Intelligent racists these days tend to defend their bigoted statements by prefacing them with a disclaimer that “What I say is only true on average,” or “This is just a statistical tendency.” Derbyshire does the same thing at the beginning of his rant. But then he goes on to make prescriptive statements about individuals based on those “statistical tendencies”! He goes from “On average, black people are more violent” to “If you see black people, you should avoid them.”
This is a statistical and logical fallacy, as any researcher knows. Descending into nerdspeak for a moment, statistical significance is not the same thing as effect size. In plain English, this means that whether a person you see on the street is black is not a useful indicator of how likely they are to attack you. There are much better ways of predicting who is likely to be violent, just from looking at them.
For example, take a look at these two screenshots (the former from the movie Boyz n the Hood, the latter from the TV show Community):
If you’re walking around and see people who look like the guys in Screenshot A, you might reasonably want to avoid them (though actually I wouldn’t, but save that for later). Why? Not because they’re black! Because they look poor, tough, and mean. Those would be good reasons to avoid someone who looked like Ice Cube looks in Screenshot A. Their blackness would be beside the point. It would not be an important consideration, on top of the other things you could already see. If they were Vietnamese, there would be absolutely no less reason to avoid avoid them.
Second, a really wonderful piece from Noah Millman of The American Conservative, making similar points, but from the perspective of someone who wants to live a full and worthwhile life:
Derbyshire gives nine specific pieces of advice that he calls applications of “statistical common sense”…. Every one of these injunctions is bad advice. To be a good application of statistical common sense, it’s not enough to know that, for example, crime rates (on average) are higher in majority-black neighborhoods. You’d need to know that the disparity was large enough, and the variance around the average small enough, so that following such a rule would actually be a decent heuristic; not to mention that there were no more finely-grained heuristics available and that the cost of applying such a sweeping heuristic in terms of the loss of experience of life and its manifold pleasures was not prohibitive.
Because here’s the thing. Granting that nobody has an obligation to be politically correct in their behavior, and granting (for the sake of argument) all of Derbyshire’s premises, what he’s still saying is that the risks are so great that it’s better simply to wall oneself off from African-Americans to the greatest degree possible. But he hasn’t actually measured the risks in absolute terms, only in relative terms: would this action reduce risk; if yes, then follow it. I wonder: does he take a similar attitude toward other risks? Toward, to take a few examples, eating raw food, bicycling without a helmet, traveling alone to a foreign country, or writing whatever one wishes for a publication like Taki’s Magazine?
I live in Brooklyn. I love living in Brooklyn. Do I live in a majority-black neighborhood? No. But there’s a large, majority black neighborhood right across the park from me. We share the park. Derbyshire’s advice to me and to my son is, effectively: don’t go to the park. Or, alternatively, don’t live in Brooklyn. But why? Does Derbyshire know what the crime statistics are like in my part of Brooklyn these days? Is he really that fearful?
The “race realists” like to say that they are the ones who are curious about the world, and the “politically correct” types are the ones who prefer to ignore ugly reality. But the advice Derbyshire gives to his children encourages them not to be too curious about the world around them, for fear of getting hurt. And, as a general rule, that’s terrible advice for kids – and not the advice that Derbyshire has followed in his own life.