1. A quick google scholar search delivers 1,540 hits for “colorism” since the year 2000. That’s quite a bit fewer than the 273,000 for “racism” but nonetheless substantial.
2. For those who don’t follow this debate, “colorism” is the new up and coming “racism.” I mean this in terms of “anti-discrimination” efforts not outcome differences, differences which have been noted in the US since the first event of racial admixture (section O, here).
3. The orthodox
position conclusion in the social “sciences” is that color outcome differences are not only the product but the proof of The Racial Discrimination System.
One can see this unsubstantiated conclusion presented in the popular press. For example, in the New York times, we are told:
The country is moving away from the blunt-force racism that once banished black people to the other side of the Jim Crow line. But we have entered a period of secondary discrimination — or “colorism’’ — that will be difficult to overthrow.
This point was alluded to in the 1995 report by the Federal Glass Ceiling Commission, entitled “Good for Business: Making Full Use of the Nation’s Human Capital”:
It should frighten us as Americans to realize that we still view one another through patterns that have a genesis in slavery. Blunt force racism may indeed be on the wane. But the battle against this more subtle and insidious form of discrimination has clearly just begun. (As Racism Wanes, Colorism Persists)
No to little consideration is given to alternatives such as: family environmental effects, genetic effects due to assortative mating, genetic effects due to ancestral genetic differences, or genetic effects due to plieotropy.
5. Incredibly — or not — this hypothesis has not been properly tested in the US. It has been so in Brazil and the results have not been, for the most part, in support.
The proper way to test if colorism accounts for the color differences is to compare the outcomes of siblings. A discriminatory hypothesis would predict that siblings who differ in color will differ in outcomes close to the degree that non siblings do. Specifically the following hypotheses predict the following:
Colorism & Plieotrophy: Same correlation between outcomes within and between families.
Family environmental effects: Zero correlation within families, regardless of sibling relatedness.
Ancestral genetic and assortative mating: Correlation lower within than between families; correlation varies by sibling genetic relatedness.
6. The dearth of tests has not been for a lack of data as such studies can be made using the nationally representative NLSY 97 and ADD Health surveys, which contain kinship, skin color, and outcome data and follow the participants though adulthood. (See my analysis as example.)
7. Nonetheless, for the first time, reports of a study in progress have surfaced, though one based on a sample from the early half of last century:
Disentangling skin-color discrimination from family background differences among African-Americans, 1910–1930
Evidence points to a significant and persistent wage gap between whites and blacks in the U.S. This gap could be attributed to discrimination, but it can also be driven by systematic differences in human capital that are correlated with race. An important determinant of human capital is family background. This paper disentangles discrimination from family background by comparing siblings who were coded by census enumerators as more or less black. We locate brothers in households that in 1910 had some sons coded as “Black” while others as “Mulatto” and we link them to the 1930 census. We then assess the extent to which brothers differ on educational outcomes in childhood and economic outcomes in adulthood. We find that outcome differences between brothers are much smaller than those between “Mulattoes” and “Blacks” selected randomly from the population. Still, darker skinned brothers are less likely to attend school in early childhood and are also less likely to be literate. Overall, our findings suggest that systematic differences in family background account for most of the educational and economic gaps observed in the population.
8. For contemporaneous results — and results which can disentangle family environmental and genetic effects, I suppose that we will have to wait for Chuck to finish his analysis.