Footnote 4 to the introduction of Part III recaps the argument. If the black/white IQ gap in the U.S. were caused by racism, you'd predict that blacks would have higher IQs in countries where they're the ruling majority, like in Haiti or most of sub-Saharan Africa, but they don't. You might think racism affects IQ by means of its affect on socioeconomic status (SES), but adjusting for parental SES only diminishes the gap by a third.
Footnote 4 to the introduction of Part III recaps the argument. If the black/white IQ gap in the U.S. were caused by racism, you'd predict that blacks would have higher IQs in countries where they're the ruling majority, like in Haiti or most of sub-Saharan Africa, but they don't. You might think racism affects IQ by means of its affect on socioeconomic status (SES), but adjusting for parental SES only diminishes the gap by a third.