[john_moscow]

>Victims of fatal shootings by police were majority white, but disproportionately black.

That depends on what you take as a baseline. I would assume that police shootings mostly happen in response to violent behavior. I couldn't find any data on the violent behavior, however there is an open FBI dataset on homicide offenders by race [0].

If you assume a correlation between the number of homicide offenders and the general probability of violent behavior, it could very much explain the disproportion you pointed out.

[0] https://ucr.fbi.gov/crime-in-the-u.s/2016/crime-in-the-u.s.-...

[jeffdavis]

The table you linked is a little confusing because it's focused on the murders, but the offenders are more relevant to your point. It also excludes a lot of murders where information about the offender is missing, so the overall numbers seem too low.

The following table is more relevant:

https://ucr.fbi.gov/crime-in-the-u.s/2016/crime-in-the-u.s.-...

And here's the one from 2018:

https://ucr.fbi.gov/crime-in-the-u.s/2018/crime-in-the-u.s.-...

So your point stands: when you talk about "disproportionate", you need a baseline. If you choose the proportion of the general population as a baseline ratio, it looks like police violence disproportionately hurts blacks. But if you choose the proportion of known murderers as the baseline ratio, it looks like police violence disproportionately hurts whites.

Note that there is a subtlety here: the baseline ratio is simply a ratio to use as a comparison point. I am in no way implying that all people killed by police are known murderers. Nor am I saying the ratio of known murderers necessarily extrapolate to the number of justifiable-use-of-force incidents.

EDIT: I am really just saying the baseline expectation matters, and that the general population is not the only meaningful baseline. I am not saying which baseline is the "right" one. That depends on which specific policy you are considering.