There are two common viewpoints on how to choose policemen. Either hire:
The best individuals who apply, or
The best racial balance
I think both arguments have some surface plausibility because the police are, as the word suggests, inherently political.
It’s not unreasonable to imagine that since such a disproportionate fraction of homicide perps, victims, and witnesses are black, police forces should strive to have a number of blacks. By way of analogy, although female firefighters have proved fairly useless (it’s not coincidental that none of the 343 firefighters killed on 9/11 were women), women cops ought to be valuable for, say, getting evidence from rape victims.
Similarly, a lot of police work in modern America is persuading black witnesses to show up to testify against black criminals.
“The superior solution would be to hire cops in rank order within their race’s quota. That way, police departments would get the best blacks and best whites rather than the random flotsam chosen by lottery.”
Strangely, though, nobody has ever produced much evidence that diversity hiring improves policing. In the largest such natural experiment, Congress ordered Mayor Marion Barry’s Washington D.C. to hire 1,800 additional police officers in 1989–90. By all accounts, this proved a disaster. As of 1994, over 100 of D.C.’s new cops had themselves been arrested.
Similarly, advocates of diversity hiring seemed compelled to lie about the evidence. For example, the federal consent decree against the LAPD in the Rampart scandal was sold to the public as being due to revelations of white racism among cops. Yet the main malefactors were all diversity hires, some of them moonlighting for felonious hip-hop producer Suge Knight. Strikingly, the Oscar-winning movies Training Day and Crash were more realistic about the race of the Rampart bad guys than was the news coverage.
