Yeah self-reporting is the absolutely least effective way to gather data. It is never viewed as reliable in any scientific or statistical context. It is barely better than anecdotal evidence.
I'm not even challenging Ron's statement that racial profiling exists as I'm sure it does. But causation is not correlation. Here's an article from the New York Times explaining some of the reasons for the disparity between AA as a % of the population vs. AA % of incarceration rates:
First, the police are at least in part guided by suspect descriptions. And the descriptions provided by victims already show a large racial gap: Nearly 30 percent of reported offenders were black. So if the police simply stopped suspects at a rate matching these descriptions, African-Americans would be encountering police at a rate close to both the arrest and the killing rates.
Second, the choice of where to police is mostly not up to individual officers. And police officers tend to be most active in poor neighborhoods, and African-Americans disproportionately live in poverty.
In fact, the deeper you look, the more it appears that the race problem revealed by the statistics reflects a larger problem: the structure of our society, our laws and policies.
The war on drugs illustrates this kind of racial bias. African-Americans are only slightly more likely to use drugs than whites. Yet, they are more than twice as likely to be arrested on drug-related charges. One reason is that drug sellers are being targeted more heavily than users. With fewer job options, low-income African-Americans have been disproportionately represented in the ranks of drug sellers. In addition, the drug laws penalize crack cocaine — a drug more likely to be used by African-Americans — far more harshly than powder cocaine.
https://www.nytimes.com/2015/10/18/upshot/police-killings-of-blacks-what-the-data-says.html?mcubz=1
This is overblown just as with most thing people grab on to and run with are.