You have quoted this guy before, and you really shouldn't, because it is some incredibly shoddy and misleading work.
A. Encounter Rate wrote:Therefore, blacks are about twice as likely to be searched as whites. Once you do a multiple regression controlling for other factors, like previous record, income, area stopped, et cetera, half of that difference goes away, leaving an unexplained relative risk of 1.5x.
C. Arrest Rates for Minor Crimes wrote:Further, blacks are more likely to live in the cities, where there is a heavy police shadow, and whites in the suburbs or country, where there is a lower one.
Hey, remember when the DoJ concluded that the Ferguson PD had been heavily overpolicing African American communities in order to generate revenue for the city from fines for petty offenses, i.e. essentially monetizing the harassment of minorities, and how they had reached this conclusion using not just the pattern of behavior exhibited by the FPD's policing but by internal correspondence between the city government and the FPD leadership? Do you know what happens if you examine this incredibly fucking racist practice after controlling for area stopped? It stops looking incredibly fucking racist while continuing to actually be incredibly fucking racist.
The fact is that our communities are heavily segregated, and police know this when choosing where to cast their metaphorical (and literal) shadows. By controlling for area, you are controlling for the influence of deliberate choices police have made about which neighborhoods to police and how aggressively to police those neighborhoods - choices which may be racially motivated, either at the level of the department's leadership or the individual officers or both.
"Controlling for potential racism, police found to not be racist." How very insightful.
C. Arrest Rates For Minor Crimes wrote:Likewise, whites are more likely to use low-penalty drugs like hallucinogens, and blacks are more likely to use high-penalty drugs like crack cocaine.
I have no idea why anyone would bring this up as evidence against racism in the justice system. The reason rich white people drugs get you a slap on the wrist while poor black people drugs get you locked away for years is completely 100% deliberate, because the War on Drugs was part of the Southern Strategy, and the architects' explicit purpose was to institutionalize the fuckage of minorities by police.
You'll note that the question Scott Alexander set out to answer at the start of his essay was "does the criminal justice system treat African-Americans fairly?" The correct answer here is, "no, because in this instance the laws themselves have been written with demographics in mind to target them with more severe penalties for comparable offenses." But instead we get that exact same fact presented to us as a mitigating factor? What the fuck?
This also does a lot to hide the true extent of the sentencing disparity, which he discusses later. I won't repeat myself there, just remember that his conclusions on sentencing also understate the problem.
D. Police Shootings wrote:[This entire fucking section]
First off, you cannot make any conclusions comparing the third column to the fourth column. The difference between the fourth column matching the third column and not matching the third column is something like
six data points. The statistical significance on that specific claim is very low, so fuck that noise.
But more importantly than that, we don't have any real data on police shootings. We just fucking don't. Every bit of police shooting data that has ever been released to the public has been released voluntarily. It is never authenticated or audited in anyway, and as a result it is often contradictory and full of errors.
Tamir Rice's shooting, the original subject of this thread, was not reported to the FBI database. Tamir Rice was a freebie - his death does not show up in any statistics on who police are and are not killing. Eric Garner's death was 1) not reported to the FBI database, and 2) not counted in the NYPD's internal statistics. Why? Because he wasn't shot, he was suffocated as a result of the method and position with which he was restrained. Eric Garner was a freebie - his death does not show up in any statistics on who police are and are not killing. Homicides by police often end up misclassified as civilian homicides with no indication that police were involved at all, meaning that while those deaths were technically reported they are also all freebies that don't make the police look bad in anyway. Fun fact, while we're talking about the NYPD, the last time it gave any numbers to the FBI at all (2006), the numbers it gave them inexplicably did not match their own internal records.
Every bit of data we have is voluntarily self-reported with no checks in place whatsoever. If you have embarrassing results, you can seriously just run them through the shredder and nobody cares. If you have embarrassing results but want to look good reporting them, there are paperwork loopholes big enough you could drive a bus full of dead black people through - and at the end it would still look like a genuine mistake (and might even be a genuine mistake). There is every reason to believe that the numbers we do have are next to worthless. They are not a random sampling. They are voluntary self-reports by police departments who may end up facing public scrutiny if the numbers in them make them look bad.
The story does not get any better when you take a more in-depth look at his sources. I have glanced through some - but not all - of them, and invariably what I find is "it looked racist, but then we controlled for a bunch of things which correlate with race and it stopped looking racist."
"Black people looked like they got harsher sentences, but then we realized that black people just happened to live in counties that were tougher on crime in general. Sure, it would be totally racist if a black person got a harsher sentence for being black. But it's totally not racist if a person from a predominantly black county gets a harsher sentence than a person from a predominantly white county. That's not racism, that's just bad luck. Bad luck that seems to happen to black people a hell of a lot of the time, let me tell you what. Woo boy. But totally not racism."