The first priority for race realists should be to establish beyond reasoned debate that there is, in fact, at least a small genetically conditioned difference in g between races. Flynn and Dickens (2001) have generously provided an extravagant model by which small differences between races can amplify in absence of disparate treatment. Small genotypic differences can, theoretically, go a long way. In addition, Gordon (1997) has shown how the impact of individual IQ differences can multiply in aggregate. Gordon’s basic conjecture has been borne out in numerous studies. The impact of IQ differences magnify as you move from individuals to households, to school districts, to states, to nations, to global regions (page 51 on). Taken together, contrary to what some have argued, small genetic differences between populations can, theoretically, account for large population differences in social outcomes. It follows, then, that by simply establishing the existence of such differences, one can shift the burden of proof in the debate about culpability, which has long been at the heart of the Race debate. The onus would be on ‘social constructionists‘ to show that various outcome differences didn’t have an ultimate genetic etiology (in a biometric sense).
Given this, it’s somewhat disheartening to see the direction that recent race realist research has taken — dubious arguments about penis size and pleiotropy and intra-racial differences supported by questionable data: