black crime
The black crime correlations are difficult. The data can be interpreted in multiple ways. sometimes math is required in the process and who does the math tends to influence how the results turn out.
For example, imagine that in a population of 100 people, 80 members of group A and 20 of group B. Imagine that in specific time period, A commits 10 robberies and B commits 5 robberies. One could look at that and reasonably come to both these conclusions:
* A committed twice as many robberies as B.
* One fifth the population committed one third of the robberies, and four fifths of the population committed two thirds of the robberies. Adjusted for population, B commits robbery at double the rate as A.
Interracial crimes have similar problems. For example, imaging that in a population of 100 people, 80 of group A and 20 of group B, in a given time period, members of A rape members of B once, and members of B rape members of A 10 times.
Again, you could reasonably come to very different conclusions:
* B rapes A ten times as often as A rapes B
* A outnumbers B four to one. So assuming each rapist chooses a victim purely at random, the odds of a member of A targeting a member of B are one in five, and the odds of a member of B targeting a member of A are four in five. Therefore, with an even distribution we would expect that if both groups commit 10 rapes, 2 of A's rapes would be on B, and 8 of B's rapes would be on A. Therefore the B on A rape is only 25% higher than what you would expect assuming even distribution, not ten times higher.
And the whole matter is complicated further by the fact that the department of justice is notorious for releasing incomplete information. For example, sometimes you'll see something that looks especially damning like, in whatever year, 50,000 black men raped white women and there was not a single recorded instance of a white man raping a black woman. Which seems unlikely. So then you read the fine print and it turns out that there were 200,000 rapes, 50,000 were interracial, but for reporting purposes they only sampled 10 of them, and of the ten they sampled at random, all ten were black on white...and therefore approximately 100% of the 50,000 interracial rapes were black on white.
Because they only sampled 10 out of 50,000.
Stuff like that happens pretty regularly with the statistics. So you have to pay very close attention when looking at them.
Unfortunately, it has been my observation that both sides of the debate tend to ignore and selectively interpret facts that don't support their opinions. For example, blacks committing X times as many interracial crimes as whites doesn't necessarily mean very much. If whites outnumber blacks by X, you might expect them to target whites by X more simply because there are that many more available whites to target. But just because blacks are arrested Y times as often as whites, doesn't necessarily mean that white cops are disproportionately targeting blacks. It's possible that they actually are committing more crime, in which case you would expect more arrests of blacks.