The Police

I'm quite conservative when it comes to law-and-order.  Even as a not-completely-crazy libertarian, I believe that the freedom to make a life for yourself is enhanced if you have a somewhat neutral third party between you and people who'd like to prey upon you.  Policing, within well-defined limits (which do not included domestic military operations!!!!!), is essential.

But it's time to acknowledge that something has gone seriously wrong.

According to the Washington Post's database on police shootings in America, blacks are killed at a rate that's roughly 2.3 times that of whites.  By some other measures, the rate is nearly 5 times that of whites.  No matter how you measure it, something bad is happening.

There is plenty of anecdotal evidence that the deaths are only part of the problem, and that there's a pervasive pattern of harassment and humiliation that swamps the killings.  But we need to be careful here, because this is exactly the sort of situation where human beings' heuristics for drawing conclusions from small amounts of data can get us into trouble.  Unfortunately, there's very little data to be had.  That doesn't appear to be accidental¹, and it's suspicious enough to make us trust our anecdotal priors even more.  I'm going to accept the idea that the police are acting in a manner where at least the result is racist.

Why is this?

My white guy impression is that most police are decent people.  Decent people know that racism is bad, and likely don't self-identify as racists.  And yet, as a group, they're engaging in behaviors that are effectively racist.

Hypotheses for what's going on have to fall into one of two broad categories:

1. There are a substantial number of police who are racist, whether they believe they are or not.
2. Something about police organizations either requires or provides strong incentives for their members to engage in behaviors that have racist effects.

As a practical matter, there's not much to be done if hypothesis #1 is what's driving things.  We're not going to disband police forces.  Fewer things will dispel concerns about racism faster than widespread disorder.  Over the long haul, it'd be nice not to hire racist cops.  That seems like a worthwhile organizational goal, and one for which future cohorts of police should be screened.  But this is no solution for either the short or medium term.

So let's drill down on hypothesis #2, where something about the practice of policing causes racist outcomes.  Let's moot some behaviors that might lead to institutional racism:

1. Police are taught how to be intimidating.  Unless you want almost as many cops as civilians, this is an occupational necessity.  Perhaps, however, the intimidation tool is relied upon too heavily.  I would think that that was a doctrinal thing that could be changed, although it's also a training thing, meaning that most good cops internalized the behavior back in the police academy.  Now that they're older, it'll be hard to retrain.
2. Intimidation is easier if you're aggressive.  If aggression is necessary to succeed, then there's selection pressure in police forces to be aggressive.  Some aggressive people are thoughtful and careful to manage their aggression.  But a lot of them are just assholes.  Not necessarily racist assholes, but assholes nonetheless. As support for this hypothesis, note that spouses of police officers experience domestic violence at roughly four times the national rate.
3. Police are taught to be very conservative when it comes to preserving their own lives and, even more important, the lives and well-being of their partners and fellow officers.  This is just Paramilitary Training 101.  But it leads to "Blue Wall" kinds of solidarity, which aren't necessarily in the public interest.
4. Police are taught not to let people committing crimes get away.  It's hard to deter lawlessness unless its consequences are certain when it's observed.  This is a fine policy until somebody winds up getting shot while fleeing or beaten for resisting arrest.  After that, there's a lot of second-guessing.
5. Police need an SOP to make hard decisions.  The trick to making as few bad decisions as possible is to rely on training and procedures whenever possible.
6. There's a natural tendency to confuse bad areas with bad people.  Residents of high-crime areas lose the benefit of the doubt.
7. Arms races between the cops and the criminals need to be won decisively by the cops.  In a society where guns bounce around like the particles in a gas, having better weapons and tactics is the only way to avoid getting killed.

All of these factors make perfect sense if your goal is to have an "effective" police force.  Perceived effectiveness is always important in government, because ineffectiveness makes the news, and news kills the careers of people it touches in most cases.

And we, the public, want an effective police force.  We want to be able to feel safe.  We don't want to deal with the debilitating effects of criminality on our lives and businesses.  I don't see any solution to the problem of institutional racism that results in more crime being tolerated.

This is a huge problem, because the deterrence of crime is wrapped up in aggressive policing.  You need police to make it difficult for criminals to operate, which requires some degree of harassment.  You need to deter street gangs, which requires harassing some assemblies of people in areas with a gang presence, and a lot of those areas don't have a lot of white people in them.  So if we want crime to stay low, at least some of this institutional behavior has to be tolerated and likely even encouraged.

That said, one way to get low crime with less intimidation and harassment is through surveillance.  The privacy issues associated with pervasive surveillance make a lot of people queasy (including me), but if I have to change the mix between public surveillance, crime deterrence, and fewer racist institutional behaviors, I'm OK sacrificing some privacy for less racism.  I'm not OK sacrificing public safety for less racism.  That seems like an awfully good way to get the public at large to be OK with institutional racism, which is not a step in the right direction.

But if we the public are going to rely more on surveillance and less on intimidation, it makes perfect sense for the police themselves to give up any right to privacy on the job.  From the time they clock in to the time they go off-duty, body cams need to be universal and on.  If the tradeoff is that cops don't get to have personal lives at all on duty, that seems like a small price to pay.

Beyond this, there are some gimmes out there:

1. Training practices need to change to de-emphasize intimidation and aggression.
2. Retraining for experienced cops needs to be a priority.  This isn't going to be hugely effective, but it's hard to hold someone accountable for doing their job the way they were trained to do it without at least giving them a chance to know that the rules have changed.
3. Standard operating procedures need to be heavily revisited, and escalation rules need to have higher thresholds, even if it comes at the expense of officer safety and apprehension statistics.
4. Maybe we need to trade the size of police forces for more professionalism and pay a more technocratic kind of police more.  A lot of the reforms I'm talking about here make the job more dangerous.  If you don't compensate people for that danger, you wind up attracting the wrong kinds of people.
5. If you increase the professionalism of your police force, then your hiring practices have to reflect it.  You need to dial down the aggression criteria substantially.  As I said above, this isn't a solution for anything in the short term, but a commitment to changing the mix sends a signal to both the public and the members of the force.
6. Complaints by the public and domestic violence charges need to be given much more weight in considering promotions and disciplinary action.  I'm reluctant to use the words "zero tolerance" here, because that inevitably leads to complaints becoming a personal weapon wielded by people who just don't like you.  But aggressive, impartial investigation of such complaints needs to be increased, and even a hint of substantiation needs to be career-limiting, if not career-ending.
7. I can't see a way of reducing the esprit de corps of police forces, and the attendant natural impulse to protect fellow officers from discipline, without completely demoralizing your force.  But body cams are a huge deal here.  They are no doubt also demoralizing, but that's why professionalizing the force isn't a terrible idea.  Note to the public:  You need to be willing to pay for this!!!!
8. Getting good intelligence (cf. "surveillance") to street cops is essential if you want to reduce the number of stops and harassment of innocent people.
9. Rules of engagement on escaping suspects need to be revisited.
10. I have no clue how to avoid the continued militarization of the police.  You simply can't let cops go up against heavily armed suspects without enough firepower to counter them.  Other countries have ways of limiting the quantities and lethality of guns in the society, but the United States doesn't.  Absent a repeal of the Second Amendment, it's not happening.  I don't think there's even a decent debate to be had on this topic.  Guns are a fact of life.  It would be nice to put an upper bound on lethality at some point, but then it'll take a couple of generations for the arsenals to decay enough to make a difference.

All of this stuff is hard and almost certainly fraught with unintended consequences.  But we seem to have reached a tipping point where this needs to be a national priority.  When that happens, the key to success is not to retreat when something doesn't work; the institutions need to keep trying.  We members of the public need to reward administrations that fail and fix things more than those that retreat to the status quo ante.  That's not something that the media is set up to allow us to do right now.  Perhaps they ought to be thinking about some reforms as well.

---

¹That this is so heavily politicized is deplorable.  On the right, there's all kinds of dodging and weaving going on to avoid getting officer-involved injuries and deaths into the uniform crime statistics in any meaningful way.  On the left, as I was googling some stuff related to the the ethnicity of gangs, I was surprised to discover that statistics pretty much vanish starting in about 2012.  Apparently, people are only fans of data when it doesn't gore the oxen of their constituents.