The Europile thread contains a quick, precise, and insightful summary of the empirical exercise conducted by Fryer to look for racial bias in police shootings. There are two distinct pools of observations: an arrest pool and a shooting pool. The arrest pool is composed of "a random sample of police-civilian interactions from the Houston police department from arrests codes in which lethal force is more likely to be justified: attempted capital murder of a public safety officer, aggravated assault on a public safety officer, resisting arrest, evading arrest, and interfering in arrest." The shooting pool is a sample of interactions that resulted in the discharge of a firearm by an officer, also in Houston.
Importantly, the latter pool is not a subset of the former, or even a subset of the set of arrests from which the former pool is drawn. Put another way, had the interactions in the shooting pool been resolved without incident, many of them would never have made it into the arrest pool. Think of the Castile traffic stop
: had this resulted in a traffic violation or a warning or nothing at all, it would not have been recorded in arrest data of this kind.
The argument against the datasets is:
1.) The arrest dataset (Set A) is made up of "a random draw of all arrests for the following offenses, from 2000 - 2015: aggravated assault on a peace officer, attempted capital murder of a peace officer, resisting arrest, evading arrest, and interfering in an arrest." Or in other words, incidents that involve justified use of extreme force. But all of those offenses are according to the police, where they have an incentive to inflate charges.
2.) The shooting dataset (Set B) involves any incident when the officer used a taser/gun. The police have an incentive to underreport those incidents against black/hispanics.
3.) The police shooting non-bias result merely shows that blacks are over-represented in set A and under-represented in set B. (This is true and the more accurate description of the result)
Hence the shooting bias result does not reflect anything.
Set A/supersets of A are where they find the racially biased results for use of non-lethal force, stop/frisk, etc.
This argument is saying/implying that there's a systemic coverup of black shootings, in that using a weapon doesn't show up on police reports at all, or made to look like an accident (controlled for). Or basically, start by assuming the police are racist and interpret the results based on that. Pretty dumb when you don't provide evidence. I'm not saying it doesn't or couldn't happen, but that line of attack is tautological and dumb when it's about data and numbers. I don't see the internal logic of the study being troubled here, or maybe I'm not comprehending.
Edit: It is true that Set B is not a subset of Set A.
Edited by accordion - 7/15/16 at 3:26am