I am going to suppose for this week’s newsletter that Joe Biden will be the next president of the United States, and there will be enough of a repudiation of the current administration’s policies that some functioning governance will follow in 2021. And if any readers are concerned that by saying this and committing it to paper it will jinx this possibility, I say to you, dear reader, that karma is real but the jinx is not. So perhaps more time and attention might be focused on the gap between your good deeds and your good intentions than on the dire consequences of imagination.
I think it is fair to say that the scale, the intensity and how distantly and broadly the police protests were felt this summer following George Floyd’s murder were a real surprise to the establishment. And by establishment I mean the political class—elected officials and the people they listen to. The murder itself was unambiguously evil and the clarity of the crime exposed deep and raw emotions. Sadly though, I don’t think the protests changed many minds in the establishment—the political class.
I have become convinced that the impediment to police reform is deep support for the bad apple theory. In formal terms, the bad apple theory states that “complex systems would be fine if it were not for the erratic behavior of some unreliable people.” The key here is an almost autonomic instinct to protect the system. In this weltanschauung, allowable tweaks to policing are constrained to some effort to limit police recruiting to people who are less likely to be erratic and unreliable and perhaps some modest restraints on allowable behaviors should people prove to be erratic and unreliable despite best efforts.
Interestingly, complex systems science has moved dramatically away from this idea. In that conceptual framework, bad systems poison the people, so the problem is not that a good system cannot hope to always root out every bad actor, but rather that bad systems inevitably lead to bad actors. And, it should be noted, these effects appear most clearly when the system is stressed. If the appeal of the Bad Apples Theory is in its simplicity, then the Bad Barrel Theory should be at least equally appealing because it is at least equally simple—if you store a bunch of apples in a moldy crate, at least some of the apples get moldy.
There is a rabbit hole here that I will avoid going down because where this gets really convoluted is in thinking about the bad, moldy apples and who is at fault. Bad Barrel theorists and Bad Apple theorists are essentially looking at the same data—data showing some officers are bad but many are not—and use it as evidence to support wildly different conclusions. It is a classical problem, what economists refer to as endogeneity, where the bad systems could cause the bad apples, or vice versa, or that they cause each other simultaneously in insidious ways.
This is where the policy discussion has gotten stuck. The political class has a preconceived truth—that this is a Bad Apples problem—and they have yet to be convinced otherwise. It is not convincing to those folks to talk about individual bad police behavior because they have already acknowledged there are a few bad apples. And talking about how many a ‘few’ is does not go very far either because all the evidence that shows there are more than a few bad actors immediately points to the system being broken and they have already taken that option off the table. Stalemate.
I think there is a way out of this paradox though. But to get police reform actually on the table we have to be clear about whether we are talking about reforming policing or reforming the police. And if it is both, which we choose to focus on first.
“The police themselves are another institution that’s fraying”
This is a quote from an excellent article by Emily Badger in last week’s New York Times. Fraying is a particularly great word to describe the present policing problem. Fray as you may know is both a noun and a verb. Fray the noun is a shortened version of affray and long ago meant a brawl, which is where ‘wade into the fray’ comes from. But is also means to sound an alarm. Not surprisingly then it also comes from the Middle English adjective afraid, meaning fearful. But also, from the French effrayer, which is to scare. And finally, as a verb, fray is to rub away, usually through overuse.
So, another way to phrase this idea is to say, “the police themselves are another institution that’s scary, overused and afraid.”
I suspect whether you’re a Bad Apples Theorist or Bad Barrel Theorist or an adherent to some other theory, having it pointed out to you that the police are a fraying institution is probably sufficiently resonant to get you to stop and think about it for a while. The evidence for that fraying goes well beyond the protests and the events that caused them. Perhaps the starkest evidence is the simplest—police departments are having trouble hiring enough officers. And this is not a new problem for law enforcement, this is a years-long problem. Even the staunchest supporters of the status quo have to acknowledge there are problems with policing that need to be solved if the big problem is that people don’t want to be cops in the way they used to. That’s the rub.
What Needs to be Fixed First: Policing or the Police?
So what is the path to fixing these problems? One obvious starting point is to choose whether the problem we seek to solve is about the police or whether it is about policing. I said earlier that no matter how you slice it, police reform has an endogeneity problem. Decisions about who we pick to be police effects what police do and choices about what police do affects who we choose to be police.
But the thing about endogeneity problems is that they often get much smaller if you look at them closely. If two things happen at the same time, a close look often reveals that one problem usually happens first. Even nature has trouble multitasking.
Do we think first about what police do or who the police are? This is a classic chicken and the egg problem. And like a chicken and an egg it is pretty obvious which comes first. Every chicken comes from an egg, but not every egg produces a chicken. So the egg must come first.
Here, policing must come first. We have to think about what it is we want police to do before we think about who we want to be police and hold the enormous power that comes from the gun, the badge and the status. And I think policing is a much easier problem to solve. I should note that throughout this essay, and throughout the debate about reform, ‘policing’ and ‘police’ are used interchangeably. There are advocates who want to ‘reform police’ and others who want to ‘reform policing’. Some want to reimagine police while others to reimagine policing. Only ‘defund police’ picks a side, although the side I think it chooses is counterintuitive, in that it prefers to reform policing because it seeks to do away with police but surely not to do away with security and safety.
Once the choice is made that the way to fix the fraying institution of The Police is to focus on policing reforms, it gets much easier for the political class to get interested. Here, you can apply some neo-classical economic reasoning to the problem, a solution set that is very relevant to that class. There is a lot to be said here, so let me just hit the highlights.
How to Disrupt Traditional Policing
The place to start is to think a little about what policing activities are service, security, social work, and preserving the peace (prevention) and what activities are something else. The something else is what is left for traditional policing to manage.
Service
A big part of what police do and a substantial proportion of calls for service from citizens are service calls. Police tend to get calls and questions from the people when we can’t think of who else to call. These are what I think of as concierge services. In almost every case someone else can answer these questions equally well, and without the need for a gun and a badge.
Security
Police perform a ton of security work. This is everything from responding to burglary calls that are almost always false alarms, to traffic duties, to literal security at parades, protests and events. In almost every case, private security or nonemergency, nonsworn law enforcement could perform these functions with equal alacrity.
Social Work
When we think about social work we often think of crisis intervention with a person who is a potential danger to themselves or others. That is certainly part of the social work model, but the heart of social work is helping to link people to services and supports that can guide them through a difficult moment or difficult years. This is actually something the current policing model does way too little of.
Preserving the Peace (Prevention)
This is another topic that deserves a vastly deeper dive, but it is troubling how much we typically equate ‘force’ with ‘peace’ in policing. Prevention is about reducing risks and leveraging strengths. Police tend to do the opposite, that is, they rarely ameliorate risky conditions. Instead, they allow risky conditions to remain and control them by threats for some period of time. I don’t think that gets said enough—traditional policing does little, if anything, to reduce risks and risk conditions, it merely controls them. There are too many prevention strategies that have better evidence of effect to name and they are better because the problems get solved not suppressed.
Traditional Policing
If all of the above activities are moved out from under the policing umbrella—and I suspect most law enforcement officers would be willing to part with those activities, there is a much narrower set of actions for police to pursue. Now, the trick is that there have to be a sufficient number of police to be available to respond in support of these other activities and to be available to respond quickly to real police emergencies. So the number of police required to do this work of policing would be quite a bit less, but not enormously less. It could be even less again with more reasonable laws around drug enforcement and prostitution. And it could be less again over time, as the negative effects of over-policing is attenuated and the vast gulf between law enforcement and communities which no longer view police as legitimate is reduced. But there is every reason to think that public security and safety would be improved. And this would be due in no small part to a reduction in our reliance on deterrence.
Coda - On Deterrence and Certainty
The general response to this kind of approach from supporters of the status quo is that crime cannot be deterred if policing was arranged in this way. I suspect this is not true and the reason it is not true is that there is a glaring misunderstanding about the relationship between the volume of policing and the amount of crime that is deterred. There is an implicit assumption that a lot of crime is deterred through the presence of a lot of police. And this may be so, but only under the specific condition that more police increases the certainty that a crime will be detected and punished.
In Baltimore in the 2000s it was not unusual for a whole bunch of officers to roll up on a corner where a bunch of young people were hanging out and arrest everybody. Now, some of these folks were certainly doing illegal things, drug dealing, low-level conspiracy and the like. But a lot of folks were just hanging out. The message to them was simple—your actions are unrelated to your likelihood of being arrested. Deterrence works by imparting exactly the opposite message, that if you commit a crime you will be arrested. If, however, the police convince you that you will be arrested whether or not you are committing a crime, the reasonable conclusion is that you might as well commit a crime because it will make no difference to your likelihood of arrest.
Deterrence then has little relationship to the number of police on the street. What matters is only whether someone entertaining the idea of committing a crime is convinced that they will be arrested if they do. This is likely why there have been many studies of the effect of more police on crime and so little evidence of an effect of the expected magnitude. More police increases the chance that some of the additional officers are doing some effective policing, but again, that is a function of what they do, and only indirectly a function of how many of them are doing it.
And that is another reason why focusing on what police do should precede a focus on how many police there are.
Graphic
A rather startling graphic via Bloomberg CityLab. The idea is to look at how police budgets changed in 2020 relative to how the overall city budget changed. The take away is that all cities above the horizontal line had an increase in the police budget, which is most of them. Those in the bottom left quadrant saw a decline in the police budget at the same time there was an overall decline in the city budget. Only LA and Boston had a decline in police budget at the same time the overall city budget was increased.
Musical Interlude
This week we lost Eddie Van Halen who was a big part of my careless youth. Earlier in September, we lost John Prine, who was one of the most beautiful people to walk with us. I suspect you have never heard of John Prine, other than perhaps giggling to Illegal Smile as a teenager. A good friend of mine likes to encourage me to listen to music more broadly with the truism: if you listened to the best music in any genre from harpsicord to yodeling, you’d dig it.
So, I hope you dig this, it is deep social commentary and should probably accompany every copy of Case and Deaton’s Deaths of Despair. Unwed Fathers, give it four minutes…