Race and IQ — Sam Harris and Charles Murray Set the Record Straight on Intelligence Testing
What this covers
This is an excerpt from episode 73 of Waking Up with Sam Harris, published on the 22nd of April 2017, titled Forbidden Knowledge, in which Sam Harris talks to Charles Murray about racial, cultural, and gender specific IQ differences..
You can listen to the entire episode here: https://soundcloud.com/samharrisorg/73-forbidden-knowledge
Source description (no synthesized summary yet).
Charles Murray argues that while IQ differences exist between racial groups with both genetic and environmental contributions, the ethical and practical imperative is to treat individuals on their merits rather than group membership, and current affirmative action policies often harm the very students they intend to help by creating mismatches between student ability and institutional rigor.
- IQ is a valid, non-biased predictor of outcomes across racial groups; claims of cultural bias in testing are scientifically unsupported
- The black-white IQ gap likely has both genetic and environmental components, not exclusively environmental causes as critics claim
- Treating people as group members rather than individuals is both irrational (given substantial within-group variance) and creates perverse incentives that harm minority students
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Regardless of the origin of IQ differences between groups, the ethical and rational approach is to treat individuals as individuals and ignore group membership, because within-group variation far exceeds between-group variation and overlap between groups is substantial.
“the truth is I learn nothing about a person's intelligence simply by being told that he's black or white or Asian you still need to treat people as as individuals and and you make it absolutely clear in your book that given the overlap in in the these bell curves there will be many many blacks who are far more intelligent than most whites I mean so this is again it all comes back down to honestly evaluating individuals”
IQ tests have predictive validity: IQ scores predict income, job productivity, and employment outcomes better than any other single datum an employer could obtain, including personal interviews, grades, or degrees.
“the IQ scores predict this outcome so they predict income in terms of employment decisions from job productivity you are better off if you're an employer and you have only one datum that you can get you can't you can't have two you are better off knowing an IQ score than you are having a personal interview having grades having degrees or anything else the the single most informative thing you can have is an IQ score this is not the result of a one or two studies the the predictive validity of IQ tests has been established over and over”
IQ tests do not systematically underpredict Black performance: extensive research shows no evidence that SAT or IQ test scores underpredict the college or job performance of Black students, refuting the claim that tests are culturally biased against Black people.
“there is absolutely no evidence whatsoever that test scores whether IQ test or sat underpredict Black Performance there is some evidence that they may overpredict Black Performance but that that raises a whole different set of issues and problematic issues but the tests are not biased against blacks in terms of predic validity the way that most people think about cultural bias though is in items that a white upper middle class kid is more likely to know the answer to than than a black inner city kid”
The within-group variation in intelligence is much greater than between-group variation, meaning there is substantial overlap in the Bell curves of different racial groups, so an employer cannot reliably infer job performance from race alone.
“there's much more variation uh within groups uh than there is between I mean the separation is much much greater within groups than it is between groups so that the overlap is very large”
Affirmative action should not place people into institutions where they will face predictable failure; the central scandal is the high dropout rates of minority students at elite schools like MIT, where admitted students face classmates of vastly superior ability.
“but at the same time that you did that and and you needed to do this on on the basis of of what's good for the people you're trying to help it needed to be one which did not put people into places where they were set up to fail the one of the great scandals that nobody talks about in Elite schools is the Dropout rates of their minority students”
Modern employment law creates perverse incentives that raise the cost of hiring minority candidates because employers fear discrimination lawsuits and regulatory penalties if they don't treat all employees identically and can't prove ironclad justification.
“what social policy is doing in an employer's mind when it is a black candidate walking into that office is all sorts of things that are raising the cost of hiring that person to that employer raising the cost in terms of vulnerability to uh discrimination lawsuits vulnerability to a variety of other regulatory penalties because all at once that person that person cannot be evaluated on that person's merits and the decision made solely on on those merits without incurring risks”
Most people do not know about the intelligence, integrity, and personality of others by looking at them or knowing their group membership; all traits vary substantially within individuals and across situations.
“you don't know whether they are people who we would like to hang out with we don't know how smart they are we don't know how uh much Integrity they have by looking at them and assigning them a group membership whether by the way it's not just race but it's also sex and it's sexual orientation and its ethnicity of a much more detailed form any group you want to name you don't know on the basis of group means what you're dealing with”
There is now an inversion of conventional versus elite wisdom: ordinary people accept that some people are smarter than others and that intelligence confers job advantages, while elites claim IQ tests only measure test-taking ability, are culturally biased, and cannot validly measure intelligence—a position equivalent to claiming the Earth is flat.
“at this point Sam it's almost as if we are in the opposite position of conventional wisdom versus Elite wisdom that we were say when Columbus was going to sail to America when when Columbus was going to sail to America it is true that an awful lot of the ordinary people still thought that the Earth was flat but among the elites it was understood that the Earth is round well now it is ordinary people are perfectly comfortable with the idea that some people are smarter than others they're perfectly comfortable that that that what we call Smart uh gets you kinds of jobs that you can't get otherwise all that kind of stuff it's the elites who are under the impression that oh IQ tests only measure what IQ tests measure and nobody really is able to Define intelligence and this and that they're they're culturally biased on and on and on and on and all of these things are the equivalent of saying the Earth is flat these are not opinions that you can hold in in contest with the scientific literature anymore than you can be an Aristotelian physicist uh in contradistinction to a Newtonian physicist this stuff is not subject to debate anymore”
If racial groups differ genetically to some degree, and most traits humans care about (including intelligence) have some genetic underpinnings (approximately 50% heritable), then it would be statistically surprising if population-level averages were identical across all racial groups for any variable.
“and because different racial groups differ genetically to any degree and because most of what we care about in ourselves the intelligence included because most of what we care about in ourselves also it has some genetic underpinnings for many of these traits we're talking about something like 50% it would be very very surprising if everything we cared about was tuned to the exact same population average in every racial group there's just virtually no way that's going to be true so based purely on biological consideration we should expect that for any variable there will be differences in the average its average level across racial groups that that differ genetically to some degree”
The blurriness of racial categories (fuzzy boundaries between groups) acts as noise in detecting genetic IQ differences, not as artificial signal; if genetic differences exist, they should survive this noisy categorization, whereas environmental-only explanations might be confounded by racial classification errors.
“the blurriness of race is noise in the signal it the blurriness of race is not going to to artificially make it look as if there are genetic differences in IQ it's going to obscure any such uh genetic differences in IQ so so in effect we are looking at a noisy signal and if you still see patterns in the data that point to the possibility of genetic roles those have survived a lot of contamination”
The Flynn effect (documented rise in IQ test scores over decades, particularly 1940s-2017) cannot be explained primarily by increasing genetic intelligence, because it occurs too rapidly and is independent of the nature of the Black-white IQ gap.
“the Flyn effect is that it was noted also by a guy named uh Richard Lynn who is uh one of the the people that uh I'm excoriated with for citing in the bell curve as being an out andout races but Richard lyard also identified this in East Asian tests and Jim Flynn identified it and brought it he did more than Lynn did to bring it to public attention namely IQ tests are rored every time they have a new addition so if you have the Wexler and let's say they had a a uh norming in 1950 they Norm it to a mean of 100 and a standard deviation of 15 what Flynn noticed was that if you give that same test to a set of people in 1960 instead of 1950 the mean is not 100 it's 102 or maybe 103”
Modern genetic sequencing has allowed researchers to analyze variation across millions of SNPs (single nucleotide polymorphisms) and perform sophisticated cluster analysis on populations, enabling far more rigorous study of genetic ancestry and variation than was possible in the 1990s when the Bell Curve was written.
“the thing that has changed most dramatically is now the genome has been sequenced and uh so much has been learned since it has been sequenced the whole discussion of uh ethnicity slra is being conducted at a much higher level of sophistication at the time we were working on the bell curve you know they would look at blood groups and things like this to try to have something besides phenotypic characteristics facial characteristics and skin color now the well the ability of the geneticist to Simply uh look at variation over a million Snips across populations and do really fascinating uh cluster analysis just the the whole conversation about uh the word populations is what the geneticists like to use now instead of race”
The mismatch hypothesis (that high-performing students admitted to elite schools where they're outmatched by peers experience worse outcomes) is a legitimately contested empirical question with credible research on both sides.
“there is a literature that has gone both ways on this issue and and for the first time this afternoon I will say I am on one side of a contested literature where there are data that can be cited by both sides there are there have been books written that said no this mismatch does not produce the U all the bad results that I've just claimed or even if it does produce some of those there are other advantages to it so I want to acknowledge the existence of that alternative argument”
The reaction to the Bell Curve was a moral panic characterized by irrational claims that intelligence doesn't exist, that differences cannot be measured, and that racial differences cannot have any genetic component—representing the rejection of a century of scientific understanding.
“it was a kind of moral panic I me you when you look at just how irrational the claims were I mean you're talking about claims that intelligence doesn't exist on some level or that it doesn't matter or that you can't possibly test for it or that there can't possibly be differences among mean differences among groups or that it could have no genetic underpinnings I mean this is like they were throwing out A Century of our growing understanding of these things and vilifying you with an energy that clearly this is what happens when you touch a taboo in this way”
The premise that group outcome differences stem entirely from racism/sexism/institutional bias rather than acknowledging real group differences has created significant harm in institutions.
“the equality premise that all groups of people only have differences in outcomes because of racism or sexism or inappropriate institutions that assumption has created huge harm”
Item-level analysis of IQ and SAT tests shows that culturally loaded items (those more familiar to white upper-middle-class students) are not systematically harder for Black students; the ordering of item difficulty is the same across racial groups, and minority students sometimes perform better on culturally loaded items than on purely abstract ones.
“you can do item analysis of the tests so that you can for example have people simply inspect the items and rate them according to their cultural loading that would be one approach and then you can extract those items and see whether test scores converge or you can see whether uh the items that are culturally loaded are harder for for black students to answer than they are for white students and the answer to th those questions is no that the ordering of difficulty of items is the same across races and that when you have tests that are empty of any cultural loading at all that it is not that the the black white difference diminishes uh it doesn't sometimes it gets greater because in this is a more complicated statement there's more complicated literature but there is some evidence that the culturally loaded items are ones that minority groups do better on than the purely abstract ones”
Dutch scholar Jelty Wicherts and colleagues found that the Flynn effect and the Black-white IQ gap operate through different mechanisms and have different patterns across subtests, suggesting they are independent phenomena rather than one explaining the other.
“the nature of the Flyn effect is pretty much divorced from the nature of the black white difference but the Flyn effect itself is a fascinating phenomenon and it's a reminder that that we don't know everything there is to know about all this stuff right”
If genes and environment contribute equally (50-50) to IQ variance in a population, then for the environment to explain 100% of the 1-standard-deviation Black-white IQ gap, the average Black person would need to experience environmental conditions approximately 1.5 standard deviations worse than the average white person—a much larger environmental gap than observed in measures of income, education, and occupational distribution.
“if you say for purposes of just thinking through the arithmetic that that genes uh and environment it's a 5050 split in explaining uh variance of an niq in a whole population that means that in order for the environment to explain 100% of a standard deviation difference mean between blacks and whites the average black would have to be at an environment that is about 1.5 standard deviations below the white mean that's a really big difference and if you take all of the measures like income and educational attainment and occupational distribution and a variety of other measures of of uh environment uh one and a half standard deviations is way way bigger than any of the observed differences are there that doesn't mean that there are un aren't unmeasured differences in the environment that are also at work it just is you start off with a really big question in your mind is that plausible that it can be 100% environmental”
Possible causes of the Flynn effect include increased exposure to visual three-dimensional rotation tasks (television, video), improved nutrition, and increased general cognitive sophistication from modern life, but improved nutrition alone seems unlikely to account for century-long gains.
“is it simply a matter of increasing of exposure to things which let you answer IQ questions an example one of the kinds of IQ questions is uh rotating mentally in your mind's eye an object in three dimensions and being able to say something about how it look when you look at it from the other side those are items which which I think I did very poorly on by the way well you know the the ability of people to answer that question is going to be different in 1930 when nobody sees routine ly objects rotating in three dimensions in front of your eyes than in 2017 where every television commercial is having you know script and other things rotating in three dimensions as part of our daily lives so it could be that kind of of Simply modernity and sensitization to certain kinds of uh mental tasks that we weren't sensitized to before could it also be an increase in G might be uh there there are some reasons to believe in the analysis of test scores that it could be partly that but not a lot that I think it's probably a fair characterization of the state of knowledge right now is improved Nutrition a thesis there or not improved nutrition could well be a contributor improved nutrition does enhance IQ although you would think you think you would hit the ceiling long before the better part of a century because nutrition is probably where it needs to be now”
The ability to mentally rotate three-dimensional objects is an IQ test item; people in 1930 rarely saw objects rotating in three dimensions, but by 2017 television commercials routinely feature rotating 3D graphics, potentially increasing the population's ability to perform well on this item type.
“one of the kinds of IQ questions is uh rotating mentally in your mind's eye an object in three dimensions and being able to say something about how it look when you look at it from the other side those are items which which I think I did very poorly on by the way well you know the the ability of people to answer that question is going to be different in 1930 when nobody sees routine ly objects rotating in three dimensions in front of your eyes than in 2017 where every television commercial is having you know script and other things rotating in three dimensions as part of our daily lives”
The Black student dropout rate at MIT was approximately 24% at some point in the past (exact timeframe uncertain), and students admitted are in the top 1-5% mathematically but face classmates in the top 0.01-1% of mathematical ability, making them feel 'the dumbest kid in the class.'
“the dropout rate among black students at MIT was about 24% now to get MIT you are going to be in the top one or two 3% of uh of mathematical ability maybe let's say everybody's within the top 5% you're going to be very smart in math but if you are let in and let's say you're at the fourth percentile in math ability and you are in engineering classes with your fellow students at MIT the rest of the students in that class are in the top on10th of 1% sometimes the top one 100th of the top percent and in that kind of setting you are the dumbest kid in the class”
The Bell Curve authors' position was that both genes and environment likely contribute to racial IQ differences, and critics are obligated to defend the much stronger proposition that the Black-white IQ difference is 100% environmental if they reject the book's conclusions.
“it seems to us highly likely that both genes and the environment have something to do with racial differences and go we went no farther than that there is an asymmetry between saying probably genes have some involvement and the assertion that it's entirely environmental and that's what the that's the assertion that is being made if you're going to be upset at the bell curve you are obligated to defend the proposition that the black white difference in IQ scores is 100% environmental and that's a very tough measure here's the way that we went about it”
Race is a valid biological concept, though blurry rather than sharp: racial groups differ genetically to some degree; these differences have phenotypic expressions (facial features, skin color) and genetic underpinnings in disease susceptibility, and genetic variation between racial groups is non-zero.
“it's a biological concept but is is a blurry one I mean is similar to the way I races races can be thought of as analogous to families in fact some people have said that that a race is essentially just a very large family that is partly inbred but you can see family resemblance in the races I mean this is it's not an accident that you can generally predict where a person's ancestors came from by just simply looking at his face right I there there are phenotypic differences between people that have genetic underpinnings and it's not merely just Skin Deep I mean there are genetic diseases that various racial groups have are more prone to you know tasac Cel anemia and this is just straight biology”
The interlocutor expresses conflict about having the race and IQ conversation because researchers investigating racial differences could be motivated by racist curiosity or could unintentionally provide ammunition for white supremacist organizations that cite the Bell Curve.
“I'm still conflicted over this issue of why study this topic at all because it's very easy to see why anyone would assume that if you're looking in an area that is producing invidious comparisons between races and you're continuing to look in that area and continuing to be interested in that area your interest must be motivated to some degree by a kind of morbid and quasi racist curiosity in just sort of kind of Shoring up a notion of difference between white and black in this case and needless to say I'm sure we can find you know white supremacist organizations who absolutely love the fact that the bell curve was published and just admonish their members to read it at the at the first opportunity”
James Flynn has argued that an environmental explanation for the Black-white IQ gap need only posit that the average environment for Blacks in 1995 matches the quality of the average environment for whites in 1945—a position that Murray characterizes as raising serious questions but not definitively resolved by current evidence.
“an environmental explanation of the racial IQ Gap need only posit this that the average environment for blacks in 1995 matches the quality of the average environment for whites in 1945 I do not find that implausible”
Education has been feminized in recent decades, with teaching methods and classroom environments that favor how girls learn over how boys learn, resulting in boys struggling in K-12 education.
“education in recent years has been taken over by essentially an elite wisdom which has feminized education and changed K12 education in ways which Boys Don't thrive in and girls do and the answer to that is not to go back into an old form of Education which was based B on how boys learn rather it's important to recognize differences between men and women boys and girls that exist to do a good job of educating them”
By the early 1980s, Murray observed that affirmative action policies were creating an atmosphere that made it harder for minority students to succeed, not easier, based on volunteering with gifted inner-city students in a tutoring program.
“I had been tutoring on a volunteer basis in the early 80s uh kids in the inner city black kids in the inner city in the uh what's called The Higher Achievement program program is run by the Catholic church I'm not Catholic but the program wasn't limited to Catholics it took the smartest kids from the inter City Public Schools and had them do intensive tutoring a couple of times a week and so I tutored those classes and this was not uh touchy FY tutoring this is really high-paced stuff and these were great kids and they were they were coming in they were attending they were working hard everything was going right and one of the things that just stuck out was the degree to which all of these attempts at affirmative action in its new form we're creating an atmosphere in which it made it harder for these kids to succeed not easier”
Murray bases his concerns about modern affirmative action policy not only on statistics and academic literature but on personal experiences where he observes that current policies make it harder, not easier, for the young people they intend to help.
“all I can say is I am taking it on the basis not just of Statistics in books but personal experiences I have had where I think we are doing enormous harm to Young People by making it harder to treat them as individuals”
Some academics who criticized the Bell Curve deliberately misrepresented its contents, knowing their criticisms were false because they had enough expertise in the subject to recognize they were lying, suggesting they acted from ideological conviction rather than honest disagreement.
“I know of academics especially who actually read the book had to know they were lying because I'm thinking of specific academics they just simp know too much about this subject not to have known that they were lying and they lied without any apparent shadow of uh guilt because I guess in their own minds they were doing the Lord's work”
Murray's 1984 article for The New Republic argued that affirmative action creates a stigma problem for beneficiaries who are assumed to have been hired due to race rather than merit, which would be psychologically damaging.
“if you go back to some of my earliest published stuff of on on affirmative action you can go back to 1984 when I did an article for the new Republic in which I was talking about the mismatch problem and a lot of that is how would I feel if I were a black kid my age and going into college and everybody thought I was there because I was an affirmative action kid I would hate that I would really hate it how would I feel if on the job I knew that everybody assumed I got that job because of affirmative action I would hate that and I would try to do my best to to prove them wrong but I find that morally repugnant”
The left's historical enthusiasm for IQ testing was based on the belief that it promised a direct path to meritocracy by identifying talent independent of class background, which is why the SAT was invented as a tool to allow talented students from mediocre public schools to gain admission to elite colleges like Harvard.
“the left used to be kind of booster is about IQ testing because it seemed to promise a direct road to meritocracy it would get us out of these class differences and people could just be judged on their own merits that's why the SAT was invented the sat was uh was going to be and in fact it did serve this function it would be a way for kids who did not go to groten and uh exitor and the rest of it to to uh get a chance to show how smart they were and they could be brought into the colleges and Harvard in particular and its Conant its president back in the 1940s were very hot on using tests for precisely that purpose”
There was genuine historical need for affirmative action to remedy exclusion of women and racial minorities from employment and education.
“had people been excluded Ed uh whether women or blacks or other minorities absolutely was there a need for affirmative action to remedy to that yes there was”
Richard Lynn and James Flynn independently identified what became called the Flynn effect, with Flynn receiving more public attention for bringing it to attention.
“it was noted also by a guy named uh Richard Lynn who is uh one of the the people that uh I'm excoriated with for citing in the bell curve as being an out andout races but Richard lyard also identified this in East Asian tests and Jim Flynn identified it and brought it he did more than Lynn did to bring it to public attention”
IQ tests are renormed each time a new edition is created, with the mean reset to 100 and standard deviation to 15 at that time.
“IQ tests are rored every time they have a new addition so if you have the Wexler and let's say they had a a uh norming in 1950 they Norm it to a mean of 100 and a standard deviation of 15”
The Bell Curve used self-identified racial categories (Black, Latino, white) to define populations rather than genetic clustering analysis, which was not yet available at the time.
“I don't blame them and I I'm happy talking about populations too that's that's just being done at a huge level that we never considered in the bell curve we simply said if people call themselves black or Latino or white we're going to believe them and we're they're going to be our samples”
Original affirmative action (in its first ~12 months of use) meant employers should make greater efforts to reach applicant pools previously excluded and actively recruit qualified minorities, not preferences or lower standards.
“the original sense of affirmative action for about the first 12 months after the term came into use was that if you were an employer you should make greater efforts to reach out to uh applicant pools that you wouldn't have otherwise reached out to and that you would take affirmative action to bring people in that that had been excluded”
In 1961 when Murray arrived at Harvard, there were fewer Black undergraduates than today, but his instinctive reaction upon seeing Black students was to assume they were probably smarter than him, based on reasoning that they'd overcome greater obstacles to get there.
“I went back to when I got to Harvard in the fall of 1961 there were way fewer black uh students undergraduates than than there are now way fewer but I will tell you this was pre-a affirmative action pre- civil rights action 1964 for that matter the kid from Newton Iowa every time he saw a black face uh at the Student Union or whatever my instinctive reaction was he's probably smarter than I am and I made that assumption because I figured that the the black kids are very likely to have had a tougher root to Ho than I had to get there”
Bell Curve authors spent considerable effort on language and presentation in the controversial chapter on race and IQ, hoping to convey that this was not a topic worthy of panic and to defuse exaggerated public fears rather than inflame them.
“when we wrote the chapter we spent a lot of time in this chapter with with trying to to get it right not just in the technical details but also in the language and when we were all done with it I actually had these hopes that when the book came out that dick and I would be applauded for having taken this inflammatory issue and treated it saying this is not such a big deal uh this is nothing to get excited about but something that it's better to to look at than for people to have these way exaggerated Notions of what might be going on”
Sociologist Steven Rosenthal called the Bell Curve 'a vehicle of Nazi propaganda wrapped in a cover of pseudoscientific respectability, an academic version of Adolf Hitler's mind.'
“the sociologist Steven Rosenthal called The Book quote a vehicle of Nazi propaganda wrapped in a cover of pseudoscientific respectability an academic version of Adolf Hitler's mind comp right this was the tenor of the response even among intellectuals and academics who were reviewing the book”