In other words: becoming a police officer increases the risk of being killed on the job.

Since that article is behind a paywall, does it describe where it gets the 36 unarmed black male victims of police shootings in 2015 from? Other sources claim 100+, which would impact whether ~18x is accurate or not.

@JeffLambert: there are a lot of sources quoted, but the author is defending her book called "War on Police." The glaring issue I see here is that using the total black population is insanely disingenuous. A truer picture would require at least an interaction requirement: stop-and-frisk, etc.

Also, not all unarmed black males who are killed by police are killed in shootings; quite a few are killed by beatings, which aren't counted here, while all police officer deaths are counted.

What is the definition of "feloniously killed" in use here? Would this include, e.g., if an officer was struck by an uninvolved vehicle during a traffic stop?

Tongue-in-cheek nitpick: These calculations assume all black males are unarmed - if a large percentage of the black male population is armed then the 18.5x figure should be lower.

@JeffLambert for 2015, it is from this database: washingtonpost.com/graphics/national/police-shootings 94 unarmed persons, of which 38 were black, of which 36 were black male.

@ChrisHayes If it's an accident, no, that's not a felony. The article should say 51 in 2014 (not 52 in 2015). The 2015 data wasn't released yet when the article was written. "Offenders used firearms in 46 of the 51 felonious deaths. These included 32 incidents with handguns, 11 incidents with rifles, and three incidents with shotguns. Four victim officers were killed with vehicles used as weapons, and one was killed with the offender’s personal weapons (hands, fists, feet, etc.)." fbi.gov/news/pressrel/press-releases

Anything about unarmed black people killing police officers so we're not comparing apples to oranges?

@Kevin: ...or, vice versa, about all black men (armed or not) being killed by police officers. (Seriously, though, the quote seems to be implicitly using "unarmed" as a proxy for "probably not justifiably killed in self-defense or in defense of the public". Which is not exactly correct, of course, but may be a reasonable statistical approximation. Presumably there's also an implicit assumption that none of the police officers were justifiably killed in self-defense, either.)

Given the question title, it might be worth using the numbers in the source (36 and 21) to compute that a police officer is 75% more likely to kill an unarmed black male than to be killed by a black male.

@DavidRicherby I edited the question to quote the video properly. The video is by Heather McDonald herself.

Awesome. I've deleted my comments. (@IlmariKaronen might want to do likewise, to save a bunch of flags.)

This doesn't answer all the questions posed by the OP. Nor does it do a proper job of pointing out probable bias in the source material, as if you were to count officers killed by white people against the number of unarmed white people killed by officers you'd arrive at a similarly lopsided figure. The stats don't say anything meaningful about who is more likely to kill whom; they just say that being a police officer is more dangerous than being a member of the general public.

How is there any value in comparing the entire black population of a country to a selected subset of police officers? This is like claiming doctors are more likely to kill patients than Catholics are to eat spinach.

@aroth I actually did some digging about the white comparison, and according to the figures from my answer below, the figure is 171x for white males.

@icc97, pity it was deleted! It was exactly what I would have expected: It must be largely because there are more whites in the population (hence there are also more whites in prison), but it seems to show that police take more care not to kill whites. Curious about the rest of the numbers, maybe you can re-write it in a more question-centered way?

To be clear I'm guessing the numbers will show that both reasons contribute (more whites, more care taken by police), but this must be shown, not asserted. Since the video cited in the question is titled "Are the police racist?", there must be room to discuss (carefully) whether the 18:1 number actually answers the question.

@alexis Can I re-write my answer and get it re-submitted? I've never had a post deleted before so I don't really understand what options I have

I've edited my answer to remove all the useless rambling, happy to change it in any other way if you see fit

@Oddthinking From your comment "Let's stick to the numbers", I've done my best to remove any pointless numbers from my answer, is there anything else I can do to improve it?

@Oddthinking thank you for undeleting it. I'll be more careful next time :)

What I mean is, you don't make explicit at the moment why the comparative numbers you post are relevant. (Maybe it's clear to readers who already share your beliefs, maybe it's clear after reading the other answer, but an explicit and self-contained answer is always better. Good luck with it!

@alexis it's based off @DavePhD's answer, and just replaces white with black numbers. So assuming that DavePhD's answer confirms the 18:1 ratio, then my answer should give a reasonable comparison