You cherry-picked one sentence that supports your position out of an entire paper that repeatedly makes the case (with ample evidence!) that race is a meaningful factor in the rates of various types of crime (mostly violent crime). You also picked the one example that's referring to a non-violent crime (drunk driving), when the GP was specifically highlighting violent crime. The rest of the paper continues to make his point, and it's clearly been examined a large number of ways but a ton of highly-motivated people. The burden of putting forth a better explanation is much larger than "one sentence I picked out of a paper that reviews dozens of papers that contradict my explanation".

I don't think this is an "omitted variable bias" thing. This is a "confirmation bias" thing on your part.

Every paper I have ever seen on the subject has concluded that controlling for SES explains racial variation in crime rates. This paper touched on it in one location but otherwise didn’t bring it up, so that’s the part I mentioned here. Feel free to check around and see what you can find, I’ve got better things to do on a Friday night.

I really don't understand what that's supposed to explain, then. Your explanation doesn't change the overall data: some racial demographics are overrepresented in violent crime statistics. If controlling for socioeconomic status eliminates the racial variation, all that really says is that the demographics that are overrepresented in the violent crime statistics are the same demographics that are overrepresented among lower socioeconomic statuses.

So... There's a correlation between lower socioeconomic status and violent crime? I don't think that's surprising. What it does not do is explain why those demographics are overrepresented in either category.

I also am highly skeptical of "every paper", unless you're specifically going out of your way to only read the ones that confirm your bias.

I think you're missing the point. If, when we control for socioeconomic status, we find that there's little to no difference in crime rates between races, then what that means is that, while yes, we can factually say "black people commit more crimes than white people", we cannot say that it's intrinsic in "being black" to commit more crimes.

That is what "$GROUP is a bunch of violent thugs" implies: that members of that group are intrinsically more likely to be violent and thuggish than other type of people, specifically because of the trait that $GROUP embodies. That statement does not hold true for black people, but I believe it does for police.

It's important to understand the root causes of things. Black people commit more crimes, because black people are overrepresented when it comes to lower socioeconomic status, because of century-spanning systemic racism and the reduced opportunities that brings.

Police tend to be more violent and thuggish because the profession selects for people who want to have (and inevitably abuse) power over others, and because it's systemically set up as an institution to enable that behavior and not hold members accountable for it.

I would hope from those two descriptions it's reasonable to say that one group's failings come from a sympathetic cause that is often out of their control, and the other group's failings are intentional and by design.

> I think you're missing the point. If, when we control for socioeconomic status, we find that there's little to no difference in crime rates between races, then what that means is that, while yes, we can factually say "black people commit more crimes than white people", we cannot say that it's intrinsic in "being black" to commit more crimes.

But we cannot say that it's not intrinsic, either! All we can tell is that there's a correlation between "being black" and "being poor". Like the person I was responding to pointed out: there's the potential for a hidden variable that connects these two things. Alternatively, one causes the other. Given the available evidence, it does not seem clear that the arrow of causality points in either direction.

This is the point that you're missing. You assume that "being poor" causes people to be "more violent", when it's equally plausible that being "more violent" causes people to be "more poor".

And if being "more violent" is something that is intrinsic to certain demographic groups, then the latter interpretation would fully explain the overrepresentation in both lower socioeconomic classes and in rates of violent crime.

I fully agree that there has been systemic racism that contributes to lower socioeconomic status for some racial demographics. However, I refuse to accept that as the only explanation for those same demographics being highly overrepresented in violent crime statistics. This is not a simple problem with a two-dimensional answer. The "systemic racism" argument may be necessary, but it is not sufficient, because it attempts to remove any kind of agency those same demographics have over their own behavior, culture, and identity.

This is both wildly disempowering and enabling. You're essentially excusing every individual in that demographic from the consequences of their violent criminal actions by putting the blame entirely outside their control, which is bullshit.