THE BARREL HAS NO BOTTOM.

Among the National Review new breed Kevin D. Williamson and Charles C.W. Cooke are pretty far out there, but don't sleep on David French, who despite lacking his colleagues' hipsterish affectations (peculiar hair balance, two middle initials) is more than a match for them twaddle-wise and seems to drift further from earth's orbit with each passing column. The title of his latest:

Are Encounters with the Police Really More Dangerous for Black Men?

If you guessed "Nah, son!" you've been paying attention. French starts with a story about how as a lad he himself was roughed up by the constabulary, but generously offers further evidence:

The results so far for 2015 show much higher numbers of police killings than previous FBI reports. They also, at first glance, seem to prove the #BlackLivesMatter thesis that police target black men. 

As of July 27, the Guardian claims, American police have killed 657 people in 2015. The large majority, 492, were armed. Some 316 victims were white, 172 black, and 96 Hispanic. (The rest were of other or unknown ethnicities.) Whites constitute a majority of the population, however, and police kill black Americans at a greater rate than whites — with 4.12 black victims per million versus 1.59 white victims per million. 

So case closed, right? Not so fast. Comparing police shootings by race with crime statistics by race tells an entirely different story: It may in fact be the case that white Americans are ever-so-slightly more likely than blacks to die in any given encounter with a police officer. After all, blacks commit homicide at eight times the combined white/Hispanic rate, and, despite their constituting roughly 13 percent of the population, represent a majority of homicide and robbery arrests. Indeed, the disproportionate share of arrests exists across all categories of violent crime — at a rate that often exceeds the racial difference in police shootings. Thus, blacks are seriously overrepresented in the most dangerous police encounters of all — encounters with violent suspects.

Go ahead, read it again. He really is saying it: That while in raw numbers blacks do get killed by cops more often than whites, you have to grade on the curve because blacks are so criminal.

The rest is also gibberish, though some of it is prime:

It’s just sheer fiction that white men enjoy some sort of shield of immunity, engaging in disrespect and defiance at will. After all, police kill white men almost twice per day.

This is where I'm supposed to lament how far National Review has fallen, but except for its arts and letters coverage it always sucked; all that's interesting about the new Review is that they've found people who are willing to say absolutely anything to keep their jobs.

UPDATE. Comments are as ever prime, and include a link to a few good explanations, as if they were needed, as to why French is full of shit: montag2 offers Jacobin's "The Making of the American Police State"; Robert M. offers, in response to an industrious troll, the insight that French's "principal error is conflating 'encounters with police' with the incidence of crime, and the incidence of crime with arrest rates" -- assuming, perhaps over-generously, that this was not deliberate.