply ingrained feature of our political culture that has taken on a life of its own. As I see it, the logical exercise of deconstructing racial categories by arguing that nothing “real” lies behind them is largely beside the point. Thus, my approach to the concept of “race” is cognitive, not normative. This perspective is supported by the theory of “self-confirming stereotypes” that I advanced in my first Du Bois lecture. My point here is a subtle one, so permit me to explain it a bit further.
Imagine that people believe fluctuations of the stock market can be predicted by changes in sunspot activity. This may be because, as an objective meteorological matter, sunspots correlate with rainfall, influencing crop yields and thus affecting the economy. Or, alternatively, solar radiation might influence the human psyche somehow, so as to alter people’s behavior in the securities markets. Each of these two accounts postulates an objective causal link between sunspots and stock prices, analogous to grounding one’s cognizance of “race” on the validity of a race-based biological taxonomy. Still, even if no such objective link as this between sunspots and stock prices exist, so long as people believe in the connection, monitor weather on the sun’s surface, and act based on how they anticipate security prices to be affected, then it is possible that a real link between these evidently disparate phenomena can emerge from subjective perceptions of market participants. As a result, belief in the financial relevance of sunspot activity will have been rendered entirely rational.
Likewise, no objective basis for a racial taxonomy is necessary in order for the subjective use of racial categories to be warranted. For, if a person knows that others in society will classify him based on certain markers, and if such classification impacts his material well-being, then it will be a rational cognitive stance on his part—not an unfounded belief and certainly not a moral error—for him to think of himself as being “raced.” In turn, that he thinks of himself in this way and that his societal peers are inclined to classify him similarly can provide a compelling reason for any newcomer to that society to adopt this ongoing scheme of racial classification. (Learning the extant “language” of embodied social signification is a first step toward assimilation for the foreigner, or the newborn, into any “raced” society.) So, I conclude that “races,” in this social-cognitive sense, may come to exist and to be reproduced over the generations in a society, even though there may be no “races” in the biological-taxonomic sense. Therefore, calling attention to a dispute on the scientific existence of races need have no bearing on the legitimacy of the social practice of racial classification. I conclude that, to understand how “race” operates within the American social hierarchy, my decision to define it as “embodied social signification” is both logically coherent and analytically useful. That is, I take “race” to be a social convention and I approach it from a cognitive, not a normative, point of view.
For me, then, the term “race” refers to indelible, heritable marks on human bodies—skin color, hair texture, facial bone structure—of no intrinsic significance, but that nevertheless have through time come to be invested with social expectations and social meanings that are more or less durable. Two distinct perceptual processes are implicated: categorization and signification. Categorization entails sorting people into a small number of groups based on bodily marks so as to differentiate one’s dealings with such persons accordingly. It is a basic effort to comprehend the social world around us. By contrast, signification is an interpretative act—one that associates certain connotations, or “social meanings,” with the categories. So, informational and symbolic issues are both at play. Or, as I put it in the book, when we speak about race, what we are really talking about is “embodied social signification.”
This brings me to the subject of my second lecture—racial stigma—which I see as the central innovative concept in this book. I eschewed use of the word “racism” not because I saw no moral problem, but because I thought the word imprecise. More useful, I argued, was my core concept, racial stigma, and the related notion of “biased social cognition.” By invoking these ideas, I was trying to connect two things: that people use racial classifications when interacting with one another, and that the “social meanings” associated with racial categories can determine how they account for any disparities they see in the social world around them. It was this linkage between perceptions of “race” and the implicit causal accounts of racial inequality that interested me. I was asking: How does the “race” of those enduring problematic social circumstances affect whether or not powerful observers perceive there to be a social problem, and what implications for our politics follow from this?
The effect I was after when talking about racial stigma, and the reason I employed the seemingly loaded phrase “biased social cognition,” was to identify the politically consequential cognitive distortions that occur when the observably inferior position of a racial group is thought to have emerged from qualities intrinsic to that group, even though the actual causation entails a system of social interactions. A disparity will not constitute a “problem” when observers see it as resulting from the deficiencies of those lagging behind. It is a species of “racial essentialism,” I argued, if observers refuse to see the systemic and endogenous interactions that lead to bad social outcomes for African Americans, and if they instead attribute those results to exogenous factors that are taken to be internal to the group and beyond political remedy. My principal argument in The Anatomy of Racial Inequality is that—due to our history of African slavery and the enduring stigma attaching to “blackness”—the risk of this particular causal misattribution for black people is great. This is a theory of “systemic racism”—using the language of our time—that I advanced two decades ago, and which I believe remains relevant to this day.
I can best make the point with a nonracial example. Consider gender inequality. We know that there is disparity in the social outcomes for males and females in two different venues—schools and jails. When compared to girls, boys are often overrepresented among those doing well in math and science in schools, but also among those doing poorly in society at large by ending up in jail. There is evidence to support both of these suppositions, but only the first is widely taken to be a problem for public policy. Why? My answer is that it offends our basic intuition about the propriety of underlying social processes when boys and girls attain different levels of achievement in the technical curriculum of our schools. We may not be able to say exactly why this outcome occurs, but we instinctively know that it is not right. In the face of this disparity, we are inclined to interrogate our institutions—that is, to search the record of our social practice and examine the myriad possibilities so as to find out where things have gone wrong. Our baseline expectation is that equality should prevail here. Our moral sensibility is offended when it does not. An impetus to reform is spurred thereby. That is, we cannot readily envision a legitimate sequence of events that would produce the disparity, and so we set ourselves the task of solving a problem.
Gender disparity in rates of imprisonment occasions no such disquiet because, tacitly if not explicitly, we are “gender essentialists.” That is, we think boys and girls are different in some ways that are relevant to explaining the observed disparity—different either in their biological natures, or in their deeply ingrained socializations. (Note well, the essentialism with which I am concerned need not be based solely or mainly in biology. It can be grounded in [possibly false] beliefs about profound cultural difference as well.) As “gender essentialists,” our intuitions are not offended by the fact of vastly higher rates of imprisonment among males than females. We seldom ask any deeper questions about why this disparity has come about. In this sense, we do not perceive there to be a problem, and so no solution is sought. That is, categories of classification and their associated significations influence our moral intuitions and thereby can either temper or enhance our inclination to demand structural reforms.
Now, we may be right or wrong to act as we do in these gender disparity matters, but my point with the example is to show that the bare facts of disparity do not, in themselves, suggest a course of action. To act, we must marry the facts we observe to some model of social causation. This model need not be explicit in our minds. It can and usually will lurk beneath the surface of our conscious reflections. Still, it is the facts plus the model that lead us to perceive a given circumstance as indicative of some as yet undiagnosed failing in our social interactions, or not. This kind of reflection on the deeper structure of social-cognitive processes bearing on the issue of racial disparity is what I hoped to stimulate with my discussion of biased social cognition. And the role of “race” in such processes is what I alluded to when I talked about racial stigma.
In my third lecture I offered a critique of colorblindness and endorsed the practice of affirmative action. I argued that colorblindness as a philosophical approach to public policy was inadequate. It alone could not accommodate the task of redressing racial injustice. This is a view I would still affirm today. I took some pains, however, to explicitly distance myself from many of the vociferous critics of racial preferences in the 1990s, including Justice Clarence Thomas, the writers Abigail and Stephen Thernstrom, and critics like Ward Connerly and Shelby Steele. In doing so, I reversed the position against affirmative action for which I had been known earlier in my career. This, I have since come to believe, was a serious error. Chalk it up, if you like, to my zeal twenty years ago to mend the breach with my academic co-racialists. In any case, I find myself these days in a sustained argument with the zeitgeist about the nature of the country in relation to the unresolved issue of racial inequality, exclusion, and discrimination. I think the discourse has gone badly off the rails, especially on issues like affirmative action and reparations for descendants of slaves. In previewing the now dated argument in my twenty-year-old book, I would like to briefly explain why.
Broadly speaking, two conflicting narratives about race and inequality compete for our attention: the “bias narrative” and the “development narrative.” This book proceeded primarily by embracing the first, but I have since come to believe that the second is the more relevant for our time. How one frames the “racial justice” issue depends on which account one emphasizes. The bias narrative stresses how white supremacy has done us black folk wrong. America “has its knee on our necks,” so to speak. Present-day anti-black racism and a litany of historical injustice are the focus. The death of an unarmed black man at the hands of police becomes the twenty-first-century extension of a lynching in Mississippi circa 1955. By contrast, the development narrative puts in center place the incomplete project of empowering African Americans who have been adversely impacted by history to acquire those capacities of functioning and performance that can allow for effective competition on what is basically a level playing field. A first-order imperative, on this account, is to abet the acquisition of the skills and the acculturation of the behaviors that afford our youngsters a greater chance for success in life. These are radically different ways of looking at the underrepresentation of African Americans, especially in the most remunerative lines of work and most selective venues of higher education. Of course, these narratives are not mutually exclusive, but I have come to believe that it matters greatly for our politics which account one chooses to emphasize. My current view is that development is by far the more crucial concern.
I set this argument within historical context: African Americans—emerging from slavery, Jim Crow segregation, and widespread discrimination—have as a group been diminished in terms of the development of our competitive and productive capacities. Generations of unequal schools and of employment and housing discrimination have impeded the realization within the African American population of our latent potential. Given such a history, one cannot expect that, at Day One, there would have been equality of the various indicia of achievement, since the background conditions did not afford us an equal opportunity to realize our full human possibilities. This is the status quo ante from which affirmative action emerged a half-century ago—the baseline from which we have endeavored to move towards something more equal. In the face of this historical inheritance, some policies of racial preference could, I think, be justified.
But time does not stand still, and we now face a difficult historical challenge. We are at a crossroads and need to reevaluate. In keeping with my argument from twenty years ago, I do not here object to affirmative action in principle, seeing it as unjust racial discrimination in reverse. It was both necessary and proper to rely upon some preferential treatment for descendants of the slaves as a temporary, stopgap mechanism. But the fundamental developmental deficits that history has bequeathed to us must also be addressed if African Americans are to compete effectively going forward. And effective competition is the only path to genuine racial equality. We are not facing up to this challenge. Instead, we have settled for a second-best remedy: less exacting criteria for selecting African Americans. It is a form of “cheap grace” for elite gatekeepers to use preferences as a cover for the consequences of our racially unjust history. Ultimately, black people are going to be the ones left with the short end of this deal, because relying on racial preferences in elite venues invites patronization, condescension, contempt, dishonesty, shame, and a general erosion of standards precisely in places where the competition for status is most keen. This may yield racially proportionate representation, but it is a long way from achieving genuine racial equality.
As for slavery reparations, this policy may seem to be a logical extension of the forgoing argument. But, stipulating that history casts a long shadow, I nevertheless think this is the wrong way to frame the problem. It is natural to try to monetize the racial depredations of history, but I nevertheless think this is a mistake—for black people ourselves, and for the country as a whole. As a practical matter, no sum of money would be large enough to indemnify those suffering the ill effects of our unlovely racial past from the consequences in their lives and the lives of their children of what history has wrought for them. However, enactment of monetary transfers based on racial identity in the name of reparations would, in effect, commodify the claim that African Americans have on the general public’s attention to the problems besetting our communities. It would constitute a transaction on the basis of which we will have been paid in compensation for what history has wrought. It would be unwise, I think, to discharge that moral obligation in this way. I would advocate for interventions aimed at reducing the most deleterious effects of our racist past, but I object to sitting across a bargaining table with the rest of America—us with our moral capital on one side, the country with its financial capital on the other side—pushing that moral capital into the middle of the table and receiving money in exchange. The ability for us to command the larger society’s attention to the certain-to-be-ongoing social maladies afflicting so much of black America will have been significantly diminished. Far better, in my view—more ethical and also more politically effective—would be to take our moral capital, combine it with the influence of others on behalf of creating of a decent social contract for all Americans, and work hard get that implemented. If that program were to succeed, then bad schools in African American neighborhoods would get attention; jails overflowing with black people would become a big social issue; and the needs of families, employment issues, health care concerns, and so forth would be attended to. They would be attended to for all Americans, and therefore for those African Americans most in need. Do not try to cut a separate deal with America is my message to black people. That is a mistake.
I am reminded, amidst the contemporary turmoil, of the period after emancipation, more than 150 years ago. A brief moment of pro-freedmen sentiment during Reconstruction in the immediate aftermath of the Civil War was washed away, and the long, dark night of Jim Crow emerged. Blacks were set back. But, in the wake of this setback emerged some of the greatest achievements of African American history. The freedmen who had been liberated from slavery in 1863 were almost universally illiterate. Within a half-century, their increased literacy rates rivaled anything that has been seen in terms of a mass population acquiring the capacity to read. That significant achievement helped to bring them into the modern world.
We now look at the black family lamenting, perhaps, the high rate of births to unmarried mothers. But that is a post-1960s phenomenon. The health of the African American social fiber coming out of slavery was remarkable. Businesses were built. People acquired land, educated their children, and developed skills. They constantly faced opposition at every step along the way —“no blacks need apply”; “for whites only,” etc.—but nevertheless they built a foundation from which could be launched the civil rights movement in mid-twentieth century that changed the politics of our country forever. As my friend Robert Woodson is fond of saying, “When whites were at their worst, we blacks were at our best.” What I want to declare here is that we can do it again!
This, then, is my final observation, as I reflect on the twentieth anniversary of the publication of my Du Bois Lectures: If we African Americans want equal dignity—if we want racial equality in its truest and highest sense—then we must realize that “white America” cannot give us what we seek. We actually have to earn equal status. Please understand that I am on the side of black people when saying this. But, at this late date, I feel obliged to report that equality of dignity, standing, honor, and security in one’s position in society—the ability to command the respect of others—these are not things that can be demanded and then simply handed over. Rather, these are thing that we must wrest from a cruel and indifferent world with our bare hands, inspired by the example of our enslaved and newly freed ancestors. We must make ourselves equal. No one can do it for us.