This next installment of the "White Privilege" series deals with the police and the prisons!
Cops shooting blacks
The point of this post is not to necessarily endorse how the cops or the courts work in a general sense. I am merely addressing the question of racial bias. The cops and courts may be unfair and capricious in a general sense, but the point of this post is to address the claim of racial bias.
Cops hesitate longer with a black suspect than with a white or Hispanic suspect:
Hispanic suspects have the shortest delay, which suggests that cops are most comfortable shooting Hispanics, then whites, and are least comfortable shooting blacks.
The year for this data is 2009.
If the years 2003-2009 are aggregated, blacks make up 31.8% of arrest-related deaths.
2009 is the most recent year I could find data for on arrest-related deaths, and while blacks made up 29.9% of arrest-related deaths, they made up 38.9% of violent crimes that year. In fact, it looks like blacks get shot by cops less than their level of violent crime would predict.
As you can see in the above chart, blacks also made up 37.8% of cop killers in 2009. And if the years 2004-2013 are all aggregated, blacks made up 43% of cop-killers.
So far it looks like cops actually go easier on blacks. But perhaps this is just another layer of bias against blacks in that cops go out of their way to target blacks for investigations of violent crimes.
This is implausible, though, because the National Crime Victimization Survey (NCVS) and the Uniform Crime Report (UCR) from 2000-2008 both basically say the same thing:
For a full breakdown of the UCR and NCVS numbers, an otherwise dumb guy did a great job of it.
If the cops are targeting blacks out of proportion to the crimes they actually commit, the victims of crime are in on the conspiracy as well.
One counter-argument to this could be the inaccuracy of eyewitness testimony in identifying individual criminals. But in dealing with claims of racial disparity, all that matters is that the victim correctly identifies the race of the attacker.
And I have to be a bit tasteless about it: blacks and whites have a very visible difference in skin color and a difference in smell. (Apparently white people have a distinctive smell; I never knew this until I was around 20. I also never thought I had an accent.) So it's less plausible that the victim got the race of the attacker wrong than that they got the individual wrong.
Perhaps it's not police bias but court bias. One piece of evidence for this is that blacks make up a higher proportion of the prison population than their total crime rate (blacks made up 40.1% of all prison inmates, but only 28.3% of all arrests).
Whites + Hispanics make up 68.9% of all arrests, but only 54.18% of the prison population.
There isn't data distinguishing whites and Hispanics on arrests, but there is for prison population. Whites made up 33.1% of the US prison population in 2009; Hispanics made up 21.1% of the prison population.
However, blacks serve longer sentences. And in fact, when you control for sentence length, the racial disparity in prison population goes away almost completely.
What this means is that the racial disparity in prison population is almost entirely due to sentence length. The residual may be due to blacks being arrested for more serious things on average - that's something that could be controlled for, but I can't be arsed. Once you control for sentence length, the prison population is what you would expect from arrest rates, which probably reflects real criminal activity as stated above.
This table puts it in a bit of perspective:
The percent ever gone to prison is estimated by taking the total prison population and controlling for length of sentence by race, which can be found in the next section.
Moreover, declination rates are the same for whites and non-whites:
These are the percentage of cases where the prosecution declines, or drops, the case and the defendant goes home free.
So the cops arrest people in about the same racial proportion as the victims of crime claim, and the courts convict people in the same racial proportion as the cops arrest them.
Perhaps juries of all blacks are more likely to convict a white and vice-versa, but on the whole it appears to even out between the races.
These two pieces of data strongly suggest that there is no substantial variation in conviction rates between the races. It would be nice to just have conviction rates by race, but alas, I cannot find that data if it exists.
Drug Law Application
The "White Privilege" claim that blacks are imprisoned for drug use more than whites comes from surveys. But people lie on surveys, and blacks could lie more.
This alone is probably enough to show that illegal drug arrests are based on actual illegal drug use. Even if it doesn't represent the number of members of each race who ever used drugs, the arrests probably reflect the overall volume of total drug use.
There's also evidence that blacks lie on surveys.
According to one study on this, blacks are roughly 25 times as likely as whites to lie on surveys of drug use when a hair and urine analysis are done after the survey. Hispanics are 2.5 times as likely as whites to lie. There are some more studies behind paywalls.
The residual 1% difference between ER visits and drug arrests could be due to the fact that when police arrest someone for committing a violent or property crime, that person is searched for illegal drugs. And since blacks commit more violent crimes and proprty crimes, they are more likely to be searched as a result.
For the record I personally don't support current drug laws. I support harsher sentences for dealers of "hard drugs" and rehabilitation for those hooked on them who should be viewed as victims. The point here is that they don't seem to be enforced in a racially biased way.
Prison Sentence Length
There seems to be one point where the “White Privilege” claim might hold, and that is in prison sentence length.
The paper “Racial, Ethnic and Gender Disparities in Sentencing: Evidence from the US Federal Courts” looked at sentencing by race and crime.
It found that 72.9% of all cases follow standard sentencing guidelines. When strictly following sentencing guidelines, blacks got 2.43 months more on average than whites for the same crime when prior criminal background was controlled for, and Hispanics got 0.71 months less than whites.
If cases where the judge deviated from sentencing guidelines are included, blacks got 5.5 months more than whites, and Hispanics got 4.7 months more than whites.
Because blacks on average serve 64.09 months for their offense, judge discretion increases their sentence by 4.8%. Hispanics on average serve 54.12 months, and for them judge discretion adds 9.5% more time. The average white prison term is 32.06 months, and the effect of judge sentencing discretion is used as a baseline to which other races are compared.
This is just one analysis; the literature on prison sentencing length is extremely muddled and I don't think I can come to any knock-down conclusion on it. Much of the literature doesn't control for prior criminal record, which is useless. If you see some infographic saying blacks get 50% longer jail times, that's almost certainly not controlling for criminal record. The real number is more like 5-10%.
What you find in the research that controls for prior criminal record is either no effect, or harsher sentencing for blacks and much harsher sentencing for Hispanics. It almost never shows whites getting harsher sentences, so all told blacks and Hispanics get perhaps 5-10% longer sentences when prior criminal record is controlled for.
But all told, the big story is that cops are probably less likely to shoot black criminals, who we are pretty sure are in fact criminals thanks to the NCVS. Cops are a bit trigger-happy with Hispanics though.
All races are just as likely to have their case dismissed or be found not guilty. If found guilty, blacks appear to get roughly 5% longer sentences when prior criminal record and nature of the offense is controlled for, and Hispanics get around 10% longer sentences.
The “White Privilege” ideas regarding cops and courts are thus mostly wrong.
UPDATE: Found a study on an otherwise dumb guy's blog that said that when IQ is controlled for, race differences in prison sentences go away.
This is evidence that differences in courtroom behavior may be part of the reason for racial disparity in sentencing, as higher IQ defendants probably behave better in court.