Editor’s note: The following is an edited version of a lecture that Professor Loury presented to faculty and students in MIT’s Department of Economics in October 2020.

Let me be provocative right at the start. George Floyd was killed in Minneapolis, Minnesota, by police officer Derek Chauvin. Chauvin is white, and Floyd is black. Was it a racial incident? What would we mean if we said it was such an incident, beyond the trivial statement that one of the participants was white and one was black? Well, we might mean that we think we know Chauvin’s motive when he put his knee on the man’s neck: that he acted out of racial animus. Alternatively, we might mean that people identify with the incident and interpret it in a particular way because of the race of the participants, quite apart from any discriminatory intent of the people acting in that situation. The fact is that the racial force of the incident is largely independent of causality and intentionality. Rather, it has a lot to do with interpretation, with narration.

There are Four Ps that I will use as my organizing principle for this talk about race and inequality in the United States.

Perennial. The problem’s been around forever. This is America.

Personal. I’m black. I’m from the south side of Chicago. These are my people that we’re talking about. How can I completely divorce that reality from the scientific imperatives? What’s my responsibility? How am I going to be read? If I speak out with a particular outlook, it’s going to be read in part in the context of my racial identity. People will understand that it’s a black economist, a black professor, a black intellectual, who says this or that. I can’t control that.

Political. The stakes are incredibly high when talking about race and racial inequality in the United States. You had people marching for Black Lives Matter in cities across the country, even across the globe. The presidential election was partly enmeshed in this argument going on within American society about race, systemic racism, white supremacy, black marginality, diversity and inclusion, equity, and all that—this is very political.

Perplexing. Because we do have problems here. We have a social-science problem. We have a challenge-to-the-country kind of problem. We’re 50 years past the Civil Rights movement. That’s almost as long a period of time as from Appomattox—where Robert E. Lee surrendered to Ulysses S. Grant—to Versailles, where the Europeans sorted out the mess that was World War I. Technology has completely changed over the last 50 years. The economy is completely different. Polity is completely different. Tens of millions of non-European immigrants have come to the country in the last half-century. Everything is different. And yet, if you look at some of the speeches that are being given, consider some of the events recorded for posterity in social media, some of the incidents taking place, and the arguments being made—and it’s as if we’re still back in the 1970s. Why is this so? It’s a puzzle.

Let me say something about my own biography. I grew up in Chicago in the 1950s and 1960s, in a working-class neighborhood. I came to MIT in the early 1970s. Before that I got a good education at Northwestern. Upon arriving at MIT, I discovered a few things. One was the deep structure of analytical economics, but I also learned that economics is a social science. It’s not divorced from policy, politics, society, or people. Paul Samuelson and Bob Solow and Peter Diamond and Franco Modigliani and many others—Stan Fischer, Marty Weitzman, Dick Eckaus, Frank Fisher: these were all among my teachers at MIT way back in the early 1970s. They cared about what was going on in the real world, not just about impressing their peers with the virtuosity of their technical practice. They addressed the great questions of their day. That lesson stuck with me.

I went on to teach at Harvard in the 1980s and at Boston University in the 1990s. I’ve been teaching at Brown since 2005. I was a black, conservative, public intellectual, for a while affiliated with the Reagan administration, and then I tacked back toward the center. Nowadays people would probably classify me as a conservative again because I’m a kind of contrarian on the issue of persistent racial inequality.

So that’s my setup. Racial inequality in America. It has been around for a long time. It is a deep, political question. It involves me personally. And it is a puzzle.

I want to preface my argument about persistent racial inequality by invoking the notion of narrative, by at least gesturing toward an appreciation for the power of the story and by noting that historical evidence does not pin down the story that we tell ourselves about the evidence. Indeed, multiple accounts can be consistent with the same facts. So, there is an inescapable element of choice about how we “narrate” those facts.

Recently, some prominent economists, UC Berkeley’s George Akerlof and Robert Shiller of Yale, for example, have also stressed the importance of narratives for understanding social outcomes. It is this viewpoint that I am invoking when I say that there are two opposing narratives on the persistence of racial inequality: the bias narrative and the development narrative.

Hands up, don’t shoot: that was Michael Brown, killed by a police officer in Ferguson, Missouri, 2014—the origin of the mainstreaming of Black Lives Matter. This is a singular event in recent history with respect to race relations and racial conflict in America. And it seems like it wasn’t “hands up, don’t shoot.” It really looks like Brown first attacked the police officer, who then shot him. The police officer probably feared for his life, and fired his gun. Two independent investigations, one by local authorities and one by the Justice Department, concluded that Michael Brown didn’t have his hands in the air when shot. Eyewitnesses have testified to this effect. My sense of the matter is that “hands up, don’t shoot” didn’t happen.

But it did happen virtually. It happened in effect. It happened because of the force of the narrative: a black man brutalized by overbearing, vicious, and racist state power—for many, that story overwhelmed all the facts in the case.

There’s a new documentary by filmmaker Eli Steele, narrated by his father, Shelby Steele, called What Killed Michael Brown? The film reviews the Michael Brown case and concludes that “hands up, don’t shoot” is what Shelby Steele calls a “poetic truth”—an account so powerfully resonant with a narrative paradigm that it may as well be true. Once it gets out there, many will have a hard time believing that it’s not true because the power of the narrative is so great.

Structural racism is a kind of narrative. What, after all, do people mean when they say structural racism? I think they mean that racially disparate outcomes are produced by a complex system of social interaction embodying historical practices that, in retrospect, were morally suspect, but that have taken on a life of their own with consequences that persist into the present. Mass incarceration, on this view, is structural racism because of the way that urban areas are organized, because of decisions that society has made about prohibiting trafficking in addictive substances, due to poor education and the inadequate economic opportunities for certain sectors of the society, all of which leaves many young people of color with fewer alternatives other than to engage in illicit activities.

They mean something like that, I think. They don’t mean that there’s a conspiracy somewhere trying to figure out how to hurt blacks. They are not talking about racism in the sense that the General Social Survey measures, when it asks questions like: “How do you feel about having neighbors or having your child marry someone of a different race?”

Still, I am not a big fan of the structural racism narrative. I think it is imprecise; I think that those who invoke structural racism are begging the question. I want to know exactly what structures, what dynamic processes, they mean, and I want to know exactly how race figures into that story. Often the people using this kind of language do not tell me this. History, I would argue, is complicated. So, racial disparities must have multiple, interwoven, interacting causes that range from culture, politics, and economic incentives to historical accident, environmental factors and, yes, the nefarious doings of individuals who may be racists, as well as systems of law and policy that are disadvantaging to some racial groups without having so been intended. So, I am often left wanting to know just what they are talking about when they say, “structural racism.” Often, use of the term seems to be expressing a disposition while calling me to solidarity, asking for my fealty, for my affirmation of a system of belief. It is only one among many plausible narratives.

If we restrict ourselves to the labor market and just talk about wages, then the structural racism narrative would be all about the demand side of the labor market. It would be about: what do employers do? What kind of information do they have? What contracts are they willing to enter into? What are the training opportunities being offered inside of organizations for employees to move ahead? Fixing this situation means anti-discrimination enforcement. We need a change of hearts and minds, on this view. We need implicit-bias training. That’s all on the demand side, where racial inequality is due to racial discrimination, and is best understood via the bias narrative.

I am offering instead, as a counterpoint to the bias narrative, what I am calling the development narrative, which stresses that patterns of behavior within the disadvantaged population need to be looked at. I speak now about African-Americans, about 35 million or 40 million people in the United States. This, of course, is a variegated, differentiated, and heterogeneous population. One size does not fit all. Nevertheless, I am willing to ask: are there patterns of behavior observable in certain communities of color that have the consequence of inhibiting the development of human potential?

Here is an illustration of why the distinction between these narratives might be important. Consider school discipline. I call attention to the Department of Education policy under the Obama administration of admonishing school districts that reported racial disparity in the frequency with which students were suspended from school for disruptive behavior. The statistics reveal that black students get suspended more often relative to their numbers. You can look at the average frequency of suspension for black and white students in a school district, that is, and you can see a disparate incidence of suspension by race.

Obama’s Secretary of Education, via the Office of Civil Rights in the Department of Education, sent a letter to local school districts warning them that they should be aware of and take efforts to reduce this disparity, or they might find themselves subject to a civil rights investigation for racial discrimination.

Now, there is indeed a disparity, and it’s nontrivial. If it reflected the differential behavior of the school districts—principals, teachers, and security officers—in how they treated disruptive behavior, such that the same behavior by a white student would be met with a less punitive response—then that would, indeed, be alarming and would warrant the attention of the authorities to do something about it. That’s one possibility.

Another possibility, however, is that disruptive behavior occurs more frequently among black students for reasons that lie outside the school. If that’s the case—if the problem is on the supply side of this “market”—then interpreting disparate suspension rates as evidence of racial bias and responding to that by disciplining the school districts, cutting off their funding, perhaps hauling them into court, would be a terrible mistake. Rather, one would want to address the sources of this behavioral differences. One would certainly not dismiss the disparity, but one would address the disparity by attempting to enhance the opportunities or the experiences of the affected young people, which shape their behavior patterns, so as to make those students less subject to disciplinary measures. (There are other possibilities. For example, one might become more tolerant of disruptive behavior across the board because a punitive reaction to disruption could be predicted to generate an unacceptable racial disparity. One can go many places with this example, but I’m using it here merely to illustrate the differences between the bias narrative and the development narrative as ways of responding to the fact of a racial disparity.)

Let’s talk more specifically now about the development problem. I’m willing to invoke the demographic observation of a high rate of single-parenthood in African-American families, where a mother is raising kids on her own. Three in four black kids, 70 percent, something like that, are today born to women without husbands. Common sense suggests that this reality can’t be unrelated to some of the outcomes, like disruptive behavior, that concern us. Perhaps it is not the main factor, but it would be an important part of the picture when talking about persistent racial inequality. The fact that I am willing to take it onboard does not, however, answer the question: what is the causal mechanism? A historical sociologist, historian, or demographer well might argue that you have these different organizational patterns within families, but they are explicable in terms of the historical experience of the respective groups. For Orlando Patterson, a sociologist at Harvard, they are a result of slavery—of the fact that families were disrupted at their core by the intercession of the master’s property claim over and against the filial and familial connections of natal bonding. It is impossible, on this view, that you could have had as intrusive an intervention into intimate social relations among African-descended people as was slavery and not see some present-day familial consequences.

Family organization matters. There is a big racial disparity in family organization. Therefore, part of the story that you need to tell to account for persisting racial inequality involves family organization. In saying that, I would not have precluded an historical argument about the sources of the family organizational patterns. I would simply have been willing to consider the supply side as well as the demand side when trying to understand persistent racial inequality. This narrative is fiercely resisted by many, but I am urging here that we consider it.

Violence, murder, homicide—huge racial disparities exist in this area. Everyone can read the newspapers. This is a reality of the contemporary urban scene. And there’s a tightly networked set of social connections among the people who are committing and are victimized by much of this criminal violence. Is that phenomenon, in any straightforward way, a manifestation of bias—of racism? Could it really be about white supremacy? Or is it about the failure of some part of a population to be socialized with the restraint, self-discipline, and commitment to civil behavior that, when widely embraced, make ordinary life and commerce in a community possible?

A willingness to ask about the behavior of the violent criminals preying on their neighbors, and the sources within a community of such behavior, is part of what it means to take seriously the development narrative. Again, I am not saying that we should forego trying to do anything about it, that policy has nowhere to go since the problem is mostly on the development side. Policy obviously has a lot to do with the development side, from better education to subsidizing child development to improving parenting skills. But we need to take seriously these patterns of behavior and their cultural antecedents.

Everyone talks about the academic achievement gap. Several groups are suing Harvard University, saying that the school’s affirmative-action practices are penalizing Asian-Americans. And the special high schools in New York City are being pressured to change their selection criteria, so as to ensure that they don’t enroll a class of more than 1,000 first-year students and have only a handful of black kids among that cohort. If you look at the National Assessment of Educational Progress, where a representative sample of American students are regularly tested for their cognitive abilities in mathematics and writing, you can see huge racial disparities in those data.

Am I willing to consider the supply side when I talk about that? Am I willing to ask: what’s going on in the homes? And: what do peer groups value? Am I willing to measure how much time people spend on homework? How many books there are in the home? Is the large disparity by race in academic achievement better understood when it is viewed in terms of the bias narrative or the development narrative?

If you are prepared to discuss the supply side—if you are prepared, that is, to talk about the extent to which members of a disadvantaged, marginalized and oppressed group are implicated in their own disadvantage—then some will charge that you are “blaming the victim.” I reject that charge categorically. It is not assigning blame to simply observe that the labor market has a supply side; that people make choices and engage in behaviors having deleterious consequences for their future economic prospects.

Of course, those behavioral patterns well may be the consequence of structural conditions and historical dynamics. On the other hand, if the reflexive response to seeing any disparity of behavior is to say: “Well, this is simply due to historical exigency,” then that has its own moral and philosophic implications in regards to agency—i.e., the extent to which people can be presumed to control their own fate, and the extent to which their communal norms and ways of living are seen as being within their ability to change.

For instance, is it a necessity that the homicide rate be as high as it is in the black communities we talk about when discussing racial inequality? Is that really how we want to talk about such matters—to say, “What can they do? Of course, there is a high level of violence. Look at our structures; our gun laws; our hypocrisy about drugs consumption and trafficking. Look at our history of racism in this country. Of course, there’s going to be a higher level of violence.” It is, in my view, morally repulsive to impute such a lack of agency to people in this fashion. It infantilizes them, makes them mere puppets at the end of strings being pulled by others. In the extreme, it robs them of their human dignity.

And perhaps worst of all, it robs a group of the ability to make social judgments. It undermines the capacity to clearly delineate right and wrong ways of living and to urge that individuals live rightly. I am not a philosopher, but I have read the Foundations of the Metaphysics of Morals several times, trying to understand what brother Immanuel Kant was talking about. I understand him to be making a principled argument for the capacity to have a theory of morals. While it is certainly true, he says, that we are all embedded within the flux and the flow of history and under the influence of forces that are beyond our control of environment, psychology, and such, nevertheless, the theorist must assume the capacity of individuals to make free-will choices about their moral life, lest there be no possibility for any theory of morals whatsoever.

I am signing on to that argument here when insisting on the necessity to engage the development narrative alongside talk about bias; the necessity for calling attention to patterns of behavior and value that are internal to a community which limit their success; and when defending myself against the accusation that I give aid and comfort to racists, or that by making these observations I am somehow blaming the victims for their plight.

I am not unmindful of the pitfalls. I can hear the retort: “But, what will the racists say if you talk like that? Whatever the merits of such a narrative, in a society like the one that we live in, where many people are much less sympathetic than are you to the well-being and the aspirations of black people, some will take your words—the words of a black man—as license to entertain their own racist thoughts about why racial inequality persists.” I cannot prove all this scientifically. But between the two paths—withholding arguments I believe true in order to manage political discourse, versus giving voice to such insight as I think that I might have, subject to rebuke, repudiation, and refutation by other critics, so as to enliven and enrich the political and public discourse—I choose the latter course. I am willing to take the risk of telling the “truth,” as best I can discern it.

One other reason to be honest about what is going on, on the development side of the equation, is that everybody can see it. People are bluffing when they say, “oh, I’m not going to talk about the black family. Out-of-wedlock birthrates don’t matter.” People are bluffing when they say, “We’re Black Lives Matter, and we’re about cops killing kids, but we have nothing to say about kids killing kids.” Everybody can see what is going on.

The fact is that, as long as race is a meaningful part of people’s identity in society and they reproduce those meanings through their patterns of association, then you are going to get some disparity by race in the structure of the social networks in which people are embedded. And when network-mediated spillovers in human capital development are important, this means there will be some persisting racial disparities of social outcome.

What about affirmative action and reparations? I have concerns—grave concerns—about these policies. I want briefly to give some hint of what it is that I am concerned about, which reveals something about my larger outlook on the age-old American dilemma of racial inequality.

I’m against slavery reparations for a few reasons. One is, okay, when the Japanese Americans interred by the Roosevelt administration during the Second World War were finally, in an act of Congress signed into law by Ronald Reagan, acknowledged as having been wrongly victimized and offered a token reparation payment, it was $20,000 a head for 80,000 people. That’s $1.6 billion, paid out of the Treasury—and it should have been paid. I have no problem with that. By contrast, there are 35 million or 40 million African Americans, and if you take the modern equivalent of 40 acres and a mule, and you bring it forward at a normal rate of return, we’re reaching astronomical sums. Maybe it is $100,000 a head, with inflation, for 40 million people. That would be $4 trillion, compared with 80,000 people and $1.6 billion.

Here’s what I’m saying. Enacting reparations for slavery would be to create a Social Security-level-of-magnitude fiscal/social policy in America, the benefits from which would be based on racial identity. That, quite simply, is a monumental mistake. It’s South Africa-esque. Our government would have to classify people and enact statutes and administer law based on people’s race. We ought not go down that path. That is the overarching moral argument that I would make.

My practical argument is that remedying racial disparity ought to be left as an open-ended commitment. True enough, this problem—which is due in no small part to our bitter history of slavery and Jim Crow segregation—must be addressed. But, in my view, it would not be the smartest thing in the world for black Americans to cash out that obligation; to have a transaction where, metaphorically speaking, we sit on one side of the table with our moral capital, where America as a whole sits on the other side with its checkbook, and a transaction is negotiated wherein the “debt” gets discharged. We ought not to be in a hurry to commodify that obligation, I would say. For then, when confronted with lingering racial disparities, the country can say “you’ve all been paid.” Rather, what we should do is to take our moral chips, combine them with other progressive political initiatives, and aim to create a decent society for everyone, whether that concerns health care, housing, food security, employment, education, or old-age security. Were these efforts sufficiently robust on behalf of everybody, the most pressing concerns about racial disparity (having to do with extreme deprivation) would be ameliorated and we will have lent our moral capital to the right cause—not a racially defined reparation, but rather a humanely defined improvement in the quality of the nation’s social contract.

One final word about affirmative action. We are now 50 years down the line with this policy. It has been institutionalized. Diversity, Inclusion, Equity and Belonging: in practice what that means is affirmative action. I have a concern, though, which is that equality of representation, when you are in the most rarified venues of selection, is in competition with equality of respect. I’m specifically referring here to selecting at the 95th percentile—the right tail of the distribution of talents, not the population median.

It is impossible that there would not be post-admissions performance differences by race in students selected at this percentile if racially different criteria of selection are used pre-admission, so long as those criteria are correlated with performance. And, if the criteria—SAT test scores, grades, advance placement tests, quality of essay, letters of recommendation, whatever indicia of performance you want to use—are not correlated with post-admissions performance, then they shouldn’t be used. But they are being used because we all know that they are correlated with post-admissions performance to some degree.

I invite you to look at the data produced by discovery in the Harvard case, for example, to see the huge disparity in academic preparation characteristic of applicant populations by race to Harvard University in recent years. There’s going to be different post-selection performance if those criteria are correlated with performance, and that’s what we see. What is the consequence of that? Either we will acknowledge the difference in post-admissions performance; or we won’t; we’ll cover it up by flattening assessment criteria and, in effect, pretending it’s not there. The dishonesty can be stifling in my view. I’m in the economics department, so let me talk in terms of economics. My point: Right-tail selection plus racially preferential selection is inconsistent with true equality. It will get you representation, perhaps, but it won’t get you equality—at least not equality of respect.

You need a closely approximating parity of performance to get equality of respect. But you’ve applied different levels of selectivity into a highly competitive and elite activity, where the selection criteria are correlated with post-admissions performance, so you’re getting disparities in performance post-admission that you’re not owning up to, or that you’re covering up.

So, many have observed that there are not enough black economists on the faculty of leading universities. We can do better. We should be more diverse and inclusive at the top departments in the country. There should be at least two blacks at each one, let’s say. Maybe I can agree with all of that. But suppose there are just not enough top-flight black economists to go around. If the way to “do better” is to make the criteria of selection into this rarified enterprise of academic economics, at the top, depend upon the racial identity of job applicants, then you’re not going to get equality. Instead, you’re going to get some degree of black mediocrity. This fact is currently unsayable. It is unsayable to observe openly that there could be racial differences in performance in venues such as this. Yet, I get emails all the time. “I’m a partner at a big law firm in New York City. Here’s what I can’t say publicly. Please don’t quote me. Many of our associates who are of color are not up to snuff, but we hired them anyway because . . . Some of them are going to make partner here, and I shudder at that prospect.” This is not equality of respect.

Here’s what we ought to do instead. We should devote our efforts to enhancing the development of African-American prospects, such that when you apply roughly equal criteria of selection at the right tail, the numbers of blacks selected still goes up, but based on achievement. You don’t increase the population of applicants by changing standards in order to achieve racial parity—that is a huge mistake.

Further, we don’t have population parity in every pursuit. How can you expect population parity in an enterprise when there are some groups (Asians? Jews?) who are overrepresented by a factor of two or three relative to their population? You cannot get population parity with equal criteria of selection when all the groups are not feeding into the pool of qualified applicants at the same rate in every activity.

My view is that the permanent embrace of preferential selection in extremely selective, competitive venues by race is a mistake. I can understand its transitional use, historically speaking, but its institutionalization is inconsistent with true equality.

I have told you what I am against: elite affirmative action that uses different standards for selection of blacks and other people, and reparations, in the broad sense of America repaying a debt to black people—for the reasons that I adduced. But what am I for?

Educational opportunity, for starters. Here’s what I would say about it. One principle of equality in the provision of educational services is influential these days. According to this principle, because local districts differ in the value of real estate and hence in their tax basis, and so are not equally situated for spending on kids’ education, the state should, through its revenue-transfer programs, redistribute resources among those districts, so as to equalize the expenditures per pupil.

But one could think about a different principle, something like equal effective educational opportunity, where the goal is to acknowledge that different districts are differently placed—not with regard to real-estate values alone, but with respect to the social conditions of the students there. So, a district with lots of disadvantaged students—more special education, more behavioral problems in the classroom, less resources at home, economic disadvantage, food insecurity, things like that—may require you to spend more per pupil there, if the goal is to try to equalize the effective educational opportunity of all students. This would be a different kind of equality principle to bring into the educational sphere and achieving it may require more—much more—than merely shifting funds between districts.

Furthermore, I would say that we ought not worry about educational opportunity for Americans primarily in terms of the fact that African-Americans are disproportionate among those ill-served by educational opportunity. It feels to me like the proverbial tail wagging the dog, to make social policy in a country of 330 million people on the basis of an effort to rectify the historically inherited racial disparity that is affecting a quarter to a third of the African-American population. (Mind you, now, we are not saying that every person of African descent is fundamentally disadvantaged purely because of the color of their skin.) Remedying racial disparity ought not to be the primary motive when making social policy. I would argue this not only from a political perspective but also from a moral perspective. I think the right theory of social justice is one in which any person’s idiosyncratic demographic characteristics should not have any bearing on the weight the social decision maker gives to that person’s welfare when formulating policy. That is to say, ultimately some version of trans-racial humanism is the right philosophical stance.

Photo by Scott Olson/Getty Images


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next