[justinlilly]

I believe there are logical holes in the article posted.

> According to data compiled by The Washington Post, 50 percent of the victims of fatal police shootings were white, while 26 percent were black. The majority of these victims had a gun or "were armed or otherwise threatening the officer with potentially lethal force," according to Mac Donald in a speech at Hillsdale College. > > Some may argue that these statistics are evidence of racist treatment toward blacks, since whites consist of 62 percent of the population and blacks make up 13 percent of the population. But as Mac Donald writes in The Wall Street Journal, 2009 statistics from the Bureau of Justice Statistics reveal that blacks were charged with 62 percent of robberies, 57 percent of murders and 45 percent of assaults in the 75 biggest counties in the country, despite only comprising roughly 15 percent of the population in these counties.

This is saying that while cops shoot more blacks than the population would indicate, we also arrest them at a higher rate. Because they are arrested at a higher rate, we would expect the shooting rate to be high b/c they have a higher-than-population-level-would-indicate interaction with police.

So.. it's unclear as to whether there's selection bias here. Are blacks given disproportionate police contact which increases arrest rate (aka differential exposure)? This would also lead to the same circular logic.

In Portland, Oregon, where I live, there was a report tracking different ethnicities through the criminal justice system called the Relative Rate Index. Article about it: http://www.oregonlive.com/portland/index.ssf/2016/02/blacks_... and source: http://media.oregonlive.com/portland_impact/other/RRI%20Repo...

In there, it says the rate of each stage of the prison pipeline is approximately the same across race. So to explain why 4x the number of black people are in the pipeline than the population would point to the pre-sentencing side of the world.