On Proposed Environmental Causes of the American Black-White IQ Gap

In this post, I will consider specific environmental variables which have been suggested as causes of the roughly 15 point IQ gap between black and white Americans. I will conclude that there are currently no known ultimately environmental variables which totally or even partially explain the IQ gap and that proximate environment causes which may contribute something to the gap account for far less of it than is commonly thought.

In order to make my conclusion clearer, we need to distinguish between proximately environmental and ultimately environmental causes. An environmental cause of the IQ gap is not ultimately environmental if it is a proximate cause which itself is caused by genetic differences between groups. For instance, if the IQ gap were partly due to differences in the rate at which whites and blacks go to college but this difference in college attendance was itself the result of genes then I would not count this explanation as being ultimately environmental. Instead it would be counted only as a proximately environmental cause. My conclusion thus means that we don’t know of any environmental causes of the black-white IQ gap which are themselves known to not be caused by genes and the variables we know of which may or may not be caused by genetics hardly account for anything.

The main methodology employed throughout this article is straight forward. To estimate the contribution of an environmental difference to the black-white IQ gap we need to know the magnitude of the group difference and the size of the effect that the environmental variable has on IQ. This effect size cannot be derived from observational research because of both environmental and genetic confounding. Experimental or genetically sensitive designs therefore must be utilized.

Of course, estimating an effect size will be unnecessary if it turns out that there isn’t a significant group difference in the relevant environmental variable to begin with. In this case, the relevant variable can be concluded to make no important contribution to the IQ gap no matter its impact on individual level IQ scores.

I should also note that I am looking for explanations of the racial IQ gap in adulthood. Because of this, if an environmental variable has an effect in childhood that does not persist into adulthood it will be treated the same as a variable that has no effect on IQ at all. Moreover, because the effects of environmental interventions on IQ generally “fade out” over time, we are justified in not thinking a variable that impacts childhood IQ also impacts adulthood IQ unless specific evidence is presented to show otherwise.

Protzko (2015)

Income

Let’s begin with income. It is well known that racial IQ gaps largely persist after controlling for variables like socio-economic status and that high income black kids have lower cognitive ability scores than do low income white kids.

(Jensen, 1998)

However, the well established genetic correlation between SES and IQ implies that these analyses will significantly over-estimate the role of the environment.

Citation	Age	Correlation	% Genetic
Krapohl and Plomin (2016)	16	0.5	50%
Trzaskowski et al. (2014)	7	0.31	94%
Trzaskowski et al. (2014)	12	0.32	56%
Rowe et al. (1998)	Adults	0.34	59%

Adoption studies and guaranteed income experiments indicate that variance in family income doesn’t impact cognitive ability in a way that lasts into adulthood. Given that family income does not impact adulthood IQ, it obviously cannot contribute anything to racial IQ disparities.

Citation	Description
Blau (1999)	Differences in family income didn’t predict sibling differences in most cognitive abilities. There was one exception: a $10,000 increase in income did predict a 0.22 SD increase in reading ability.
Maynard and Murnane (1979)	A guaranteed income experiment on poor African Americans increased reading scores by .23 SD and had no effect on GPDA for grades 4-6. It had no effect on reading scores and a negative effect on GPA (.18SD) for grades 7 – 10.
Maynard (1977)	A guaranteed income experiment on children in North Carolina and Iowa produced no effect on GPA in Iowa and a 6.2% increase in GPA in North Carolina for young children. No effect was found in either state for high schoolers.
Duncan et al. (2011)	An analysis of 16 welfare experiments found that increased income improved teacher-ratings of student performance, but had no effect on test scores.
Scarr et al. (1976)	Family income failed to predict IQ in a sample of 130 black children in adoptive homes.

Some may express skepticism towards this conclusion on the basis of the fact that some of this research is on GPA and academic test scores. Such concerns should be mitigated by the fact that GPA correlates quite highly with IQ.

Citation	Age	K	N	R
Roth et al. (2015)	Elementary School	71	18584	0.45
Roth et al. (2015)	Middle School	75	49771	0.54
Roth et al. (2015)	High School	71	15427	0.58
Postlethwaite (2011)	High School	32	13290	0.65
Postlethwaite (2011)	College	78	16449	0.72

More importantly, academic tests virtually are IQ tests.

Citation	Test	Correlation	N
Brodnick and Ree (1995)	SAT – V	0.8	339
Brodnick and Ree (1995)	SAT – M	0.7	339
Brodnick and Ree (1995)	ACT	0.87	339
Frey and Detterman (2004)	SAT	0.86	917
Frey and Detterman (2004)	SAT	0.72	104
Beaujean et al. (2006)	SAT	0.58	97
Deary et al. (2007)	GCSE	0.81	70000

Thus, while this data is imperfect it is still quite significant evidence and in the absence of better evidence to the contrary, we are rationally obligated to think that parental income does not impact IQ. Given this, parental income must explain none of the racial IQ gap.

Quantity of Education

Let’s next consider education. We might think that parental education explains some of the gap.  Parental education does predict IQ in adoptive samples and this justifies thinking that the effect is real and not the result of genetic confounding.

Citation	Description
McGue et al. (2007)	A family SES measure that included parental occupational status and parental education failed to predict adolescent IQ in adoptive families. (r = -.02, N = 409)
Beaver et al. (2014)	Among young children, maternal education predicted IQ in an American adoptive sample (beta=.21, 286), but this relationship was non-significant in adolescence (beta = .10, N = 229).
Weinberg et al. (1992)	Adoptive parental education predicted IQ among young children (r = .20 for maternal and .38 for paternal) once the same kids reached adolescents the correlations were non-significant (r = .11 / .14, N = 101).
Kendler et al. (2015)	In a sample of adopted Swedes, parental education correlated with offspring IQ at .18 at age 18 – 20 (N = 2,821).
Leahy (1935)	Parental education correlated at .20 in a sample of 193 children living in adoptive families.
Burks (1928)	In a sample of adoptive children, paternal education correlated at .01 with IQ (N = 173) and maternal education correlated at .17 (N = 194)

However, the existing evidence can’t tell us whether this effect lasts significantly after children separate from their parents in adulthood and even in childhood the effect size is too weak to significant contribute to the black-white IQ gap.

We might also think that personal (rather than parental) education explains part of the IQ gap. Like parental education, personal education also seems to impact IQ. Different study designs indicate different answers about whether this effect fades with the time, but even if it does the rate of decay is not great enough to wipe out the effect entirely even in late adulthood.

(Ritchie et al., 2018)

Thus, unlike income, both personal and parental education seem to have a causal impact on IQ. Given this, we might conclude that whites having 1-2 years more education that blacks accounts for something like 2-4 points of the black-white IQ gap.

(Ryan et al. 2016)

However, the full standard deviation IQ gap between blacks and whites is already present among high school students and college applicants. Obviously, in these samples neither group has yet reached their mean level of educational attainment and everyone’s degree of educational attainment is roughly equal. This line of data suggests that personal educational attainment doesn’t account for any significant portion of the black-white IQ gap.

Roth et al. (2001)

One possible means of reconciling these seemingly contradictory lines of evidence is to posit that IQ tests measure more than one sort of cognitive ability and that the cognitive abilities that education impacts are not the ones that races differ in. Saying that IQ tests measure more than one ability is self evidently true. One way of dividing these abilities is to say that some are general, in the sense that they underlie performance on many seemingly different mental tasks, and some are specific, essentially just meaning that they are not general.

General intelligence can be measured separately from IQ test scores by applying factor analysis to the inter correlations between IQ subtests. We can then ask which subtests are most correlated with general intelligence and whether those happen to be the subtests on which racial groups differ the most strongly. Meta-analysis have shown that this is the case, suggesting that racial groups differ primarily in general intelligence rather than specific abilities.

Citation	Group 1	Group 2	K	Correlation w/g loading
Nijenhuis et al. (2015)	Whites	Amerindians	25	0.62
Nijenhuis et al. (2014)	Whites	Jews	4	0.8
Nijenhuis et al. (2014)	European Jews	Oriental Jews	4	0.87
Nijenhuis et al. (2016)	Whites	Blacks	15	0.57

Returning to education, research has found that schooling improves IQ scores via the improvement of specific abilities and not through the improvement of general intelligence (Ritchie et al., 2015). Thus there is reason to think that education effects IQ generally speaking but does not effect the cognitive abilities that races differ in and so plays no role in racial IQ disparities. This is the best explanation for all the available data and so the one we are justified in believing until presented with evidence to the contrary.

It is worth briefly setting education aside to note that many variables have been tested to see if the degree to which they correlate with different abilities is higher the more that ability correlates with general intelligence (its g loading). The only variable which has exhibited a practically significant correlation in this direction is genes (and maybe brain size). This should make us less optimistic about any of the other variables tested proving to be important causes of the black-white IQ gap.

Citation	Predictor	K	Correlation w/ g loading
Nijenhuis et al. (2014)	Headstart Gains	8	-0.8
Nijenhuis et al. (2015)	Adoption Gains	4	-1
Nijenhuis et al. (2013)	Flynn Effect	11	-0.38
Nijenhuis et al. (2007)	Retesting	64	-1
Woodley et al. (2019)	Lead Exposure	16	0.1
Woodley et al. (2016)	Brain Size	3 or 4	.35 or .07
Nijenhuis et al. (2014)	Heritability	6	0.38
Nijenhuis (2019)	Education	12	0.13

Quality of Education

Returning to education, it might be objected that the data I’ve looked at thus far deals with quantity of education rather than quality of education. We might also note that even if we granted (for the sake of argument) that personal education is a significant environmental explanation we must next ask whether it is an ultimately environmental explanation or just a proximately environmental one. I bring these two questions up at once because the same lines of research provides answers to them both.

Some people mistakenly think that minorities in America are disadvantaged with respect to educational opportunity. This belief may cause them to think that education is an ultimately environmental explanation. For instance, they might cite research showing that school districts with lots of minority students get less funding than average. This is misleading though because within districts there is a tendency to allocate resources towards schools with more minority students.

To get around this, we have to look at spending per pupil based on school, rather than district, level data. Murray and Rueben (2008) calculated spending per pupil for US schools between the years 1972 and 2002. They found the following: “In 1972, the ratio of nonwhite to white spending was .98; this trend had reversed by 1982, as spending per pupil for nonwhite students was slightly higher than for white students in most states and in the United States as a whole and has been for the past 20 years”

Thus, since 1982, spending on non-white students has been greater than spending on white students.

This issue was revisited by Richwine (2011) who found that spending on black students was 1% greater than spending on white students, while spending on Asian and Hispanic students were a few percentage points lower.

Turning to more specific measures of school quality, racial differences in class size were non-existent by the early 1970s.

Cocoran et al. (2003)

In fact, even in he segregated south racial class size differences were probably too small to matter by the late 1940’s.

Card and Krueger (1992)

Moreover, blacker schools have more experienced teachers with more formal education and more pay.

Cocoran et al. (2003)

Even in the segregated south, black and white teacher pay equalized in the 1950’s.

Card and Krueger (1992)

Thus, if anything, black students are advantaged relative to white students in their pre-college education. There is also significant pro-black bias in college admissions. When comparing people of equal qualifications, Black applicants are roughly 21 times more likely to be admitted into an American college, while Hispanics are 3 times as likely, and Asians are 6% less likely.

(The race columns show the odds of admission compared to those of white applicants when qualifications are held constant.)

Citation	School	Type	Black	Hispanic	Asian
Nagai (2008)	Arizona State	Law	1115.43	84.95	2.18
Lerner and Nagai (2002)	University of Virginia	Law	730.8	1.1	1.86
Nagai (2008)	University of Nebraska	Law	442.39	89.63	5.78
Armor (2004)	William and Mary	Law	267	0.66	0.66
Nagai (2008)	University of Arizona Law	Law	250.03	18.15	2.54
Lerner and Nagai (2002)	William and Mary	Law	167.51	2.47	3.29
Danielson and Sander (2014)	Berkeley	Law	121.6	18.2	1.6
Armor (2004)	University of Virginia	Undergrad	106	2.81	0.94
Nagai (2006)	University of Michigan	Undergrad	62.79	47.82	0.81
Lerner and Nagai (2002)	University of Maryland	Medical	20.63	2.51	0.68
Armor (2004)	North Carolina State	Undergrad	13	1.93	0.64
Lerner and Nagai (2001)	SUNY	Medical	9.44	4.08	0.76
Nagai (2011)	Miami University	Undergrad	7.99	2.16	2.14
Danielson and Sander (2014)	UCLA	Undergrad	5.15	1.92	0.85
Lerner and Nagai (2006)	US Naval Academy	Military	4.44	3.32	0.67
Lerner and Nagai (2001)	University of Washington	Medical	4.01	4.86	0.9
Nagai (2011)	Ohio State	Undergrad	3.33	4.3	1.47
Lerner and Nagai (2006)	US Military Academy	Military	1.94	1.2	0.68
Lerner and Nagai (2002)	George Mason	Law	1.13	1.09	1.74
Median	All	All	20.63	2.81	0.94
Mean	All	All	175.51	15.43	1.59

In selective colleges, it’s been estimated that the proportion of students who are white would increase from 66% to 75% if admissions were based solely on test scores.

Carnevale et al. (2019)

Once minorities get into college, they are given greater access to grants. Specifically, Minority students account for 38% of the student population and 40.4% of grant funding. White students account for 61.8% of all students and 59.3% of grant funding.

Kantrowitz (2011)

There is also a slight trend such that white students are more likely than black students to be employed while in college. Asian students are less likely than white and black students to be employed.

NCES (2019)

Black, Hispanic, and white students also have similar chances of their parents paying for a significant proportion of their college education. Asians are more likely than others to have parental aid.

Rathmanner (2017)

In sum, the evidence suggests that minorities, especially blacks, are actually privileged with respect to educational opportunity relative to whites. This is true in terms of the ease with which they move into higher levels of education as well as the quality of the educational institutions they learn in when young.

It is possible that these advantages are more than made up for by some other environmental deficiencies from outside the educational system which impacts educational attainment. But we’ve already seen that income plays no role in the racial IQ gap and so we should not be tempted to name any deficiency that is the result of poverty.

The absence of any such deficiency is suggested by the fact that black people have higher educational attainment than white people do once IQ is controlled for. Thus, black people have a net advantage in educational opportunity with respect to the set of factors impacting educational opportunity other than IQ.

Herrnstein and Murray (1994)

But of course they have a disadvantage with respect to IQ itself leading them to have a net lower level of educational attainment than do white people.

Thus, racial differences in educational quality seem to favor blacks and thus cannot be a reason why they score lower on IQ tests.

Returning now to the question of whether educational quantity is an ultimately environmental explanation, it should be clear that we cannot answer this question without first answering the degree to which the IQ gap is due to the environment excluding education. If the IQ gap itself is entirely due to genes and educational gaps are caused by IQ gaps then educational gaps will be a proximately environmental but ultimately genetic explanation for (part of) the IQ gap.

Importantly, even if the IQ gap were partly environmental education could still fail as an ultimately environmental explanation. For educational opportunity to succeed as an ultimately environmental explanation, the educational loss generated by the environmental portion of the IQ gap must be larger than the educational gain generated by affirmative action. If this condition is not met, we will have to conclude that if environments were equalized across races group disparities in education would increase and, as a result, racial disparities in IQ would increase as well.

Regardless, we cannot properly classify education until we’ve formed an opinion on the heritability of the group IQ gap excluding the impact of education first. More importantly, we still have no good reason to think that education impacts the sorts of cognitive abilities that the races differ in, and have positive reasons to think that it does not, and so are most justified in thinking that education does not contribute to the black white IQ gap.

Nutrition

Another variable worth considering is nutrition. In the US, there are significant racial discrepancies in the frequency of deficiency for vitamins D, iron, and iodine. To the degree that such deficiencies impact cognitive ability, they will contribute to racial IQ disparities.

Reference	Nutrient	Whites	Blacks	Hispanic
CDC (2012)	Vitamin A	0.2	0.5	–
CDC (2012)	Vitamin B12	2.2	1.2	1
CDC (2012)	Vitamin C	7.1	4.3	3.1
CDC (2012)	Vitamin D	3.6	31.1	11.3
CDC (2012)	Vitamin E	0.6	1.2	–
Caldwell et al. (2011)	Iodine	25.7	31.2	21.9
CDC (2012)	Iron	9.8	4.9	11.4

Vitamin D is not thought to importantly impact cognitive ability, but iodine is and iron might, as evidenced by meta-analytic reviews of what happens to the IQ of deficient subjects when they are given iron or iodine supplements.

Citation	Nutrient	Effect of Supplementation
Sachdev et al. (2005)	Iron	0.41
Protzko (2017)	Iodine	0.53
Protzko (2017)	Iron	0.032

Using the racial gap in the rate of deficiency and the effect sizes from supplementation research we can estimate the role that iron and iodine play in racial IQ disparities.

Nutrient	B-W Gap in  Nutrient	W-H Gap in Nutrient	Impact on IQ Gap (BW)	Impact on IQ Gap (WH)
Iron	0.049	0.016	0.30135 points	0.0984 points
Iodine	0.055	0.038	0.43725 points	0.3021 points
Iron	0.049	0.016	0.02352 points	0.00768 points

What we find is that iron and iodine may contribute as much as a third of a point to the Hispanic-white IQ gap and perhaps as much as 0.7 IQ points to the black-white IQ gap.

Importantly, these nutritional differences by race cannot be assumed to be environmental in origin. There are obviously genetically caused racial differences in nutritional intake. For instance, it is well known that dark skin is thought to directly interfere with people’s absorption of vitamin D from the sun. The link between genes and iodine follows from the fact that, in America, the major source of iodine is dairy, and most African Americans are lactose intolerant while the vast majority of whites are not. Thus, some portion of the .44 point IQ disparity between blacks and whites that may be caused by differences in iodine consumption is probably itself caused by genetic differences between racial groups. The reasons for group differences iron deficiency are not clear, but there are no grounds for assuming that they are not genetically caused. Until further research is done, we cannot say with justification that any of the nutritional explanation of racial IQ gaps is ultimately environmental and, as explained, we have good reason to say that at least some of it is ultimately genetic.

Lead Exposure

Next let’s consider exposure to lead. There is no doubt that lead negatively impacts IQ. The best estimate of the effect size I’m aware of comes from Lanphear et al. (2005) who combined data from 7 longitudinal studies which looked at lead and childhood IQ. These studies controlled for an impressive list of potentially confounding variables including race, sex, birth weight, birth order, maternal education, maternal IQ, maternal age, marital status of parents, prenatal smoking status, prenatal alcohol use, and HOME inventory score which is a measurement of how cognitively stimulating a child’s home environment is. This research is not experimental, but it is as good of an estimate for the effect size that we are likely to get when dealing with a neurotoxin.

Given these effect sizes, we need only to know the size of racial disparities in lead exposure to estimate the contribution that lead exposure makes to racial IQ disparities.

White et al. (2015) reviewed 5 national samples taken between 1988 and 2004 and found that African American children had a mean BLL about 1.4 ug/dl higher than non-Hispanic White American children did. More recent data from the CDC shows that in 1999-2004 Black children had a mean BLL which was 1 ug/dl higher than White children and that this BLL gap shrank to 0.9 ug/dl in samples taken between 2003 and 2006, and to 0.5 ug/dl in samples taken between 2007 and 2010. The black-white gap in blood lead levels diminishes to virtually nothing by adulthood. CDC numbers [taken from Tsoi et al. (2016) and Pricket et al.(1994)] show that BLL levels no longer differ significantly by race among the total population.

(The CDC and White et al also looked at White/Hispanic children differences in BLL and both found that there were no consistent differences between these groups suggesting that lead does not play a significant role in the Hispanic-white IQ gap.)

Given that black-white differences in blood lead levels are not present in adulthood, lead could only impact the black-white IQ gap in adulthood via the lasting consequences of childhood lead exposure on adult IQ.

Research has not demonstrated that this sort of lasting effect exists. But, for the sake of argument, let’s assume that the effects do persist. Based on the analysis from Lanphear et al., a 7.6 ug/dl difference in BLL (2.4 to 10.0) leads to an IQ difference of 3.9 points. The most recent CDC data available suggests that Black children have a mean BLL about 0.5 ug/dl higher than Whites which is 6.57% as large as a 7.6 ug/dl difference. This, in turn, should have an effect on IQ about 6.57% as large as -3.9 points, which is 0.26 points.

One complication is that the effect of lead on IQ may even be stronger at this very low level of BLL (1.8 vs 1.3) than is indicated by Lanphear’s estimate of the effect within the 2.4-10.0 ug/dl range. However, this effect is almost certainly less than one IQ point. Thus, we are justified in thinking that lead exposure may be a small environmental cause of the black-white IQ gap assuming childhood lead exposure has effects that last into adulthood with no decay.

Now we must ask about the cause of racial disparities in lead exposure. The most obvious explanation here is poverty. But we’ve already seen that childhood poverty does not impact adulthood IQ. If racial differences in lead exposure really are caused by poverty, then this implies that the racial lead exposure differences in childhood do not in fact lead to a significant difference in adulthood cognitive ability.

It isn’t clear what the cause of racial differences in lead exposure would be other than poverty. Until a better explanation is produced, we are justified in doubting that lead explains any of the adulthood black-white IQ gap. If a better explanation is found, we would then need to assess its ultimate cause as well. And even if the cause were ultimately environmental, it would explain less than a single point of the 15 points gap.

Education Oriented Culture

The next environmental variable to consider is culture. Sometimes, it is suggested that Blacks and Hispanics lag behind whites and Asians because they have cultures that place less value on education.

The first problem with this hypothesis is that it isn’t clear that such racial differences in culture actually exist. Black parents are more likely than white parents to say that it is important that their child gets a college degree.

Pew (2016)

Consistent with this, Black and Hispanic students are also more likely than whites and Asians to have parents who check to see that their homework is completed.

NCES (2011)

There are some differences that favor whites when you ask students to rate, on a 4 points scale, how far they intent to go in school. But these differences are less than 0.2 SD and so are practically negligible (SD = .49).

Mau et al. (2000)

In another survey, racial differences on measures of family involvement in school, commitment to school, and family attitude towards education, where consistently found to either be practically insignificant (d<.20) or favor minorities.

Kue (2019)

Thus, stereotypes about large racial differences in the value placed on education appear to be unjustified.

Of course, even if such differences did exist we’d have to ask what causes them to exist, and the answer to that could very well be genes. And even if we determined that such differences were due to the environment, there is no reason to think that they would have a large impact on IQ.

Actually, there’s some reason to think that they wouldn’t. While not a direct measure of how much parent’s value education, adoption samples show that related constructs like parental permissiveness, and parental involvement, have no impact on adolescent IQ scores.

Beaver et al. (2014)

Another way of examining the impact of culture, as well as some other variables I’ve already considered, is to look at trans-racial adoption studies. If the white home environment is the cause of the black-white IQ gap, then the gap should not be present among blacks raised in white homes.

However, when interpreting trans-racial adoption studies it is important to remember that people do not adopt children at random. A particular concern here is that smart parents may tend to prefer to adopt children who seem bright. This may lead to children with abnormally high genotypic IQs having an elevated chance of being picked for adoption both in general and particularly by smart parents and this in turn will reintroduce a genetic correlation between IQ and SES even in adoptive homes.

This is the likely explanation for some of the studies in the trans-racial adoption literature on race and IQ. These samples are easy to spot: in them all the groups compared, of any race and whether adopted or not, will have very high IQ scores. For instance, in Moore’s study (see below) even black children raised in black homes had a mean IQ score of 118. Thus, in this study the normal racial IQ gap is not present in any context and so cannot be studied.

Excluding such studies leaves us with only a few studies which lack large samples. However, they consistently suggest racial groups end up with the normally observed sized gaps by adolescence.

Thomas (2016)

Thus, trans racial adoption research does not support the view that the home environment causes the black-white IQ gap, but it is low quality evidence due to lacking sample sizes.

In sum, the relevant empirical evidence fails to justify the existence of these supposed cultural differences and the effect of such differences on IQ scores. We are therefore justified in thinking that this aspect of culture plays no role in racial IQ gaps until we encounter better evidence to the contrary.

Self Image

Another cultural hypothesis is that some groups have internalized negative stereotypes about themselves which in turn inhibit their cognitive abilities. There is no empirical support for this view. Meta-analyses show that Blacks score higher than Whites on measures of self-esteem.

(Twenge and Crocker, 2002)

In fact, this has been true for the last 50 years.

(Twenge and Crocker, 2002)

(Twenge and Crocker, 2002)

It may be objected that self esteem is too general of a concept since what we are really concerned with is specifically how intelligent people see themselves as being. This objection can be easily countered by examining questions which specifically ask about self-perceived intelligence:

“Black students estimate their own academic competence more highly than Whites despite their own objective and self reported lower academic achievement (Hare 1985: Table 3; Tashakkori 1993: 97). Black high school seniors in the Coleman study were more apt than Whites to classify themselves as “among the brightest” and less likely to agree that “Sometimes I feel that I just can’t learn” despite poor academic performance (Coleman et al. 1966L 287-288, Tables 3.13.11, 3.13.12). Remarkably, southern rural Blacks, whose academic performance fell below that of all other blacks as well as all Whites, were more apt than other Blacks to classify themselves as Bright and able to learn” – Levin, 2016, page 75

Thus, these sorts of cultural explanations fail to explain racial gaps in intelligence because they are based on false accounts of how self image differs between races.

Racism

The final environmental variable I want to consider is racism. In large part, the idea that racism causes racial IQ differences has already been refuted. This is because racism would need to impact IQ through some mechanism like education, poverty, or self image, and we’ve already seen that each of these hypotheses fail.

Sometimes, it is suggested that slavery provides a mechanism by which racism might impact IQ. Of course, no African Americans alive today were slaves, so this effect must be indirect. We are thus left once again wondering how this effect is supposed to work if not through socio-economic status or self image.

Now, there is evidence that slavery had a multi-generational effect on African Americans. However, this effect ended long ago. This is evidenced by comparisons of the descendants of free blacks and those who were enslaved. Beginning soon after emancipation, those black people who were freed from slavery were less literate, and poorer, than black people who were born free. This difference persisted for some time, but after two generations the descendants of enslaved and free black Americans did not statistically differ in terms of both education and economic success (Sacerdote, 2002). This suggests that the effects of slavery had largely faded by the time we get to the grandchildren of slaves and makes it highly improbable that such effects still linger today.

Further evidence against racism as a cause of the black-white IQ gap can be found by examining how that gap has changed with time. While it narrowed earlier in the 20th century, among adults the black-white IQ did not significantly narrow between the 1970s and the 2000s.

(WAIS, , AFQT, Wonderlic, and NAAL data from Chuck (2013). The WAIS 2000s sample comes from Holdnack et al. (2013) p.190. WAIS 1980s sample came from Vincent (1991).)

During this same time period, there was a steep decline in racist attitudes among whites.

Krysan and Moberg (2016)

This suggests that racist attitudes among whites is probably not a cause of the black-white IQ gap.

Because the word “racism” is not a scientific hypothesis and the channels by which racism might be thought to interact with IQ have already been disconfirmed, we are justified in thinking that racism does not explain any significant portion of the black-white IQ gap.

Conclusion

In conclusion, most environmental variables people think might be involved in the black-white IQ gap are not either because the variable does not impact IQ (or G) or because it doesn’t differ across races in the needed direction. Some variables, like nutritional intake, do impact IQ and disadvantage blacks relative to whites, but even if we sum variables like nutrition and lead exposure together their contribution to the black-white IQ gap is less than a single point. Further more, there is no evidence that group differences in things like lead exposure and nutrition are not themselves due to genetics and there is some reason to think that they probably are. Thus, there are no known causes of the black-white IQ gap that are known to be ultimately environmental and the non-existent or near non-existent contribution of variables often thought to be involved with the gap should decrease our confidence in environmental explanations for the black-white IQ gap in general and increase our openness to entirely or nearly entirely genetic explanations.