What’s worse than violent police?

Violent racist police.

What’s worse than violent racist police?

Inconsistent violent racist police who become more brutal when they perceive threats.

This way you don’t get beaten up under some universalistic pseudo-fair consistent treatment.

You get beaten up based on whether the policeman you are dealing with is scared of blacks on your particular day.

Police comfort themselves emotionally when they are afraid by beating up blacks.

This is strong stuff.

But it is the primary finding of an important study by Joscha Legewie, a sociology professor at Harvard University. Legewie obtained data on all cases of police stopping a pedestrian in New York City between 2006 and 2012. For each of these stops, he knows the race of the suspect, the timing of the stop, the geographical location of the stop and whether the police used force on the suspect.

There already existed a significant number of studies showing that minorities are more likely to be stopped by police, that the rate at which minorities are stopped is greater than any racial difference in criminal offending, and that police are more likely to use force against black suspects. (See the reading list at the end for the relevant citations.)

Legewie’s data confirmed the general pattern of police being more likely to use force against minority suspects. New York City is 45% white, 25% black and 28% Hispanic. Despite the black population being less than half the size of the white population, only 10% of suspects in police stops were white. 54% were black. There were five times as many stops of blacks than whites.

It was not the case that blacks were committing five times as many criminal acts. Most police stops find no contraband, no weapons and lead to no arrests. Only 6% of police stops lead to an arrest. Only 2% lead to the finding of contraband. Only 1% lead to the finding of a weapon. Most people who are stopped have not done anything criminal and are let go.

Of those people who were stopped, whites were far more likely to have weapons and contraband than were blacks. Despite whites being more likely to have weapons, police were more likely to use force on blacks than on whites. The differences were modest: 16% of white stops led to the use of force while 22% of black stops led to physical force. But this is still a reversal of expectations, since the white suspects were more likely to be armed and dangerous.

*  *  *

Now for the issue of racialized panic.

There were four shootings of police officers during the period of Legewie’s study. Two of the killings of police involved black suspects. One involved a Hispanic suspect. One involved a white suspect. Legewie was able to look at the use of force against suspects before and after these killings to see if the killing of a police officer made a different in police procedurals.

The killings of police officers by white or Hispanic suspects had no effect on violent acts by police. The New York City police used the same amount of force on suspects after the white and Hispanic killings as they did before the incidents. They just continued doing their normal business.

The reaction of New York police was completely different when one of their own was killed by a black suspect.  The use of force went up. Worse, the use of force went up only for black suspects. There was no change in the way in which New York police treated white or Hispanic suspects. However, they were more likely to use physical force against black suspects.

Legewie notes that the police were not objectively in more danger after the two black shootings of police officers. Neither incident produced an increase in black suspects found with arms. Neither incident produced an increase in black gang activity (measured indirectly with drug-related crimes in black neighborhoods). Nor did the increased force come from any top-level down directive to increase police surveillance of black neighborhoods. Nor did the increased force come from black suspects being more violent or more resistant of arrest. The records measure the degree to which the suspects resisted arrest. The rate of “non-cooperative” suspects was the same before and after the shootings.

The odd thing here is that police violence did not go up when one of their number was shot by a white person or a Hispanic person. Presumably, those were written off as individual incidents of no particular relevance to everyday policing. When the same crime was committed by a black suspect, there was an immediate generalized perception that blacks were dangerous. The police began using greater force against blacks simply because they were more afraid of blacks.

Never mind the racism involved. That would be bad enough. This is a sign that the decisions being made by police about whether to tase someone or choke someone or tear gas someone or shoot someone are not being made on any kind of professional criteria. They are being made in the heat of the moment with the emotion of the movement. The question of whether to put a citizen into a chokehold should never be an emotional call.

The new protests against police violence are correct that police have far too much access to lethal force – and that that lethal force is used unwisely. Police in much of the rest of the developed world do not carry weapons; still they are able to produce crime rates that are even lower than ours.

“Defunding” police is a political slogan. “Rethinking” police is close to the intention of “defunding” police. Giving police more mental health training and introducing protocols that de-escalate rather than escalate conflict would be obvious first steps. So is the immediate termination of any officer who uses violence inappropriately. Right now, abusive cops get second, third, fourth and forty-eighth chances. They are kept on our police forces forever. Killer cops should serve long prison sentences just the way killer civilians serve long prison sentences.

A cop who kills should not get a second chance. His victim does not get a second life.