Holly's recent post on race, Wade's book, and HBDers identified some of the core issues that separate many anthropologists from their often openly or eagerly racist colleagues. Some differences are scientific, but others really are emotional differences about sociopolitical views, whether or not that's actually stated or even understood. This is so contentious, and indeed has been since Darwin himself if not, in pre-evolutionary terms, for the entire history of human societies, that measured discussion is rare. People take positions and, in their righteousness, suspect (or worse) anyone who might challenge their faith. Even scientists are, after all, people.
Of course we know from clear, recent (as well as deep) history what can happen with explicit or covert racist views of human variation and evolution -- that was one motivation for our recent post on Mussolini, science and race. Policy can use the scientific rationale to treat all members of a labeled race as if they were equally defective or superior, even if some weak caveats are added that this is only an approximation of average differences between groups. Based on categorical thinking, policy decisions can allocate resources to enhance what is judged by the supposed experts to be superior so we can help Nature's own evolutionary path, or to deprive members of the inferior group. Depriving need not be so draconian as some lethal episodes in race history, because we could simply deprive a group, as a whole, of resources such as investment in education. Many, even some commenting on or about posts here, have made such statements. This is part of the historical reality of these race-based ideas, and those making the assertions know this very well, or should.
One might agree that different categories of people, as s/he constructs them, inevitably have statistically different traits on average. This is almost inevitably the case, and the differences will include genomic effects. But why not make policy based on individual traits rather than on a group basis? Many arguments-of-convenience are standard: it's a waste of money to invest in inner city schools because, even if some of the kids have above average IQ most won't, etc.; or don't give aid to Africa because they inherently cannot measure up to our standards.
Even accepting standard geographic categories as the definition of 'race', such as European, Asian, African, it is obvious that there is large overlap between the groups in most of the kinds of traits being discussed (like mate choice, sexual behavior, criminality, IQ, athletic ability....), which means, for example, that many members of the inferior group score higher or behave like members of the supposed superior group, and vice versa. So from a social policy point of view, unless the one defining categories just wants reasons to keep resources in his/her category, individual based policy would seem to be far more equitable (if one believes in being equitable). Even diseases with a genetic basis that may be group specific (until recent admixture) are far from present in every member of the affected group, and again, determining on an individual basis who's got the relevant allele is much more sensible than assuming it on the basis of membership in a group.
There are problems here even in legitimate disease-related contexts, and they merit quiet, measured attention which they don't usually get. Many racist or commercial expedients seek to milk categories for drug sales, and others for sociopolitical reasons deny the levels of variation that we know exist. For example, we can't just identify individuals with a given genetic risk variant without taking their geographic ancestry into account when making medical diagnoses or treatments. The same genetic variant may be more common in one population, say, Europeans, than in another such as Africans. But the variant often if not usually turns out to have different average effects on the individual that are origin-specific. This can have to do with cultural differences in life-history exposures, which are often very hard to tease out, but can also have to do with polygenic background differences related to variants in the genome that are not measured but are different in different populations.
Such issues are real and important and have to do with whether medical practice should view people in terms of 'race'--even in a benign world where it was recognized by everyone that this is only a substitute for geographic origin (and takes admixture into account somehow). If we could stop Pharmas from trying to use categories to cash in on putative 'race' differences in drug efficacy, and so on, and if we could disengage from the strident Darwinism that is prevalent in discussions of racial variation and its inherency and value, we might actually be able to address very important and scientifically legitimate issues. But we can't really do that amidst the invective in today's arena.
The realities of human genomic variation, and the reasons
The obvious fact is that the presence and frequency of genetic variants, genomewide, vary over geographical space in ways related to population history: the flow of variation by mating patterns of many generations, geographic barriers affecting migration and mating patterns (mountains, rivers, seas), physical and ecological conditions, climate, and so on. This is obvious and has been known long before we had any genetics to invoke.
Roughly speaking, the farther apart your ancestors the more different you'll be across the genome. This means that if you sample from distant areas, you can easily conclude that the people are distinctly different--indeed, statistically, this is true even if based on variants that are shared among the groups but have different group-specific frequencies. But this does not point to categorical differences, only to quantitative statistical ones. Even within each group, there is about as much overall variation as between them, a point long known, current Lewontin-bashing notwithstanding.
Analyzing data from discrete, widely separated samples automatically treats, statistically, the populations as categorical units. This obviously can lead to what appears to be categorical variation among human populations (we talked about this a few days ago here). But this is an illusion of grouping decisions and statistical analysis of the geographic pattern of variation need not use such meat-axe approaches because there are statistical methods that more realistically depict the actual more continuously distributed pattern of variation (and this, of course, is rapidly being modified by large-scale, long-range migration---such as by anyone living in North America rather than in Asia, Africa or Europe as our ancestors did).
These considerations are neither racist nor politically correct, and relate directly to Holly's point about species designations, a subject typically of heated controversy when it has nothing to do with the emotionally loaded subject of modern humans.
Given this variation, across the genome, one has to expect some of it to reflect local adaptations, and some to have serious effects on the individual. Much gene-specific disease susceptibility is found only, or more frequently, in one population compared to others.
The opposition to 'political correctness' that is manifest so typically if not in rabid self-satisfied proclamations by those advocating a strongly deterministic view of our species and its genomic evolution is, no matter if it's denied, essentially a confusion of two agendas. One is the reality that some political-correctness denies or minimizes aspects of reality or the nature of causation in humans; the other agenda is a very clear and cogent societal problem of class and other discriminatory distinctions in the quality of life in complex societies. In fact, it is ironically the social politics, and not the technical science, that largely underlies much of the heat in the 'discussion' of this subject.
We could, of course, have an honest political discussion of whether it's OK for some groups to be allowed to have a disproportionate fraction of the material and cultural wealth and deprive other groups, or whether we should be more egalitarian. If you agree on how to define the groups, this can be at least a fair discussion. Arch capitalists argue that inequality is a good thing, though usually they don't couch this in racial terms. Elitists have a right to view elitism as good, but this is a subjective value judgment and not something that can be justified on genomic or evolutionary grounds. Why, for example, is depriving someone of his or her access to food or education justified just because the person isn't as good as the professor is, at calculus? Even if that is based on a genetic difference, why does it suggest that one person can decide how many resources the other has? We are, after all, here as human individuals, and our genomes are the result of a shared history of evolution. How can one person declare the inherent worth of another, and on what kinds of legitimate scientific grounds--other than his/her personal, and hence subjective, opinion? Or are you just defending power and privilege that you happen to enjoy? Or are you denigrating differences because you want to gain power that you don't now have?
We'd be a lot better off, we think, if these distinct, scientific and societal, areas of discussion or disagreement were kept separate and treated on their own merits. But that doesn't seem to be in the cards at present.