Charles Fain Lehman on the Fight Against Crime
He sees the the fight against disorder as paramount
I recommend listening to Charles Fain Lehman's talk on crime and disorder. His talk, recorded at UATX offers a unique theory of how a society can maintain a low crime rate.
Lehman starts out by asking the students if they know what “broken-window policing” means. It turns out that they don’t. He is not surprised, but I am. Perhaps much of my value added from being the classroom would come from being old enough to know such things, and to include them in stories.
Anyway, his theory is that police do not provide most of their value by apprehending criminals. They instead provide value by upholding order. This is a radical theory.
In an orderly society, people obey social norms. In terms of game theory, you have an equilibrium in which most people in most situations cooperate. Defection is rare.
In an urban area, people look around for signs that show whether or not the equilibrium is one of predominant cooperation or widespread defection. Broken windows are a sign of the latter. People who want social order instinctively shy away from neighborhoods with broken windows, leaving those neighborhoods to be taken over by defectors. Neighborhoods with broken windows, or with other signs of social disorder, become high-crime areas.
Lehman argues that police are relatively powerless to deal with crime. First, criminals are hard to catch. Second, they are hard to convict. Third, our bias is to be lenient and to give second chances.
It is not that conventional crime-fighting is futile or should be abandoned. Incarceration is a useful tool for reducing crime.
But Lehman says that much of the time, police are called to deal with disorder. They arrive at a situation in which it is unclear whether a crime has been committed, but people are arguing and disorderly conduct has taken place. It is those situations in which police earn their keep. If they can restore order, the neighborhood calms down. But if the police themselves panic or get involved in the fray, the neighborhood becomes more volatile and tempers become more heated.
I remember the point made by management guru Edwards Deming that the least cost-effective way to improve the quality of manufacturing output is to wait until the end of the process to inspect widgets for defects. Instead, fix the manufacturing process to minimize the number of defective widgets produced. Even better, design your widget so that it is easy to set up a reliable manufacturing process. The earlier you intervene to promote quality, the more effective your intervention will be.
Listening to Lehman, the analogy would be to say that preventing crime requires earlier intervention. If police can help maintain a sense of order, then the area will be dominated by cooperators helping to maintain and enforce social norms. Instead, in a disorderly area, by the time you make an arrest it is too late in the process to make much of a difference.
We think of police as being there to “enforce the law.” This suggests that they should be operating by strict rules.
Lehman argues that strict rules are almost the opposite of what makes for good policing. He says that it requires good judgment in ambiguous situations that require restoring order regardless of whether an actual crime has been clearly committed.
It is important for the public to believe that police are law enforcers and not law breakers. But in the grand scheme of things, what keeps crime low is an orderly society. That is one in which cooperation with social norms becomes the dominant social strategy, and it becomes difficult and unusual to break norms. Formal laws are just a subset of the norms that need to be upheld if you are going to have an orderly society.
In Lehman’s model, small breaches of order are not harmless. People sleeping on the streets or engaging in substance abuse in public are creating a sense of disorder. As I wrote in March about a Scott Alexander essay,
The last time I walked around Manhattan, I found the constant acrid smell of marijuana demoralizing. When I see the big increase in reckless driving since the pandemic, I find highway travel demoralizing. Like Scott, I find people who impose their loud music demoralizing. I find subway fare-jumpers demoralizing.
Lehman is saying that what we should expect police to do more than just apprehend criminals. We should want them to contribute to enforcing what I call the fundamental social rule: reward cooperators and punish defectors.
I can imagine this being carried too far. Michele Gelfand’s concept of tight cultures and loose cultures comes to mind. In America, we probably do not want our culture to be as tight as it is in Japan. But my impression is that we have gone too far in the direction of loose.
substacks referenced above: @



I do wonder whether convictions will be easier in an age of AI/mass surveillance. Perhaps the leniency bias will persist though that is also likely regional specific.
For whatever it’s worth it does sound like El Salvador has seen notable gains in safety (or at least homicide rates) by rounding up and putting away gang members very aggressively.
But yes I do think/seems intuitive that seeing law enforcement usually promotes better behavior (whether on the highway or in a city).
I think MR has also blogged on literature that there are benefits to more certain convictions /shorter sentences rather than unlikely conviction /longer sentences even if the expected value jail time is the same.
Policing serves different purposes in middle class upper middle class and lower class communities.
Most middle class members of society rarely interact with the police so they are unaware how policing really operates.
But their belief that the police operate as they expect them to do help keep societies and communities functioning.