Not necessarily. Mind you, I’m not disagreeing that there needs to be way, way more thorough and systematic collection and analysis of data on police shootings. But there is nothing intrinsically more “honest” or “non-sensationalist” about comparing racial disparity in shooting to racial disparity in crime rates than about simply pointing out any such disparity as a separate fact.
Especially when one takes into account the fact that racial disparity in crime rates is affected by racial bias in conviction and sentencing, we can’t just assume that the fact that black males are convicted of crimes at a higher rate than white males automatically explains why police are killing them at a higher rate. Lott’s critique of the ProPublica analysis based on a very selective focus on racial differences in criminal conviction rates, while ignoring other factors that influence those differences, is no more “honest” or “non-sensationalist” than you claim the original article to be.
There is. If some group commits crimes at 10 times the rate of another group, then it is natural, and not surprising, that they would encounter the police 10 times more and get shot by the police 10 times more. Roughly 10 times of course. Plus/minus.
Propublica not mentioning and not adjusting for it is pure dishonesty and sensationalism. Propublica not mentioning that the data set they worked from is utterly statistically invalid, since it is only from 1.2% of self-selected police departments in the country and heavily slanted to urban police departments is even more dishonest.
And everyone repeating the bullshit “21 times” as if it was actually true after these deficiencies are pointed out is just as dishonest.
Given that all we know about how often any group “commits crimes” is how often members of that group are arrested and convicted for crimes, and that racial bias demonstrably influences arrest and conviction rates,it is not legitimate to use arrest and conviction rates unquestioningly as a reliable proxy for genuine group differences in criminality.
Your (and Lott’s) not mentioning and not adjusting for that is pure dishonesty and sensationalism.
Really, Terr, you don’t have a leg to stand on here. I’ve already pointed out that the ProPublica article itself already acknowledged that the data is very defective and consequently any conclusions drawn from it must be very provisional and oversimplified. Your eagerness to argue that it must be solely black people’s own disproportionate lawlessness that’s getting them disproportionately shot by cops is just highlighting the superficiality and distortion of your own approach to the data.
It is not legitimate not to even attempt to adjust for it.
If your data is “very defective” you should not present conclusions from it at all. And I didn’t see you mention the “very defectiveness” of the data in your original post. In fact, you presented it as “from the data as a whole”.
You’re really reaching now. I was simply pointing out that larger trend of racial disparity in comparison to your four isolated incidents (that you mistakenly believed were five because you didn’t notice that one of the shooting victims survived).
I certainly wasn’t making any claim that “the data as a whole” have been thoroughly sifted to produce incontrovertible conclusions. My use of noncommittal terms like “it appears” and “indicates” rather than “it is proved” or “demonstrates” or “establishes” should have tipped you off as to that.
Mind you, I’m not chastising you for posting a few isolated incidents instead of trying to draw statistically sound general conclusions from as much available data as possible, because isolated incidents were what the OP was inviting. But it was definitely reasonable to put them in the context of the broader picture of statistics on racial disparity in police shootings, even if that broader picture is admittedly still very blurry due to the defectiveness of the data.
That’s not all we know about who commits crimes, not even slightly. There are victim surveys, surveys where people anonymously say what, if any, criminal behaviour they engage in, sociological studies of various kinds, and so forth.
You realize that you just agreed with me that you and Lott are trying to make an argument that’s not legitimate, right?
You don’t get to decide that you can look at only the statistically confounding factors that you think support your point while disregarding the ones that undermine it.
If you think it’s mandatory to include racial disparities in arrest and conviction rates in any discussion of racial disparities in police shooting rates, then it’s also mandatory to include racial-bias impacts on racially-disparate arrest and conviction rates in such discussions too.
Very true. I meant that arrest and conviction data are all that constitute the legal record of crime commission, and they are by no means infallible or comprehensive guides to what crimes are actually being committed.
Yes, because there’s roving bands of black cops looking to blow off nuggets in whitey.
I’ve heard this same thing on conservative talk radio and it’s a pointless argument. There is not enough history of systematic, nationwide abuse of economically and historically oppressed white people by police officers that would cloud every white suspect death with suspicion.
First of all, you are not allowed to draw conclusions from a statistically invalid data set.
Second, it is mandatory to include racial disparities in crime rates when discussing racial disparities in shooting rates, because one obviously influences the other. The arrest and conviction rates are an obvious proxy for crime rates (unless you can find a better one). If you’d like to adjust crime rates by “racial bias” feel free to do so (if you can quantify it), but not adjusting for the crime rates at all is dishonest.
Because it seems that every time lately that you hear about one of these police killings that makes no sense, it’s always a black guy on the receiving end. So I was honestly wondering, how often does this happen to whites?
But no, I wasn’t going to go through >2100 posts in the other thread to try to sort out a particular subset of those incidents. If you think that’s a good approach, I’m not stopping you.
It happens all the time. If you expect that someone will scour the Internet for you and give you a hundred incidents, you would have to pay for the work. If you’d like, I will email you my hourly rate.
