Episode 31: Race, Genetics, and a “Most Dangerous Myth”

The concept of distinct races came from European naturalists in the 1700s. It’s now recognized as a social construct, rather than a biological classification. Nonetheless, genetics researchers sometimes use race or ethnicity to stand in for ancestry. This practice has been criticized for creating discrete categories where none exist and for underemphasizing the ways that environment and other nongenetic factors can contribute to ill health.

In March, the National Academies of Sciences, Engineering, and Medicine weighed in with a consensus report. It documented the problems of using race as a biological category in genetics studies and suggested more appropriate approaches. One of the report’s authors is Ann Morning, a professor of sociology at New York University. Over a decade ago she wrote the book The Nature of Race: How Scientists Think and Teach about Human Difference. She spoke with Issues editor Monya Baker about why race is a poor—but persistent—shorthand in genetics studies.

SpotifyApple PodcastsStitcherGoogle PodcastsOvercast



Monya Baker: Welcome to The Ongoing Transformation, a podcast fromIssues in Science and Technology. Issues is a quarterly journal published by the National Academies of Sciences, Engineering, and Medicine and by Arizona State University.

Race is often thought of as a biological category to explain human differences, but today’s concept of races actually comes from European naturalists in the 1700s who created a ranking with white Europeans at the top.

Modern scientists have long agreed that race is a social construct, not a biological one. Genetic data do not show evidence for biological races, and relying on race as a discreet category can cause researchers to focus on genetic differences while ignoring other important factors to health.

Why do genetic researchers continue to use race and ethnicity as criteria in their studies? In March 2023, the National Academies released a report titled Using Population Descriptors in Genetics and Genomics Research. It explores appropriate ways to use race and other population descriptors and studies.

My name is Monya Baker, senior editor of Issues. I’m joined by Ann Morning, a sociologist at New York University and one of the committee members on the report.

Welcome, Ann.

Ann Morning: Thank you. Glad to be here.

Monya Baker: It’s well known that race is a social construct, so why is it still being used in genetic studies, and what problems does that cause?

Ann Morning: Well, first of all, I would say, Monya, that we can’t take for granted that it is so widely understood that race is a social construct. I would actually say that there is still pretty widespread thinking about race as a matter of biological difference, whether it’s skin color or something that runs deeper. Because in the US more broadly, race continues to be associated with biological difference, it’s not surprising that race is still appearing in scientific research many times as a measure or marker of biological difference.

Now, there are two main problems with this. One is the problem of race as a scientific tool. The problem is, simply put, that race is a really poor proxy for human biological variation. And it stands to reason, as you said, the race categories that we have today come to us from the 1700s. They’re categories that came out of really long-standing thinking about the four humors, or this old belief that the kinds of liquids that people had in their bodies—like blood or phlegm or what was called black violent yellow bile—determined people’s health and personalities and all kinds of other outcomes.

And that led us to our black, white, yellow, red framework that is basically the heart of our racial categories today. It’s not surprising that an eighteenth-century taxonomy that has just basically four categories—five categories to it, maybe—is actually not a very good way for mapping the subtleties and complexities that we know of today in human genetic variation.

Race is a very blunt tool, in other words, to try to get at human biological variation. But it’s also a problem, not just as a scientific tool, but as a cultural belief. Because when scientists use racial categories, for example, in genetic research, it sends a message to the wider society that races are actually biological groupings as opposed to social constructs or socially invented groupings.

And that also comes with the idea that belief in races as biological groups tends to come then with idea that racial differences are hardwired. That any differences in the life outcomes for different groups are saying something about their permanent qualities, their essential characteristics.

And that conveys the wrong idea that racial differentials also can’t be addressed. If we see outcomes in education levels or health outcomes or income or whatever it is between our so-called racial groups, that tends to be assumed that these are permanent kinds of differences that we can never change or narrow.

The idea of race as a matter of biology—as biological groups—is still very much with us. It’s with us in US society broadly, which means, of course, that it is also with us in the sciences to a certain extent. And it poses problem both in terms of how we conduct science and in terms of the way we think about inequalities and diversity in our own society.

Monya Baker: What you just said reminded me of something I was reading in the New Yorker, and it was talking about the concept of BMI, which came from this Belgian mathematician. So BMI is body mass index, which people use to assess whether or not your weight is a healthy range.

And when this mathematician was coming up with his concepts, he was basically looking for the ideal European man, and all of the data is based on that. And this article concluded that it’s not as accurate for people who are women or people who are Black or Asian or other groupings.

And it almost seemed like, oh, so each one of these should have their own number, but I think that that’s getting to the problem that you just talked about.

Ann Morning: What I take away from that excellent example is that if we limit our samples to one very restricted grouping of people—whether we’re doing that by race, we’re doing that by gender, we’re doing it by nationality, we’re doing it by zip code, whatever it might be—whenever we limit our samples in some way, of course that means that we have to ask ourselves the question of how well it represents the broader scope of humanity. So, I think that research that’s conducted solely, with a particular group and just one gender and so forth, is always going to raise that question of its representativeness.

However, that doesn’t then automatically mean that racial categories are a good way to describe the variation that is out there. And I think that is what is difficult for people to wrap their minds around. When we say race is socially constructed, what we’re saying is that these are categories of people that we made up, that as you said, were made up centuries ago.

We know, of course, now more than we ever have about human genetic variation. But we also know from that, that these patterns of variation are much subtler and much more complex than our four or five or three racial categories can get at.

Monya Baker: You’ve been thinking about this for a long time. In 2011, you wrote a book called The Nature of Race: How Scientists Think and Teach about Human Difference. Can you just tell me a little bit about what drove you to write the book? What problems do you see as being most persistent into the 2020s?

Ann Morning: I wrote the book because I was a graduate student, and in my doctoral studies in sociology, I was definitely hearing the message that race was a social construct—that is, that our racial categories are ones that we have historically made up, and we could have made them in different ways.

But I was struck that whether in the academy or outside the academy we’re still continuing to use race as if they captured patterns of human biological variation. For example, around that time, the FDA approved a drug for heart disease called BiDil that was marketed exclusively to African Americans. That’s something that certainly seemed to send a message that racial categories had a biological basis to them.

And of course—there are all kinds of other areas of US society—we can talk about commentary in the sports world about race and physical difference or just everyday beliefs that people have.

So basically, I got interested in this gap between what I was hearing in my doctoral studies about these new understandings of racial categories as historical and social inventions, and then just sort of the everyday ways in which people were talking about race.

At first, I thought that there was this gap because the idea of race as socially constructed must have been a new idea. I just thought, maybe it hasn’t really caught on yet—but then, I came across a really fascinating book written in 1942 by a British American anthropologist named Ashley Montagu. His book was called Man’s Most Dangerous Myth: The Fallacy of Race.

And he basically made the arguments about races being socially constructed and not biologically grounded that I was learning in my classes in the early 2000s. I thought, wow, if this idea of race is socially constructed has actually been around for quite a while, it’s been around since the early twentieth century—why hasn’t it made more headway? Why don’t more people seem to have absorbed that idea?

That’s really what got me on the path to exploring what people believe about race. It led to that first book. In a nutshell, what I concluded from my interviews with scientists, and also with college students, about what they thought about race, what I concluded was that we educators were not getting out to the public a clear message about race being a social construct.

And we weren’t getting that message across because in the academy itself, we hadn’t really all gotten to be on the same page about it.

Monya Baker: Tell me more about how this National Academy’s report came about. Sounds like it might be a continuation of these kinds of efforts.

Ann Morning: My understanding is that, like many other institutions in American society, there’s been a growing concern in recent years—certainly after the death of George Floyd in 2020 but even going back before then—there’s been this growing concern about the ways in which race plays a role in our society, the kinds of inequalities that it leads to, and injustices. And people in every kind of institution have asked themselves tough questions about how their particular organizations might be playing a role in these patterns of reproducing and maintaining these patterns of inequality.

The National Institutes of Health initiative, to bring this committee together, I think is really part of a broader concern. And I would say that I think there are two principle concerns at work here. One is the questions, the concerns about diversity, equality, and inclusion. In what ways can we make sure that this kind of research really is serving all people and not just certain communities?

But then there’s also, I think, the concern about how ongoing genomic research might be reinforcing certain ideas of human difference of race that are detrimental and actually make it harder to achieve the diversity, equity, and inclusion outcomes that we’re interested in.

Monya Baker: I was doing some background research before this interview, and I came across this 2012 article that was looking at how race was being reported in the medical literature. And it found in the 1980s only 14% of articles in the Journal of the American Medical Association reported on race. And by the 2000s, that was up to 46%. So 14% to 46% from 1980 to the mid-2000s.

And then I found a separate analysis of 2021 papers, and that found that 90% of the articles were reporting race. I just think a lot of clinicians and researchers have heard the message that they should be recruiting diverse populations, and somehow that’s getting conflated with this idea of race as a biological category.

How do you reconcile this instinct for equity and representation with accurate science?

Ann Morning: I agree absolutely that there is this risk of conflation. That on one hand there is this drive to, for example, put together more representative samples of people for research. But what our report tries to do is to cut through some of this confusion by basically saying two things.

So on one hand, we understand that race and other population descriptors might have a role in certain stages of research. For example, race might be a useful tool in recruiting individuals for studies. It might also be a tool that a researcher is required to use to describe the individuals in their sample. So for example, funders like the NIH will require that researchers use, for example, the United States Official Racial and Ethnic Categories, the ones that have been established by the Office of Management and Budget, to describe the individuals who are in their sample.

But our report suggests that while those uses of race or other population descriptors may make sense, that doesn’t necessarily mean that researchers have to use race as part of their analytical tools. Race might be the framework you use to describe your sample—but that doesn’t mean that in your analytical research, you are required to use racial categories, for example, to capture patterns of human biological variation.

And in fact, one of our recommendations is very straightforward on that count—where we counsel researchers to absolutely not use race as a proxy for human genetic variation for the very reasons that I spoke about before: it being an incredibly blunt tool for trying to get at the subtleties of human biological variation.

Monya Baker: I was reading the report about how if people are looking for Mendelian genes—just one gene trait—that they don’t recommend using racial categories at all. There’s some really cogent text about how there are different concepts of difference, that it could be political, that it could be genetic, and that it could be geographic. And how important it was to understand what concept of difference researchers were working from and drawing their conclusions from.

Can you tell me a little more?

Ann Morning: What we’re really advocating for here are two things. One is a careful tailoring of the population description that’s appropriate for the information that you want to capture and matching that with the study that you have in mind. And second, in a sense what we’re calling on researchers to do is to bring the same level of care and precision to their choice of population descriptors as they bring to so many other facets of their genomic research.

What we try to do then is give people tools for thinking carefully about what each of these population descriptors represents. It’s really important to clarify that this report, it’s not intended to be a laundry list of rules. Really the gist of this is to get people thinking very carefully about what each population descriptor actually represents conceptually, and then give people tools to think about which studies would benefit from one descriptor or another.

For example, there’s often—not just in the US but elsewhere in the world—there’s often confusion between race and ethnicity. People often use those words interchangeably. And so, one of the things that we do in this report is talk about how these really are different concepts and they capture different things. Race is really grounded in this historical way of thinking about the human body and physical differences, whereas ethnicity traditionally has been something that measures more cultural characteristics, traditions, religious practices, differences in food or dress, or what have you.

Both of those things are different from yet another population descriptor, that of geography, which might refer to the location of a person’s birthplace or their parental birthplace or the region in which their ancestors of certain generation were thought to inhabit. We really try to break these descriptors down in very clear ways so that we can help researchers think about, what is it that they really need to get at for this study that they’re conducting.

Monya Baker: Right. Do you care about the shape of somebody’s nose or what they tend to eat or how hot it is when they go outside?

Ann Morning: Exactly—that’s really well put—exactly. What is it you’re trying to get at? And in fact, one of the things we also say is that we think it’s important for people to get the best, most precise data that they can for any of these things.

If what they really, really want to know, for example, is—as you said—how hot it is, what the climate is in a place that people have grown up, then we want something other probably than an ethnicity marker or a race marker. And even, maybe we want something more precise than geographic location. Maybe we need actual temperature readings or climate readings from a particular place.

Again, what we’re really pushing for is great thought, care, and the most precision possible in measuring the kinds of data that researchers actually need for the research they’re conducting.

Monya Baker: One of the clearest recommendations is that people need to always consider environmental and other factors when they’re doing their studies. I feel like when people hear this word race as a category, they’re going to first assume genetic explanations.

Something else I was reading that is that Black women are three times more likely to have fibroids. But if you used another category like population density or income or region, that would pull up a lot of other biological mechanisms to explain this increased rate of fibroids than where if you say Black, I think the first thing that a lot of people think of is genetic.

What are some examples of research questions where the subject’s race is relevant, and where could it be uninformative or worse?

Ann Morning: As I said before, one place where we’re very clear about race not being useful is if it’s intended to be some subtle measure of patterns of genetic variation—it’s just way too blunt a tool to do that. But where race might be useful instead—I spoke already about how people might feel that they need to use race when putting their sample together or describing their sample to funders. But the other place where we think that race might actually be used in analysis might be as a variable for environmental conditions.

It might be that researchers feel that racial categories help them get aspects of the environment, which I understand from my geneticist colleagues on the committee—just the environment for them means anything that’s not genetic. Using that very loose grouping, we can imagine how race might be used as a proxy for environmental conditions like exposure to discrimination, access to health care. Ethnicity might be a variable that would be used to capture something like group norms about seeking health care and so forth.

There are ways in which researchers might feel that they can get enough mileage out of racial or ethnic categories to try to get at these environmental conditions. But again, our recommendation is really to try to be even more precise. So rather than, let’s say, use race to presume something about people’s exposure to environmental toxins or their disposable income, our recommendation is to actually go out and collect the data on those more precise measures and not just fall back on racial or ethnic groupings that are blunter and less precise. Again, because the risk here is that even when race is intended to just capture something about the environment, about structural inequalities, and so forth and so on, just using it in research that focuses on biomedical outcomes—that usage tends to reinforce this often automatic association that people draw between race and biology.

So it unfortunately runs the risk of reinforcing the belief that racial groups are simply kind of discreet, uniform, monolithic groups of biologically similar human beings.

Monya Baker: The National Academy’s report comes up with 13 recommendations, and we’ve talked about some of them. I’m wondering, what’s the low hanging fruit? What’s going to be the easiest of these recommendations to implement?

Ann Morning: A lot of our recommendations—really the bulk of our recommendations—are directed towards individual researchers, so to researchers using genetic or genomic data. What we’re asking researchers to do is basically to think very carefully about their use of population descriptors. To make sure that they’ve thought carefully about which descriptor or descriptors are most appropriate for the work that they want to do. Which group labels or categories are actually appropriate? And also very important for us is to encourage researchers to be very transparent about how they went through this decision process, and why they choose the descriptors that they do.

You might say that on some level, that’s not so hard to do. And again, it’s basically a matter of bringing the same care and precision and attention to this aspect of research design that geneticists are routinely bringing to other very complex aspects of their research.

I think it’s also important, though, to underscore that as I said previously, the report is not intended to be a straightforward list of dos and don’ts. It’s not a simple list of “don’t use this word but use that label instead,” so it’s not that easy. It is really asking people to think carefully.

There are also other aspects of the report which are probably going to be harder to implement. We’re speaking in some of our recommendations to funders, to journal editors, to journalists, to educators, and so forth, and really recommending that they invest in some way in the education or training of scientists or providing guidelines for decisions around the choice of population descriptors.

We are encouraging them to consider interdisciplinary approaches or working groups or even advisory groups that take up these issues. That definitely is asking for another level of institutional work.

Some researchers might be concerned that all of these are recommendations and the ways they’re implemented by other institutions—that this might all mean an increase in bureaucracy for researchers. Will they have to fill out more checklists when they’re applying for funding? And so forth and so on.

But I think the question to ask is really, what is the alternative? That is, if we don’t really prod researchers and ask researchers to think very carefully about how they’re using population descriptors, then are we okay with a situation where researchers continue to use these population descriptors—like race and ethnicity and others—in unconsidered ways and in ways that are unclear or confusing or even misleading?

Monya Baker: It’s so easy to just look at what people measured in the past and just use the same thing, because it means you can get started with your research earlier. And I think what you’re saying is think before you choose those variables.

Ann Morning: I agree very much, and I’ll say that similarly in the social sciences. A basic introductory tenet of what we teach students in methods classes is that they should think carefully about the concept that their variable is meant to capture. And then that they also think carefully about how they operationalize it.

We all know that genetics research grapples with incredible challenges of complexity. We’re not talking about a field which is wedded to simplistic thinking or easy answers. Far from it. But because race, certainly in American society—which means all of us, whether we are laypeople or we are bench scientists—all of us have been inculcated with this taken-for-granted way of thinking about race.

And we just don’t often stop to ask ourselves what we mean by race or whether it’s different from ethnicity, and what do we mean by this or that? That’s the challenge of really asking people to rethink things that they think they know or that they’ve taken for granted, but just take a step back and reexamine them.

Monya Baker: Yeah. And in terms of reexamining what terms to use, the report doesn’t make a lot of explicit recommendations. It talks more about how to make those choices. But one explicit recommendation is to avoid using the term Caucasian. Tell me more about that.

Ann Morning: Caucasian in particular is still a term that gets used when people think that it’s actually getting at a scientific and objective measurement of a human group. It still has that connotation. That’s one of the reasons that we actually target that term in our report.

This is one of the places where we actually do wade into questions of semantics. Like I said, that’s not mainly what we’re about in this report, but with this term, we do. Not only because Caucasian gets used by some as a supposedly objective and scientific biological term, but also because of its particular backstory. That Caucasian is a word that comes to us from [Johann] Blumenbach, considered one of the founding fathers of anthropology, who in the 1700s coined that term because he felt that skulls from the Caucasus—when he was serving his large collection of human skulls that were the basis of his thinking about human races—he felt that a skull from the Caucasus was the most beautiful skull in his collection. And so that made it the perfect symbol of the white race because at that time, beauty was considered to be an objective, surefire indicator of a person’s racial membership. He thought that this beautiful Caucasian skull was the perfect symbol for whiteness.

So just the term of Caucasian, aside from erroneously giving the idea that there are some objective sciences to racial categories, it also comes loaded with that historical hierarchy, that historical white supremacism that’s built into it.

We also cautioning against using phrases like “the Black race,” which conveys the impression that there are these objective groupings that just exist out there in the world, ones that we invented, and frankly that we also reinvent with time.

Right now, the US Office of Management and Budget is going through public hearings, public exercises to consider changing the framework of racial and ethnic categories that it uses and that everyday people we encounter on the census most obviously. These categories are ones that we fiddle with, that we change, and we can do that precisely because they’re not objective categories, but they are these social creations that we invent and we tinker with and we discard or reinforce.

Monya Baker: My last question is, what final thoughts do you have?

Ann Morning: So just to underscore that what we really try to do in this report is get people thinking hard about what the concepts are of human difference, and particularly descent-associated human difference, right? Because there are many kinds of ways in which human beings discuss “different”—we’re not talking about all of them in this report.

We’re talking about the kinds of categories and taxonomies that get used to described groups that we think of as being different in some way because of their ancestry or descent.

We really want people to understand that we’re really trying to get them to operate and reflect on that conceptual level and not think that this is simply a matter of word choices or group labels. That there’s something deeper, more far-reaching and more important, but also ultimately more fruitful by taking that more foundational route.

Monya Baker: Yeah, not what term do you pick, but how do you pick what term do you pick?

Ann Morning: Yes, exactly. And what are the characteristics or the differences that you think you’re capturing when you choose this label or not?

Monya Baker: To find the National Academy’s recommendations and learn more about this topic, check out the report Using Population Descriptors in Genetics and Genomics Research. You can also check out Ann Morning’s latest book, An Ugly Word: Rethinking Race in Italy and the United States.

Find links to those resources and more in our show notes. Email us at [email protected] with any comments or suggestions, and I encourage you to subscribe to The Ongoing Transformation wherever you get your podcasts.

Thanks to our podcast producer, Kimberly Quach, and audio engineer Shannon Lynch. My name is Monya Baker, senior editor of Issues in Science and Technology. Thank you for joining us.