Hacker News new | past | comments | ask | show | jobs | submit login
[flagged] Research database explores racial bias in police shootings (2019) (msu.edu)
40 points by tomohawk 37 days ago | hide | past | favorite | 51 comments

This study mirrors others, namely, police shooting are rare and don't show racial bias in the aggregate, and usually are 'justified' in some form (i.e. aren't a very clear case of wrong doing like the George Floyd case). It isn't great for police reform policy, because you don't have metrics to judge success or aspire to.

Having said that, this is an emotional issue and does need some solution. There was a recent article about a black Google VP who was hiding in her office when there was some sort of active-shooter threat on the YouTube campus. She was as much afraid of being shot by this threat, as she was of mistakenly being shot by a police SWAT team... that's a real fear in a person of privilege (highly educated and positioned high in a Fortune 500 corporate hierarchy) and that's not good. There is a lot of trust that needs to be regained that all the calls to 'defund the police' miss. Maybe we need more funding (not less), for things like training in community interactions and deescalation. Why not have police officer spend 1 day out of every month, doing some sort of simulation training on deescalation and community outreach and .. heck .. 'customer service' - it's going to be expensive, but social strife is expensive too.

Police killings in the US are around 40 times higher per capita than in the rest of the developed world. Murder Rate is also around 5 times higher than in the rest of the developed world.

Clearly this are problems that needs to be addressed.

My opinion is, that the first step to solve this issue are good safety nets for everyone.

Sentence lengths need to be reduced and victimless crimes should not be punished.

The prison system must be rebuild from the ground up to reintegrate offenders to reduce the risk of re-offending.

Police training needs to be improved as well. It can't be that two sober cops end up shooting a drunk that is sobering up in his car in the back because they are unable to handle such a ordinary situation.

You have subcultures that idolize gang violence combined with gun ownership being common. You don't get that combination in most of the developed world, or not in the numbers that you do in the US.

I'm also curious about metrics on racial bias of police. From what I've seen (in other studies, I've not yet looked at this one) shootings don't look immediately biased. However I often hear how biased police are in their attention towards minorities. But I'm unsure how to quantify that, or what metrics we can use to research that.

Is selling or buying drugs a "victimless crime" when people's lives, families and communities are destroyed by said drugs? What exactly is a "victimless crime" then?

Your characterization of the recent killing of Mr. Brooks is incomplete. Brooks fought against the police and fired a taser (a deadly weapon according to the Atlanta DA) at the officers giving him chase. They were absolutely justified in shooting him.

The problem we have in America is that many people think it's ok to fight the police. Almost all publicized police killings in recent years are a result of suspects resisting arrest in a violent manner, escalating things to the point where force is required to control them. Of course this doesn't fit the narrative, so it's ignored.

So the taser is deadly? Why was it pulled in the first place? How many times can a taser be fired? Brooks is running away, are you seriously suggesting that he posed a serious thread to the officers?

Nevertheless, the incident shows at best how grossly incompetent the police in the US are,however manslaughter is certainly also a reasonable interpretation.

Alcohol is the most dangerous drug. Are merchants responsible if they sell alcohol?

He was out of his mind enough to fight with two police officers, steal a taser and shoot it at them. Whether or not he was a threat to the officers in that exact moment after shooting the taser he clearly could have posed a threat to the community.

So yes, he did pose a serious threat and had to be stopped. He brought his death upon himself through his own actions. My feelings on this are completely independent of his race or other superficial characteristics.

You know, the US is the only developed country where police officers that are called on an unarmed drunk person, end up killing him. You can find excuses for nearly every instance all day, but in the end you can't expect model behaviour from a drunk person. You can however expect two well trained sober offers to handle something like this without an escalation to deadly force.

I mean 40 times higher than the developed world. Do you agree, that this is an unacceptable level of violence?

>So the taser is deadly?

It can be, depending on how it is used.

> How many times can a taser be fired?

This depends on the model but ones which can fire multiple shots are issued to police.

>Brooks is running away, are you seriously suggesting that he posed a serious thread to the officers?

Yes, because he did.

>Police killings in the US are around 40 times higher per capita than in the rest of the developed world. Murder Rate is also around 5 times higher than in the rest of the developed world.

Police killed in the line of duty is higher than anywhere else as well. The per capita murder rate may only be 5 times higher, but the variance is high too. If you sort this table[1] by murder rate (crime rates per 100,000 of top 100 urban center in US), the number are quite stark. The top 4 cities in that list, are in the global top 50 of cities with highest murder rates (St.Louis is actually in the top 10).

>My opinion is, that the first step to solve this issue are good safety nets for everyone.

Social spending is through the roof already. Federal budget alone allocates the vast majority to healthcare, welfare and social security (and defense). Then you have state and municipal budget putting in billions more. It's not enough to say, 'spend more'. I don't subscribe to the notion that crime is a result of lack of social spending. Hong Kong has immense income inequality, and lots of poverty, but is a very safe city. There are obviously other factors as well.

>Sentence lengths need to be reduced and victimless crimes should not be punished.

There is a common trope that prisons are filled with either innocent people or non-violent drug offenders. That's not true, at all. There is no meaningful prison reform (if your goal is to reduce incarceration rates) without also reducing sentences for or releasing of violent offenders (who comprise the vast majority of prisoners).

People tend to say they want shorter prison sentences, and then you hear about some horrific rape or murder and there are calls for jailing the individual for life. Harvey Weinstein was sentenced to 23 years in prison for third-degree rape. In most countries, he would get probation, at most. In most developed countries, life sentences are limited to 25 years with parole typical after 15 years (and no stacking of sentences). Are you sure the American public would be OK with that?

>Police training needs to be improved as well.

Agreed. It's going to be expensive, but it's worth the cost.

>It can't be that two sober cops end up shooting a drunk that is sobering up in his car in the back because they are unable to handle such a ordinary situation.

That was a not a clear-cut circumstance. I'm not 100% sure police in any other country would have acted differently in that situation. I fully agree about not shooting an unarmed individual running away. But this situation had an individual who assaulted police officers and stole their weapon and fired it at them. That complicates things for you immensely. This is also where police officer should get the benefit of the doubt because they need to make a risk-analysis to the public and to themselves in the span on 1 or 2 seconds. With no weapon involved, we should expect them to err on preserving a life. With a weapon involved, it's different.



Police with bad training kill more and are killed more, not really a surprise there. A better trained police would reduce the number of violent encounters and be less likely to be killed in a violent encounter.

I mean it is factor 40. This are third world numbers. I don' really think there is a valid excuse for that.

It may be that the majority of prisoners are there for violent crimes. But how many of them were in prison for a non-violent offence the first time?

The percentage of innocent people in prison is probably below 10%. But the number of exonerations of the innocent project and the difficulties and issues they uncover, make it hard to believe, that it is in the lower single digit range. Most of the time it is just pure luck that a false conviction is overturned. So there must be many more.

Today we know just how unreliable eye-witness testimony is, especially of a different ethnicity, yet it has been a major way criminals are identified.

There is definitely room for improvement (de-escalation) and less aggressive approach towards people who are not being aggressive (but may be playing dumb and other irksome attitudes). And also better understanding of “unstable” subjects —“oftentimes” the police shoot at provocation that doesn’t pose immediate harm [if someone waves a knife, back up 20 feet].

Also sentencing reforms are needed as well as more effort into rehabilitation and preparation to become contributing members of society for non violent felons.

On the other hand the issue is way over politicized and then you have opportunists who use this as an excuse to agitate and cause rioting and looting (as opposed to political protesting).

Moreover there is a coöpting of the movement by totally unaffected people (non-black non-poor whites) and its to the point in LA inner city “gangs” run off the middle class protesters/agitators because they don’t find affinity in them.

There are people being fired for doing studies like this, I think there was a case in the UK. A review of the decision process that led to these cases is also sensible, otherwise any discourse is quite difficult.

I would caution with attaching privilege to anyone. We have seen Harvard graduates accusing people of profiting from privilege while having day-to-day wage-jobs. If the concept is too complicated for "elite"-graduates, it probably doesn't help in a general discussion. Especially not if the topic is emotional as you said.

I think more funding for better training is indeed a good idea. That police officers are accused of murder by shooting someone resisting arrest and shooting back is also a catastrophe. We also have seen people cooperating being shot. These should be handled differently of course.

> It isn't great for police reform policy, because you don't have metrics to judge success or aspire to.

I feel that the idea of a racial bias behind police killings in the US is misguided. The issue exists, but as this study seems to prove, it's not an issue of police killing too many blacks- the issue is of police killing too easily. This simply doesn't happen in other countries.

If you approach the problem as one of racial bias, while still expecting police officers to escalate confrontations and meet any potential threat with lethal force- which is the real issue IMO- then the problem is never going to be solved.

> as she was of mistakenly being shot by a police SWAT team

But the study suggests that this fear, while real, is misguided: she's getting this from sensationalized news reporting and hyperbolic movies/television rather than anything based in reality.

It may be a fear that is not statistically supported, but claiming it is not "based on reality" is, well, not based in reality. This does happen:


That's also a completely different circumstance to what she found herself in.

It's quite similar. Don't be black near a police response, it could be fatal.

Right, because humans in fear always carefully consider all aspects of a situation under time pressure, with perfect recall of details they read in the news. I'm sure you'll appreciate people splitting hairs to critique your rationality and recall under pressure next time you react to something.

That's a lot of words to concede that it wasn't based on reality, as the other poster said. :)

I conceded no such thing.

You tried to defend the irrationality of someone experiencing fear.

Is assuming you're in a vacuum based on reality for someone on Earth who is experiencing symptoms of hypoxia like astronauts would?

On the other hand, it may be the case that steps such as those this Google VP was taking are what is reducing the likelihood of a bad outcome.

She was. The reporting had built up an inaccurate view of reality in her mind.

you cite "very clear case of wrong doing like the George Floyd case," but the only reason that is clear is because of civilian video recording.

Here is how the police described it in the official MPD report: "Two officers arrived and located the suspect, a male believed to be in his 40s, in his car. He was ordered to step from his car. After he got out, he physically resisted officers. Officers were able to get the suspect into handcuffs and noted he appeared to be suffering medical distress. Officers called for an ambulance. He was transported to Hennepin County Medical Center by ambulance where he died a short time later."

Transparency should be an important step, often time the body cameras were not on or the videos were lost. There also seems a lot of cops covering their mistakes so more transparency and fairness could help with trust.

> She was as much afraid of being shot by this threat, as she was of mistakenly being shot by a police SWAT team...

When we underwent active shooter training at work, part of it was dedicated to how to behave if/when there is a police response. Fear is unfortunate but being acutely aware of your actions in a situation like that is down right responsible.

Even if the studies show no clear bias (the article mentions some criticism and a forthcoming reassessment of the data based on that) the problem clearly extends all the way down to the most minor of police interactions where it becomes harder and harder to explain away.

I think theres a certain set of people who hear "no statistical proof of bias, and emotional issues arent real issues"... so while I agree with you, I just want to emphasize the huge space of systematic problems that cause the emotional issue.

How the police have responded to the protests is by itself indicative of a problem. Responding to accusations of police brutality with televised beatings is so tone deaf that it must speak to an institutional issue. If there were no evidence of police brutality, except for video taken in the past few weeks, It would still be enough evidence for reform.

I don’t think this behavior is particular to US policing [French police vs Yellow vests, Greek police vs anti austerity marxists, Korean police vs protesters, etc]. Yes there is room for improvement but also this result is predictable in this kind of antagonistic interaction. It’s like treating disease with radiology, you cannot avoid some of the accompanying risks . They may be minimized but not eliminated. It’s like teenagers and parents. It will result in rebellion, it’s part of that relationship.

Sam Harris did an excellent podcast (in my opinion) touching on the topics you mentioned.


or text


I especially wonder what qualifies as "justified". The statistics can be only as good as the source data.

It's telling how many "justified" shooting have had video evidence come out afterwards that directly contradict the police report. The question isn't whether the police lie and plant evidence, it's how often do the police lie and plant evidence.

Particularly when dead men tell no tales.

The critical response to this research, which they helpfully link, seems to imply the authors (and arguably others) using racial shootings data to conclude that PoC are not in fact shot at a higher rate - have fallen for a base rate fallacy or selection/sample bias, where the base rate of upstream neutral encounters between all people and police is not captured in the data. Without that frame, there isn't enough information to infer causality in the selected data.

The linked critical letter: https://www.pnas.org/content/117/3/1261

That's my layman reading of this, as the language is a bit opaque. If that is their assertion, this kind of error seems like such a rookie mistake I would sympathize with someone who thought that neglect was ideological or misrepresented. Personally, I think in politics and policy, data is used for support not illumination. The only way data ever causes anyone to change their mind is if it comes from a poll. That's my skepticism, but a more capable reading of the criticism would be useful.

1. The study assumes that the statistics provided to them by police departments is accurate. That is to say, they're assuming that when an officer shoots an unarmed man without provocation, the officer will report it accurately.

2. This study focuses on police shootings, not use of force. So choking a civilian to death wouldn't count.

3. It also doesn't measure how often the officer involved escalated the situation.

Why do people always look at "shootings". The black experience with police extends to all interactions from fining, arresting, charging, overcharging, railroading all the way up to killings. That's like having a quality metric that ONLY considers times when production had to be shutdown, and never when people couldn't login, when transactions were lost, and when people got someone else's shopping cart's contents in theirs.

Unfortunately leadership want to grab into the most salient examples and purposely use raw language like “police brutality”. Not denying there is police brutality because there is, but pointing out the leadership themselves are pushing this angle, so how do you expect people to look at the finer points?

But I get it, it’s not sexy, it doesn’t draw headlines, it doesn’t draw middle class protesters.

Shootings are both extreme enough and infrequent enough (though still all too frequent) that one could plausibly assemble a nearly complete national dataset, plus they are more likely to get covered in the news/on social media. I’m not familiar with the literature, but I’d bet there are dozens of high quality (though more limited) studies on the racial disparities in the interactions you mentioned for each study on Police killings.

Yea I just mentioned that in another post, I'd love to see studies like this that try to analyze police and biases towards minorities.

Are you aware of any?

This appears to be convincing argument and research. However, someone here(?) pointed out recently that most people are unconvinced by facts that leave them emotionally unsatisfied. Information like this likely won't help when people are upset already.

That this article was flagged, QED

The press could provide more context and balance to reporting rather than the Robo-cop like apocalyptic narrative that gets people riled up.

The police haven't been doing themselves any favors by actively attacking members of the press though. If you don't want a narrative about police brutality to spread, perhaps you shouldn't beat the people responsible for generating the narrative while they scream that they're 'Press'.

But that’s exactly the point. The majority of reporters don’t get beat up, but it’s presented as a routine thing to happen when more often reporters get beat up by protesters —but even that is a minority of interactions.

I have been somewhat disappointed by the number of people not being willing to even look at the data when it comes to these racial and police issues. It's certainly an emotional issue where the data can be inaccurate or misleading, but that doesn't mean it can be completely ignored. There are big changes that need to be made and the efforts will be much more productive if they're backed by research and not pure reactionary emotion.

it seems more likely that people who are invested in not changing anything would look at data uncritically (what is the source of the data? police? do they get to say what is justified use of force?) and conclude nothing needs to change.

i suspect people who feel their life is on the line are considering seriously all methods for making change, including reviewing all available research

note that the article mentions criticism of their work, includes a response, and a note about reassessing certain aspects.

People in this thread who are engaged in rapidly shifting the goal posts to fill the vacuum left by the race-based narrative, please stop and consider why you and many in the US were so incensed by widespread beliefs that were easily disproved by clear and available data. Consider what else you believe on this topic that may also be wrong. This is a teachable moment.

Similarly, people in this thread who so readily provide their own interpretations of this data as supporting their existing beliefs: please stop and consider that data in itself doesn't prove anything. Interpretation is what matters. You'll see plenty of excellent questions both elsewhere in this thread and even posted as a response to the research. The point is not that the data itself is wrong - though it may well be, given the sources - but that the interpretation and conclusions may be far less straightforward than they first appear.

Consider what else you may be ignoring as you seize an answer that feels concrete and rational, and just happens to agree with your existing political stance. This is, indeed, a teachable moment.

Police in the USA literally toss hand grenades into the cribs of sleeping babies. It is not a question of race, white or black they will come for you and they will injure or kill your children.



It's astounding how many problems are immediately apparent in this article.

> Reports of racially motivated, fatal shootings by police officers have garnered extensive public attention and sparked activism across the nation. New research from Michigan State University and University of Maryland reveals findings that flip many of these reports on their heads – white police officers are not more likely to have shot minority citizens than non-white officers.

> “Until now, there’s never been a systematic, nationwide study to determine the characteristics of police involved in fatal officer-involved shootings,” said Joseph Cesario, co-author and professor of psychology at MSU. “There are so many examples of people saying that when black citizens are shot by police, it’s white officers shooting them. In fact, our findings show no support that black citizens are more likely to be shot by white officers."

First of all, none of this research purports to "flip [any] of these reports on their heads". The preceding disclaimer makes that explicit.

Second of all, the charge of racial bias in policing (and racial bias in police violence as a consequence) is not and has never been about the race of the police officer. So far as I'm aware, the race of the police officers is a focus mostly of white people, sometimes from a misguided assumption that black police won't produce similar results, but usually from a perspective intended to discredit the charge of racial bias. This article beginning from that assumption is a really bad sign.

The charge of racial bias in policing is institutional, structural, and historical. Policing in the US was created to reinforce racist social structures. People who are nominally targets of racism are capable of participating in and reinforcing the racist institutions and social structures which would target them, and there's a long history of that. The fact that people of color as cops reproduce the same outcome in no way discredits the claim that policing is racially biased.

> “Many people ask whether black or white citizens are more likely to be shot and why. We found that violent crime rates are the driving force behind fatal shootings,” Cesario said. “Our data show that the rate of crime by each racial group correlates with the likelihood of citizens from that racial group being shot. If you live in a county that has a lot of white people committing crimes, white people are more likely to be shot. If you live in a county that has a lot of black people committing crimes, black people are more likely to be shot. It is the best predictor we have of fatal police shootings.”

This cannot be supported by the data sources they attribute. The police are the subject of scrutiny and can't be relied on to accurately represent the "rate of crime" of their victims. Nearly 90% of entries in the Washington Post's database has no body camera data. The Guardian's database doesn't provide filtering by accused criminal activity, but a quick glance at a few entries shows descriptions straight out of police reports. The "rate of crime" claim is directly attributable to police, and to levels of policing.

This is obviously unreliable when scrutinizing the police. It bears mentioning, since the murder of George Floyd was a catalyst for recent protests and scrutiny, that the original police report was drastically different from what was actually filmed, and that an independent autopsy reached a drastically different conclusion from the official autopsy.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact