[lukeschlather]

It's important to note the distinction between "blacks commit 50% of the murders in the US" and "50% of the murder convictions in the US are black people." It's well-documented that 40 years ago a white person could openly kill a black person and they would not be convicted. I've read a lot of studies, reviews of court cases, and court judgements about the subject, and it is clear that prosecutors are far less likely to charge a white person with a crime than a black person.

Prosecution figures are simply unreliable as a metric for judging how many crimes are committed by whites vs. blacks. There are many interesting cases that show this. The most interesting one is McCleskey v. Kemp. It's not my favorite in terms of the specifics of the case, but it is my favorite because in the intervening years it has become pretty clearly a bad ruling, to the point that the justice who wrote the majority opinion wishes he could reverse his own vote. The unfortunate thing being that he cannot, and his vote has made it effectively impossible to detect and address racial bias in prosecution.

https://en.wikipedia.org/wiki/McCleskey_v._Kemp

[Pyxl101]

Other data support these ratios. For example, in surveys of crime victims such as the National Crime Victimization Survey, the race reported by victims of their attacker reflects the proportion of conviction rates:

"The UCR and NVCS show similar trends regarding the races of offenders. Generally, the number of minorities identified as offenders is disproportionately high compared to their overall numbers in the population. As Ronald J Berger et al. advise the NCVS, 'Data are consistent with the UCR. The offenders in these types of crimes are disproportionately young, nonwhite, and male.'" (NVCS page 312, can be found by Google search-inside-the-book)

I would be interested to research the topic further. Could you provide more information about the studies you mentioned? I have not previously seen convincing evidence that bias in the justice system is responsible for the racial disparity in these crime statistics. From what I understand, a lot of the murders are black-on-black violence, as well. In 2013, 90% of blacks were murdered by blacks. A majority of people are killed by one of their own race in general (though black-on-white violence is much higher than white-on-black). People are typically murdered by someone who knew them intimately.

  [Updated per correction]
  Blacks killed by blacks: 90% [1]
  Blacks killed by whites: 8%
  Whites killed by whites: 82%
  Whites killed by blacks: 15%

In what way are these murder statistics going wrong? White people are committing murder and getting away with it? Whites are committing murder, but it's falsely attributed to blacks? There are single examples of all sorts of crazy things happening in individual cases, but I haven't seen evidence or argument supporting that this happens systematically to a degree anywhere near enough to explain the discrepancy. Murder is a very serious thing, and it's hard for me to believe that prosecution bias is responsible for statistics saying that blacks commit ~4-5x as much crime as whites, and 10x as much murder as whites.

I find it plausible that there are effects along the lines of what you're saying, but I find it hard to believe that it would result in a murder per capita discrepancy of 10:1. It seems more plausible that that kind of bias would result in a discrepancy of much smaller proportions, especially for murder. I would be glad to review whatever evidence is available.

[1] https://www.fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2014/...

[flopto]

Those statistics are just false. More reliable numbers are:

  Blacks killed by blacks: 90%
  Blacks killed by whites: 8%
  Whites killed by whites: 82%
  Whites killed by blacks: 15%
  Ratio of whites killed by police to blacks killed by police: 1.857

[1] http://www.politifact.com/truth-o-meter/statements/2015/nov/...

[2] https://www.fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2014/...

[Pyxl101]

You're right, that table was inaccurate. I've corrected it to use your numbers. That's more consistent with the other data I was reviewing, which suggested that most homicide was intra-racial rather than inter-racial.

"For here we are not afraid to follow truth wherever it may lead, nor to tolerate any error so long as reason is left free to combat it." (Thomas Jefferson)

[lukeschlather]

Read "The New Jim Crow" by Michelle Alexander. I'd suggest starting with chapter 2 since the introduction as well as the first chapter are more of a manifesto, the data which thoroughly demonstrate the point are in chapters 2-5.

One particularly damning statistic: black people pulled over and searched for drugs are significantly less likely to have drugs than white people.

http://www.nytimes.com/2015/10/25/us/racial-disparity-traffi...

This is a pretty good digest of the figures, but really read The New Jim Crow:

http://www.naacp.org/pages/criminal-justice-fact-sheet

[dragonwriter]

The UCR aren't independent data that confirm conviction-based statistics, they are based on law-enforcement reporting, and share the same biases as prosecutions do.