37 Comments

The US need to reduce the number of crimes. Legalize all drugs, remove all firearm restrictions on non-felons, legalize prostitution, etc.. The police should invest almost of their resources on violent crime and white collar crime that crosses a high monetary threshold.

Expand full comment

I have absolutely no idea about what you are trying to suggest. Given you're a libertarian, I assume this is advocacy for debtor's prison for poor people of all races rather than simple acceptance of the fact that blacks do more crime and thus get arrested more.

Expand full comment

This will probably be viewed as too simplistic, but it seems like policing is overly concerned with getting collars than it is in maintaining social order and public safety.If it were around public safety, for example, there would never be a high speed car chase anywhere, because it is never worth it. It seems like rather than supervising individual cops around how many collars they make, departments should be supervised for overall order in their territory.

Expand full comment

It's like insurance. Economically efficient but unfair to innocent people who happen to be in a high risk category but individually are low risk.

Which is more important is a pretty subjective question. Many people draw the line at applying this unfairness to law and order along class and/or racial lines. This is for reasons of basic fairness, and also because of negative externalities - there are real social cost to unfairness (perceived or otherwise), as we're seeing just now.

Robin thinks efficiency trumps fairness, and this is a difference in values, which he does seem to acknowledge. But it's odd to me that such an obviously intelligent chap, can write a whole article and brush over this whole issue with "it makes sense to me"

One last point. We are inherently group animals. We cannot think ourselves out of that, and so the negative externalities will exist. I do not think Robin's analysis accounts for them. (Please correct if it is included?)

Expand full comment

I think the point about maximizing crime detection and prosecution is fine, but you started by talking about maximizing crime reduction, which is not the same thing:

"The point of my post here, however, is just to point out that we shouldn’t expect proportionate stops, arrests, etc. in the context of policing designed to max crime reduction."

It could be that something close to proportional, or even disproportionate in another direction is better for reducing crime overall. Disproportionately focusing on certain people might catch more crime, but could mean more crime overall.

By the way, the CAPTCHA's are getting cut off to the right here for me.

Expand full comment

Considering how harmful crime is, and how many homicides there are in the U.S, and how much policing we could afford, that seems doubtful. We might be policing badly in a qualitative sense though.

Expand full comment

I said what I meant to say; I wasn't trying to indirectly say something else.

Expand full comment

Also: I think it might be helpful if you wish to communicate to readers to indicate whether this is just something you happened to think about relating to theoretical ideals or if you really meant to suggest it explains the apparent different effort.

I mean it seems to me that if you want to assert the later you need some actual plausible theory of how police end up making these rational resource allocations overall.

I mean isn't the natural assumption that what police do is optimize their own welfare and thus respond to the incentives given by their employer and a desire to do less work not global concern with crime minimization. Nothing I've seen in police dept policies nor in the considerations of police attitudes/reasoning seems to make it plausible that anyone is even sorta doing this kind of optimization much less than it accounts for the observed effect.

OTOH if you merely mean to make an abstract point about the value of equality it seems to me this misses the boat. People in that debate either care about equality as an instrumental psychological value or have some other moral foundation that values things besides efficency.

Expand full comment

That seems correct as a matter of theory as far as it goes but I think the implied relation to the current situation gets it wrong (reporting the relative crime rates is also wrong since that itself is a function of police choices and less relevant to the random stops) . I'm not a fan of almost all writing on the policing data but tthis paper is a good framework and the stanford paper the had 3rd parties blind to race code interactions between police and drives was really convincing that it really is about attitude or expectation or simple pattern matching not some grand policing strategy (disrespect just hurts you here)

First, it seems to me this effect becomes pretty insignificant if you measure it in terms of something like number of people 'harassed/targeted' rather than disproportionate police attention. After all, the police should focus on the crimes that have the fewest plausible suspects and that means that it will be relatively few people who get way more attention.

This brings up the second point. If you start digging into the data and papers it's just not even slightly plausible the effect is the result of the approach you advocate here. Indeed, the broken windows theory (which stated or not) is behind many of the random stops is basically asserting the exact opposite. Namely that you don't respond to the individual effects and investigative costs of the crimes but presume some kind of network effects.

Most importantly even if you were right there is a real serious global effect based on perceived bias such that the *perception* of police bias itself likely impedes crime deterrence far more than non-optimally distributing investigative resources to avoid that perception.

Expand full comment

The time is ripe for criminal justice reform. Is your suggestion to police higher-crime groups more than the status quo?

Expand full comment

Judging from history, creating enmity between groups is a large cost. If an analysis concludes that disproportionate policing makes sense, but doesn't mention that cost, I'm not convinced by it.

Expand full comment

I suspect that had the NYT written the article with data about "men" and " women" instead of "black" and "white", no one would have had an issue with disproportionality.

Expand full comment

Absolutely.

What is missed in focusing on the relative resource allocation is the possibly more important question of aggregate policing level. The open question being, is the US policing too much? I suspect it is, although it’s a hard sell politically; that the bullet we need to bite to reduce miscarriages of justice is tolerating more crime.

Expand full comment

I have no objection to considering such issues, but I don't see how they overturn the above analysis.

Expand full comment