Earlier quoted context omitted.
"in proportion to the violent crime they commit" This comes across as tone deaf when people are talking about victims who were not committing violent crimes. For example, allegedly passing a counterfeit $20. Furthermore, the violent crime statistics exist in the context of people not being able to trust the police, and to the extent they do, the police not doing anything in response to shootings. So it's easy to see…
This is ridiculous trying to pin the prevalence of violent crime in some communities as the fault of the police. What evidence supports this? Maybe you can say that violent crime is an offshoot of poverty, but a lot of poor communities don't have the same violent crime as others..so that's tenuous at best. The big experiment would be to reduce the poverty of all groups and see what happens to violent crime, maybe the…
First you say it's ridiculous to blame the police.
Then you suggest poverty might be to blame, but you doubt it.
Then you propose the experiment of reducing poverty.
But you don't believe that's the cause! And I don't either. So it seems to me like you're not providing an alternative hypothesis to blaming the police.
I think in order to select evidence, two alternative hypotheses are a pre-requisite. And if you think the one is "ridiculous", then it implies you have another one. You just haven't stated it so I am aware.
Also, your last two sentences really bother me, despite being far from an expert in statistics. Do you think it's reasonable to say that correlation is symmetric, while causality is not? And how in general, do you think you resolve a chicken-and-egg problem where people dispute responsibility? It doesn't seem to me like the answer is simple or purely a matter of statistics.