Percentage of population: 63......13
of total offenses charged : 69......27
Deaths Due to Use of Lethal Force by Law Enforcement: 54......32
Inmates in prison by race: 57.....38 — ssu

If deaths or prison sentences would have no racial bias, then the total offenses charged would be a good indicator in telling how many go to jail or how many are killed by the police.

32% differs from 27% by 5%, which is noticeable, yet 38% (percentage of inmates) differs from 27% by 11%, which is huge. — ssu

A better naive hatchet job of the numbers would at least adjust for per-capita rates by making an odds ratio. The percentage of the population numbers are right there.

63/13 gives you about 5. So the left column numbers should be about 5 times the right column numbers under random assignment of outcomes given that ratio. Instead, the left column numbers are about 2.6 times then 1.7 times then 1.5 times the right column numbers.

Every one of the left column is about under half of what it "should be" under the assumption of random assignment. And the bottom two are the most out of line with that assumption, not the least.

That's the "are there racial disparities" question.

Rereading your post, I saw a different claim. That instead of adjusting for the per capita percentages, "total offences charged" should be conditioned on for calculating whether there is a disparity or not in deaths due to use of lethal force by law enforcement. Outside of the issue of whether the causal chain:

X is charged with an offence -> X is killed by police.

actually makes sense as an explanation here, which is the modelling assumption underlying that conditioning. adjusting for that does make the numbers more in line. Whether that's a numerical coincidence or not remains to be seen; though it's certainly plausible that whatever variables cause the police to charge people with an offence being racially loaded explain some of the effect of the racial disparities in police killings. If you gotta be in contact with an officer to be killed by one, anything that raises officer contact probability raises officer killing you probability; so it could be be a preferential sampling thing (read; racial profiling + police effort + economic variables + other demographic variables).

Conditioning like that doesn't explain this kind of thing though, taken from the paper you referenced:

Further, although force was employed in fewer than 4% of contacts for all racial/ethnic groups in 2008, blacks were nearly three times more likely than whites to experience any use of force during an LE encounter.

Which suggests that the simplified causal chain X is charged with an offence -> X is killed by the police is over simple. I believe it suggests that because the racial disparity from preferential sampling is already conditioned on (the people in question have encountered the police) and yet there's a disparity in the application of lethal force.

X is charged with an offence -> X is subject to lethal force -> X is killed by the police

If race also influences whether X is subject to force in an offence charging encounter, it'll have an effect over and above the preferential sampling effect. But at that point, we really need to start talking about models, rather than comparing data in a naive hatchet job way.

Should also consider whether being charged with an offence is an adequate way of representing an encounter with police that may turn lethal - Breonna Taylor says otherwise. But that's also a question of proportion.

Can you adjust for economic status?

Not from what I looked at for the above.

That would be helpful, though.

Aye.

Relevant report.

Higher poverty among the black population accounts for a meaningful, but relatively modest, portion of the black-white gap in police killing rates. In contrast, higher census
tract poverty fully explained the Latino-white gap, and the police killing rate among Latinos
was lower than expected given their relatively high rates of census tract poverty

Wow. Ask and you shall receive. :up:

Rereading your post, I saw a different claim. That instead of adjusting for the per capita percentages, "total offences charged" should be conditioned on for calculating whether there is a disparity or not in deaths due to use of lethal force by law enforcement. Outside of the issue of whether the causal chain:

X is charged with an offence -> X is killed by police.

actually makes sense as an explanation here, which is the modelling assumption underlying that conditioning. adjusting for that does make the numbers more in line — fdrake

Thank you reading and understanding my point correctly, and as this was done in a simple google search, it surely wasn't meant to be a thorough statistical inquiry. Only that where there is more crime, there are likely more police encounters and likely more excessive use of force and this should be taken into account.

If race also influences whether X is subject to force in an offence charging encounter, it'll have an effect over and above the preferential sampling effect. — fdrake

And of course there also is the question if "total offences charged" has in itself already a bias that makes charges made more likely towards blacks than white, which could be the case. The question that comes up to me is how big role does the war on drugs have to play with this.

Conditioning like that doesn't explain this kind of thing though, taken from the paper you referenced:

Further, although force was employed in fewer than 4% of contacts for all racial/ethnic groups in 2008, blacks were nearly three times more likely than whites to experience any use of force during an LE encounter. — fdrake

What that encounter is might differ, but as I've said there's an obvious difference and there is a statistic that shows it.

But at that point, we really need to start talking about models, rather than comparing data in a naive hatchet job way. — fdrake

Before models, best to understand underlying issues like the impact on war on drugs, as I mentioned already, or how broken communities really go into free fall in the US making a huge divide between the prosperous and poor communities. Poverty goes through racial lines still in the US.

Before models, best to understand underlying issues like the impact on war on drugs, as I mentioned already, or how broken communities really go into free fall in the US making a huge divide between the prosperous and poor communities. Poverty goes through racial lines still in the US. — ssu

You'd enjoy the report I linked to frank above. It's actually adjusting for economic variables in the context of police killings.

@NOS4A2 is sharing Project Veritas videos churning up ersatz agitprop about rigged elections, meanwhile an actual data leak from Trump's 2016 campaign shows how they attempted to deter 3.5 million Black Americans from voting.

You'd enjoy the report I linked to frank above. It's actually adjusting for economic variables in the context of police killings. — fdrake

I'll read that.

Excellent investigative journalism by Channel 4 UK tonight revealing voter profiling by the Trump campaign prior to the 2016 election. Targeting 3.5 million Black voters across the US through profiling aided by Cambridge Analytica and Facebook. Each voter identified was targeted to deter them from voting by the use of false and decisive Facebook posts. Racial voter suppression on a mass scale.

https://www.channel4.com/news/revealed-trump-campaign-strategy-to-deter-millions-of-black-americans-from-voting-in-2016

Oops, Maw beat me to it.

@Maw believes you can suppress a vote with Facebook ads, confusing voter suppression with political campaigning.

meanwhile an actual data leak from Trump's 2016 campaign shows how they attempted to deter 3.5 million Black Americans from voting. — Maw

believes you can suppress a vote with Facebook ads, confusing voter suppression with political campaigning. — NOS4A2

Hey can a third party here point out where I said "suppress" or "suppression"? Can't seem to find it!

Before models, best to understand underlying issues like the impact on war on drugs — ssu

But not the long history of racism in the US, Lord no!

My mistake. I thought “deterring people from voting” meant “voter suppression”. It does mean that in regular parlance but perhaps not in your world.

My mistake. I thought “deterring people from voting” meant “voter suppression”. It does mean that in regular parlance but perhaps not in your world. — NOS4A2

Deter (verb): to discourage or restrain from acting or proceeding:
Example Sentence: Trump campaign strategy to deter millions of Black Americans from voting in 2016

Your mistake indeed.

Trump campaign strategy to deter millions of Black Americans from voting in 2016 — Maw

Apparently it's impossible to influence someone's decision to vote and so political campaigning and buying ads is a waste of time and money. They'd achieve the same results if they just sat at home all day and waited until after the election. :roll:

We have a major problem here in American. 1) Trump will not accept loss. 2) The Republicans will join him in this stance. 3) The Justice Department will back this any way it can. 4) His storm troopers will rush into the streets with their guns to protest. Maybe someone can add more to this grim picture?

We have a major problem here in American. 1) Trump will not accept loss. 2) The Republicans will join him in this stance. 3) The Justice Department will back this any way it can. 4) His storm troopers will rush into the streets with their guns to protest. Maybe someone can add more to this grim picture? — JerseyFlight

Trump will accept loss if A.)it is too great in both popular and electoral votes to be contentious, and B.)he believes it will somehow benefit him(that it's best for him to concede/resign). A Biden pardon in private conversation used as a carrot would do the job.

I do not believe that the spineless Republicans are willing to continue feigning support of Trump if he loses by a wide margin. Rather, I suspect that there are many who would be more than willing to abandon his ship as it's sinking.

My 2 bit(coins) are on resignation (especially if it's a Biden landslide) and quick Pence pardon (after, of course, Pence is also pardoned) ...

:victory: :mask:

The U.S. Military doesn't back him. The U.S. Intelligence & National Security establishment doesn't back him. The FBI doesn't back him. 24 Democratic Governors and, by extension, their State National Guards don't back him. Chief Justice Roberts won't even hear challenges to lower court rulings the way Chief Justice Rehnquist had broke precedent in 2000. And if the election is inconclusive and gets thrown into the U.S. House, the next (117th) Democratic-controlled Congress no doubt will Impeach & Remove him. A shitshow - amid a pandemic & depression which he's overwhelmingly blamed for? Fuck yeah. But, absent the essential elements as well as other mounting legal, political, & diplomatic impediments, any wannabe "coup d'etat" has already failed.

I do not believe that the spineless Republicans are willing to continue feigning support of Trump if he loses by a wide margin. Rather, I suspect that there are many who would be more than willing to abandon his ship as it's sinking. — creativesoul

I hope this analysis is correct. Lots of emotional people ready to pounce, they want a fight so they can express and therefore relieve some of their anger. The story will be that Trump lost because of the most massive voter fraud in history, therefore he didn't really lose, and must stand up against this fraud to save American democracy. This is how narcissists roll.

I can only hope that there are some brilliant apt professional advertising and/or political campaigning experts out there capable of flooding the airwaves with soundbites consisting of nothing more than Trump's own words...

The key is getting all the people who want Trump out to vote.

That's all it would take.

Strong points, hopefully this populist experiment in political anger is about to come to its end, and hopefully it will take the entire Republican party down with it. These have been brutal American years, full of suffering for so many people.

The U.S. Military doesn't back him. — 180 Proof

I think that this is a crucial consideration. If only the retired generals would come out together publicly and condemn his use of military as a means for suppressing freedom of speech amongst other things such as his being a national security risk to the nation itself.

Deter (verb): to discourage or restrain from acting or proceeding:
Example Sentence: Trump campaign strategy to deter millions of Black Americans from voting in 2016

Your mistake indeed.

I guess the civil rights groups who said this was voter suppression made the same mistake. Nonetheless, one cannot deter someone from voting, or suppress a vote, by showing anti-Clinton ads on Facebook.

Nonetheless, one cannot deter someone from voting, or suppress a vote, by showing anti-Clinton ads on Facebook. — NOS4A2

So, why do it then?

You're such a chump.

So, why do it then?

Why campaign on facebook? Ask the winner of the last election.

What I find rather incredible is that you can collect "race" data in the first place and then even share it with third parties. This would be so incredibly illegal in the EU it wouldn't even be contemplated.

All the more reason we should avoid having our data stored in the US or even handled by companies established there. The only protection is to make sure your data is encrypted in transit and in storage.

Nonetheless, one cannot deter someone from voting, or suppress a vote, by showing anti-Clinton ads on Facebook.

You can, a black guy in Milwaukee admits that he was deterred from voting by a fake anti Clinton add in the report.

Now we know where the real fake news was.

Everything about Trump is turning out to be fake, hollow, smoke and mirrors.