Is skin tone correlated with job prestige?

The General Social Survey rated the skin darkness of a sample of black Americans, ranging from "very dark brown" to "very light brown." Biologically oriented researchers might see the question as a rough measure of the percentage of European ancestry, while sociologists would see it as a measure of discrimination--lighter-skinned blacks getting better treatment. Respondents were also given a job prestige score that ranges from 16 to 80--16 is a shoeshine and 80 is a physician. Here are the job prestige means listed by skin tone:

You can see that average job prestige tends to rise with lighter skin. The mean for blacks with "very light brown" skin is roughly two-thirds of a standard deviation higher than the mean for "very dark skin" blacks. 

The pattern can be interpreted in at least two ways: 1) genetic--blacks with more European ancestry tend to rise in the status hierarchy much more than African blacks, or 2) sociological--whites discriminate more against darker blacks, and perhaps light-skinned blacks have white (privileged) relatives who gave them advantages. 

For several reasons, I'm inclined toward the genetic explanation. For one, my experience is that when a white person is interacting with a black person, his thought is, "I'm talking with a black person," not, "I'm talking with a light-skinned black person." For another, how do those Nigerian immigrants do so well in the US when their skin tends to be so dark? The sociologist would predict severe discrimination. The biologist would argue that African immigrants are a select group of Africans--above average in IQ and drive--and this overwhelms any bias they might experience. 

If IQ were skin color, how visible would the racial gap be?

I often describe the differences between groups in terms of standard deviations or Cohen's d, but maybe we can get a clearer picture of what these mean with a black-and-white example; namely, skin tone.

Skin color is a very noticeable racial difference and tends to dominate discussions of race even though it is just one characteristic. People focus on it because it is obvious, and I suspect race deniers like it because they can treat it as something superficial, something only shallow people would focus on. People like Dinesh D'Souza who I often agree with will say idiotic things like race is nothing but a "coat of paint." Never mind that skin is our largest organ and carries out many important functions, and that skin color appears to have important functions as well. And anyone who claims that aesthetics aren't important doesn't know human nature.

Anyway, General Social Survey interviewers were asked to rate the skin tone of respondents with a number ranging from 1 (lightest) to 10 (darkest). The lightest ethnic group is Americans of Danish descent with a mean of 1.33. (I'm weird because I'm all NW European and 1/8th Danish but look like an Arab). We'll use Danes as the comparison group. What I show below is the gap between the Danish mean and the mean for the focal group in terms of standard deviations:

Degree to which group is darker than Danish Americans (in sd units)

Blacks  2.2
West Indians  2.2
Asian Indians  1.3
Mexicans  0.8
Filipinos  0.8
American Indians  0.8
Puerto Ricans  0.7
Chinese   0.6
Japanese   0.6

Total Sample  0.6

Spanish   0.5
Arabs   0.5
Jews  0.3
Portuguese  0.2
Italians  0.2
Greeks  0.2

A gap of 2.2 standard deviations between Danes and blacks is so huge, it's the reason why people decided long ago to define groups as the opposite of each other: black and white. If a study reports a 2 sd gap, be impressed.

Another way of describing the gap goes like this: not a single black in the sample is lighter-skinned than any Dane, and 97% of Danes are lighter than 94.5% of blacks.

If you want a visual that matches the size of the black/white IQ gap (1 sd), think of the color gap between Danes and a typical Mexican and typical Asian Indian blended together--a very large difference. Another way of saying this is that the black/white IQ gap is only half the size of the color gap.

Even a difference that is only 10% of the black/Dane gap is still visually easy to detect: the Italian and Greek means are 0.2 standard deviations darker than Danes--0.2 considered a very small difference--but southern Europeans are known for being darker than northern Europeans. The higher Spanish mean is probably due to darker Hispanics claiming Spain as their homeland.

Data: Lighter blacks tend to be smarter

Recently, I documented that of the four racial groups examined -- whites, Hispanics, Asian Indians, and East Asians-- four showed a positive correlation between lighter skin and IQ.  But what about blacks?

This graph shows mean IQ for blacks born in the US at increasing dark skin tones (as rated by General Social Survey interviewers, N = 658).  I am using skin tone as a rough measure of the degree of European ancestry:

Tones 6-10, especially 10, have lower average IQs than tones 1-5.  The difference between level 1 (IQ = 93.6) and level 10 (IQ = 86.4) is well over half a standard deviation.

So blacks with more European ancestry tend to have higher IQs. The explanation that light-skinned blacks receive better treatment than dark blacks is unconvincing.  When someone of another race encounters a black person, you say "black person" to yourself, not "light-skinned black person."  And anyway, there's no evidence that poor treatment makes somebody dumber.  If it did, Jewish Holocaust survivors would be morons, not people with above average IQs.

Data: Among East Asians, are lighter-skinned people smarter than darker ones?

Three times in a row (for Hispanics, whites, and Asian Indians), I've shown that the lighter skinned members of a group tend to be smarter than darker members.

I now find the same pattern for a small sample of East Asians born in the US (n = 24). The correlation between darkness and IQ (based on a vocabulary test) is -.16 -- a small effect.

Note: A potential confound occurs to me. Do lower IQ people spend more time in the sun?  With larger samples, it might make sense to look at the sexes specifically.  Lower IQ men might work outside more, but I wouldn't expect this for low IQ women. 

Data: Are lighter-skinned Asian Indians smarter than their darker counterparts?

As I've done in the last couple posts, I correlated the skin tone of General Social Survey participants with IQ scores -- this time for people whose ancestors are from India.

The sample is small, only 39, but the correlation is impressive: -.40. In plain English, the tendency for lighter Indians to be smarter is fairly strong.

This is the third group in a row (Hispanics, whites, now Indians) that has showed the same positive lightness/IQ correlation.

Rushton on the skin color/behavior correlation

Phil Rushton and colleagues at the journal Personality and Individual Differences continue to push the the theory that genes explain the worldwide correlation between skin tone and important behaviors:

Pigmentation of the hair, skin, cuticle, feather and eye is one of the most salient and variable attributes of vertebrates. In many species, melanin-based coloration is found to be pleiotropically linked to behavior. We review animal studies that have found darker pigmented individuals average higher amounts of aggression and sexual activity than lighter pigmented individuals. We hypothesize that similar relationships between pigmentation, aggression, and sexuality occur in humans. We first review the literature on non-human animals and then review some of the correlates of melanin in people, including aggression and sexual activity. Both within human populations (e.g., siblings), and between populations (e.g., races, nations, states), studies find that darker pigmented people average higher levels of aggression and sexual activity (and also lower IQ). We conceptualize skin color as a multigenerational adaptation to differences in climate over the last 70,000 years as a result of “cold winters theory” and the “Out-of-Africa” model of human origins. We propose life history theory to explain the covariation found between human (and non-human) pigmentation and variables such as birth rate, infant mortality, longevity, rate of HIV/AIDS, and violent crime.

As Razib Khan has explained on this blog, pleiotropy as the explanation of these correlations simply does not work. So what plausible theories remain? Any ideas?

Skin color and occupational prestige among blacks

As a follow-up to the last post, I looked to see if skin color among blacks is related to occupational prestige. Interviewers rated skin color from very dark brown to very light brown. Here are the mean prestige scores (GSS data, sample size = 454):


Mean occupational prestige scores

Very dark brown 28.3
Dark brown 28.4
Medium brown 35.3*
Light brown 33.2
Very light brown 42.7*

*significantly higher than the very dark brown group

Lighter groups tend to have better jobs than darker groups. This is consistent with the common observation that elite blacks often look like they have more European ancestry.

Skin color and desirable traits

Donald Templer and J. Phillippe Rushton have recently published a study of the 50 U.S. states which finds strong correlations among the following varaibles: skin color, IQ, infant mortality, life expectancy, violent crime, HIV/AIDS, and income. This pattern mirrors what Templer has observed internatioanlly. Skin color tends to be more strongly correlated to the list of variables than does income, so the authors conclude that this points to biology rather than economic environment as the more important explanation of the pattern of correlations. They rely on Rushton's life history theory and Lynn's cold climate theory to make sense of why the variables hang together, but whatever the explanation, there seems to be a worldwide pattern of desirable (or undesirable traits) running together, and this is closely associated with skin color.  

They review some interesting research by Ducrest et al. (2008) which found that in 20 wild vertebrate species, darker individuals are more aggressive, sexually active, and resistant to stress. They also tend to have a larger body mass and greater energy and physical activity. For example, darker maned male lions are reported to be more aggressive and sexually more active, and darker barn owls show stronger immune response. Darker coloration is associated with enhanced fertility, female sexual receptivity, male sexual motivation and performance, and higher levels of testosterone. Ducrest et al., however, cautioned that these findings might not apply to humans.  

Skin color and educational level among blacks

ADD Health interviewers judged the skin color of respondents as black, dark brown, medium brown, light brown, or white. Limiting the sample to just African Americans, let's treat the measure as a rough indicator of the amount of European ancestry. Here is mean educational level by skin color (sample size = 1,211):

Mean educational level

Black 12.8
Dark brown 13.0
Medium brown 13.3*
Light brown 13.5*
White 12.8

*significantly higher than black category


Medium- and light-brown blacks are significantly more educated than black-skinned African Americans. White blacks--only 16 respondents--are the exception to the trend.

Skin color and IQ

I've copied the correlation table and the results from a 2008 study in Personality and Individual Differences (Templer, Donald. 45, 440-444). From the author's conclusion:

The finding that skin color shows substantial correlations with life expectancy, birth rate, and infant mortality confirms the importance of this variable, as originally identified by Templer and Arikawa (2006). It merits further research given the suggestion made by Jensen (2006) that pleiotropy(genes having more than one effect) may underlie both IQ and skin color. Skin color and IQ have been found to be correlated in several studies of people of mixed-race ancestry Jensen, 1998; Rushton, 2008; Rushton & Jensen,2005).

The nature of a latent common factor such as the r–K dimension requires inference and interpretation. Life history theory provides a central organizing theoretical principle for generating specific predictions as to which traits should be correlated and it suggests physiological processes such as brain functioning that provide a substrate (Jensen, 1998; Rushton, 2004). This latter approach was adopted by Figueredo et al. (2006) who proposed an integration that traced the development of a life history from genes to brain to the various associated components of the reproductive strategy. They explored the psychometrics and behavioral genetics of the K factor, and specific areas of the human brain, including the frontal lobes, amygdala, and hippocampus. They then applied life history theory to predict patterns of development within the brain that are either paedomorphic (i.e., development begins later, proceeds at a slower rate, and has an earlier cessation) or peramorphic (i.e., development begins early, proceeds at a faster rate, and has a later cessation). The theory of r–K reproductive strategies may eventually provide an evolutionary understanding of many seemingly disparate results in personality research.