Racism is not limited to American society.

https://en.m.wikipedia.org/wiki/Racism_by_country

If I said “there are 100 million cats in America” (made up number; I’m not a cat statistics expert), then “cats exist elsewhere” would not be a very useful reply. File under “true but contextually irrelevant”.

> Racism is deeply embedded in the fabric of American society.

Yes, I'm sure, and racism is for sure "a significant aspect" of police brutality, I don't contest it. But racism is nonetheless merely a stupid pattern of thought, and I oppose the notion that combating it is to apply more racism.

You'll note that the database doesn't analyze racism, it analyses race.

> You'll note that the database doesn't analyze racism, it analyses race.

It's impossible to do any analysis of racial disparity unless the race element is specifically recorded. And, as data collection is as much a political endeavor as it is technical, not every agency or state tracks race.

You can see a not-quite complete collection and comparison of nationwide police data in the Stanford Open Policing Project (disclosure: started by one of my former colleagues) [0], and which states do/don't track driver race. Not all states have agreed to release their data.

I wonder how you propose that a database "analyze racism"? A database contains data. If it's hard as it is to define and get data on race, you think there exists an arbiter and resulting data source for racism?

[0] https://openpolicing.stanford.edu/data/

The Stanford data set suffers from the same problems as the one in discussion: It opens up for correlating 'race' with foul behavior. Of course there may be such a correlation per se, just as there may be a correlation between heavy drinkers/medication abusers and excessive use of force. The problem with isolating the 'race' factor is that it promotes race hate, which is just stupid, when there are factors that are amendable ('race ' is not amendable). There is obviously a correlation between racism and inappropriate police work, but these kinds of data sets don't show that.

> I wonder how you propose that a database "analyze racism"?

I note that we are in agreement: you can't surmise racism from this kind of data, still that's what the presentation implies. To be able to establish the extent of racism one would have to conduct attitude research (psychological profiling) on the officers.

>The problem with isolating the 'race' factor is that it promotes race hate

How is your stance any different from the common and absurd refrain that discussing racism is racist? We can't discuss racism without looking at race.

I'm not familiar with that refrain, and agree it sounds absurd. FWIW I really do believe we need to discuss racism, however not with racist arguments.

I don't think anyone here will disagree with you on that either. But I fail to see the relevance of that in the context of this article.

Yes, data can be misinterpreted and abused. But you seem to throw out the possibility that data can be correctly interpreted to cast doubt on previously held notions. The prevailing belief is that white cops punish non-white suspects. Data frequently shows that this is either not the case, or is significantly more complicated (e.g. similar numbers of non-white cops being involved in incidents).

You're not in agreement. GP is saying that "racism" isn't a quantifiable metric, so you have to deduce it from data that include race. You're saying that it's not "racism" without officer intent, which is entirely semantic; data can still show racial discrimination.

Noting that race is a factor in police brutality is not racist, it is an anti-racist act.