> but I smell a weasel in the phrase 'crime rate', because law enforcement is discretionary and also/therefore racist
> This is why homicide rates are useful. For most cases, it is pretty obvious a crime of some sort occurred.
Tldr; homicide conviction rates are subject to other artifacts and can't support either side very well without additional data.
Law enforcement is discretionary not just in deciding what counts as a crime but in who they decide to investigate and other tactics employed. Not that this hugely affects any relative statistics, but the murder false imprisonment rate is estimated at 4-11%, and false imprisonment (as far as we can tell from subsequent DNA exhoneration and whatnot -- this might not be representative of all murders) disproportionately affects black people. Not that this necessarily says anything about broad crime racial statistics, but it could throw any direct comparisons off.
More importantly though, 30-40% of murders go unsolved, and it's probably not reasonable to assume that distribution to reflect the distribution of successfully convicted murders (if for no other reason than the race<->income correlation). That's such a huge monkey wrench in any comparative statistics that murder conviction rates without any additional information don't strongly support either side of things.
> this violent crime rate disparity exists even when controlling for SES
That's exactly the kind of additional data I was asking for. I just didn't think murder conviction rates by themselves were sufficient evidence to support the claims they were being thrown against.
> Criminologists consider homicides to be reliable because it's difficult for police to ignore homicides.
That side-steps my point a bit doesn't it (in hindsight I should have been more clear)? Being difficult to ignore makes homicides generally reliable for some kinds of comparisons (e.g. comparing violent crime between countries), but that reliability doesn't transfer well. E.g., when a plausible suspect is thrown in jail the spotlight is removed from everyone involved. I gave specific examples where despite a high reliability in overall murder rates, piggybacking on that reliability to lend credibility to the racial distribution of conviction rates isn't justified (without additional information). The rest of your comment notwithstanding (I did appreciate it and do generally agree with you), rebutting by reaffirming points I already agree with and pulling in criminologists as experts in the vaguest way possible isn't satisfactory.
> Note again that virtually no one is arguing that blacks are inherently criminal, but that these disparities are a vestige of centuries of severe institutional racism--it would be shocking if no disparities remained only 60 years after the formal end of Jim Crow.
No argument here. I tend to agree with the claim, but the evidence (based on murder conviction rates) was suspect.
>The unfortunate truth is that black people are more likely to commit violent crime in the US.
It doesn't make much sense to cite the proportion of black people convicted of a crime by US law enforcement as evidence refuting the claim that the US law enforcement's interaction with people changes according to their race. The proportion of black people convicted of [category] crimes could be high for many reasons, including actually higher incidence of such crimes among blacks, racial prejudices at various points along the legal path of alleged criminals, and any combination of those or other factors.
> IIRC, blacks are 13% of the population yet 30% of the police killings.
Yes, and blacks are 13% of the population but 38% of the murderers (to use your & the OP's numbers). The point is that in two random encounters with two random people (one black, one white), the black person is more likely to be a murderer than the white person. That's not racist: it's just a fact.
As an example, I'll use the 2013 FBI crime statistics (https://www.fbi.gov/about-us/cjis/ucr/crime-in-the-u.s/2013/...) and 2015 U.S. Census data (https://www.census.gov/quickfacts/) — yes, there will be some inaccuracy due to using different years' data, but the 2015 crime data isn't yet available by race that I could find. In 2015, 77.1% of the U.S. population — 247,813,910 people — were white; 3,799 whites were arrested for murder in 2013, for an arrested-murderer rate of 1:65,200. In 2015, 13.3% of the U.S. population — 42,748,703 people — were black; 4,379 blacks were arrested for murder in 2013, for an arrested-murderer rate of 1:9,760.
Assuming the correctness of those numbers, and assuming that black and whites are fairly arrested for murder, any random black person one encounters is 6.68 times as likely to be a murderer as a white person.
Now, there are no doubt some very powerful arguments that blacks and whites are not equally treated when it comes to murder arrests, but even if only half of all blacks arrested for murder are guilty and as many white murderers are never arrested for murder as are, the multiplier would still be 1.67, and honestly those assumptions are a stretch.
Given the previous assumption of fairness, police killing blacks 3.49 times as often as whites would indicate that they are actually almost twice as cautious of killing blacks (or twice as quick to kill whites) as the actual random-encounter risk would predict: 6.68/3.49 = 1.91.
> a huge issue is racism in law and police policies (not necessarily in individual cops) - crime statistics need to be taken with a grain of salt
Homicides are one way to dis-entangle this (in jurisdictions not prone to To Kill A Mockingbird asshattery). It's objective as to when it happens. And our courts are relatively better, at least in the post-DNA era, about due process when it comes to such crimes.
It appears blacks committed 52.2% of homicides from 1976 through 2005 while representing 12.3% of the population [1]. The 2013 statistic for "murder and nonnegligent manslaughter" arrests of blacks is also 52.2% [2]. That said, (a) arrest does not mean conviction and (b) we decided long ago, and rightfully so, that projecting population characteristics onto someone who did not choose to be a member of that population is morally wrong and therefore, oftentimes, illegal.
> Now we can debate why the black population commits more crime, but it’s not a racist policing discussion.
Criminal statistics are a record of who is arrested and prosecuted, not a record of who commits the most crime. If a black criminal is more likely to be arrested than a white criminal, then the statistics will reflect that.
"It doesn't make much sense to cite the proportion of black people convicted of a crime by US law enforcement as evidence refuting the claim that the US law enforcement's interaction with people changes according to their race. "
It actually does.
The rate of violent crime among African Americans is massively disproportionate. There's no way any kind of racial bias on the part of police could account for it.
A major study done in the late 2000's actually showed that while African Americans did receive more punishment for similar crimes - once 'criminal history' was taken into consideration, the difference evaporated.
Given that there is a massive difference in 'recorded convictions' (which is arguably biased because of 'racist cops') and numerous attempts to study the prevalence of racism in the system - it's pretty safe to assume that there's no way on earth that 'racist cops' are why there is why African Americans are about 700% more likely to commit violent crimes than others.
It's a sad situation, surely, but it's important to keep the facts in check.
About 50% of violent crime in America is committed by African American men, though they are only about 7% of the population (about 13% including women). This is quite radically high and it's not even explained by poverty.
The sheer size of the discrepancy between ethnic groups is unsettling.
> of all crimes committed 26.9% were by black perpetrators. So in relation to that statistic
If you accept the possibility that police killings may be skewed by racism, why are you simply assuming that arrests (and therefore trials and convictions) are not?
> Crime stats aren't collected in a scientific vacuum. They reflect the previous strategies the police used to fight crimes.
It may also just reflect reality. Let's say you have a group of people that commit a specific kind of crime significantly more often than another group, i.e. white male investment bankers are more likely to commit tax fraud than female black nurses. Any reasonable policy fighting this kind of crime would have to look biased against white men when it comes to enforcing tax fraud. I think that no one would reasonably call such enforcement policies bad or racist.
But when black men are significantly more likely to be prosecuted for violent crime, suddenly it's a racist policy and must be the racist polices fault, because that's the only acceptable answer.
> And crime statistics are directly dependent on the policing and prosecution of crime--they don't reflect the objective reality of crime, they reflect the outcomes of an imperfect system that detects and punishes crime.
That's true for drug use crimes and to a lesser extend drug selling crimes, which make up way too many of the crimes prosecuted. Similarly for other crimes where police have a large amount of discretion to decide whether or not to arrest someone.
However there are also big differences in crime rates by race for murder. It's very hard to dismiss this as the result of how murder is policed. Police generally have much less discretion in how to deal with murder.
> If blacks commit so much crime, why, and if they don't, why are the numbers inaccurate?
We don't know this because the figures cited are biased towards successful convictions. If white folks are able to afford better, more competent lawyers, you can expect them to mount successful defenses. I've never, ever, seen anyone post the 13/50 stat without acknowledging this fact in their comments or replies. And I've seen it a lot.
> The real incidence rate is hard to tease out because of the presumption of criminality baked into the American system
I don't think it's as hard as you make it out to be. We have multiple independent sources collecting this data. The FBI collects racial data on perpetrators and victims, and multiple national surveys ask victims of crimes about the demographics of the perpetrator. The numbers from all these independent sources are fairly consistent.
Murders are especially easy to get reliable numbers on because there's a body. We know black people are murdered at dramatically higher rates than any other race and that most murder is intraracial. The reported racial crime statistics are accurate unless we believe that white people are sneaking into inner city majority black areas to murder black people undetected in massive numbers and the FBI is conspiring with multiple independent data collection groups to cover it up.
> the latter in the absurdly skewed prison population the US is known for.
Are you sure? Since most homicides are intra-racial [1] (despite the impression one gets from those the media choose to focus on), we can use victim race as proxy for offender race (I'd use offender race directly, but then you'd just blame it on police racism). Since it's hard to manufacture a corpse, or fake its race, we can further assume that data is largely free of police bias.
So white+Hispanic are 50% of homicide victims [2] (which we use as proxy for perpetrators), and 58% of the prison population [3]. For blacks, its 44% and 37%, respectively. Doesn't look particularly skewed to me. There is some anti-white+Hispanic bias, but since we're looking only at homicide to avoid police bias, we're not seeing the whole picture, which may explain the disparity.
Of the 12,664 murders 4,077 were cases where offender's race was unknown leaving 8,587. Males were responsible for about 90%. 90% of 8,587= 7,728 US murders we know to be committed by males. White males committed 45.2% of 7,728 murders which = 3,143 murders. Black males committed 52.4% of 7,728 murders which = 4,049 murders. Black males make up 7% of this nation's population with about 19m individuals. White males make up about 28% of this nation's population with about 112m individuals.
Odds of a white man being a murderer: 1 in 35,634.
Odds of a black man being a murderer: 1 in 4,693.
African American male is 9 times more likely to commit homicide. It's very unlikely that biased policing is the cause of 9x homicide conviction rate.
How is this even possible? Aren't black responsible for more than 50% of the murders despite being only 13% of the population? These stats don't add up.
All of these are considered crimes in all jurisdictions I know of in the USA: theft/burglary/robbery, violence or threat thereof, and reckless endangerment. If you commit one of those acts, you have definitionally committed a crime, even if you are not convicted in a court of law.
> dismisses the overwhelmingly higher rates of arrests for non whites as not being a systemic racism issue
I encourage you to re-read what I wrote, because that isn't what I said.
My post raises a question of causality. There are several common proposals for why arrest rates for blacks are disproportionately high:
1. Police are racist, so they arrest black people at higher rates than other people, regardless of criminality.
2. Criminal law is written specifically to target black people.
3. Black people, due to systemic racism elsewhere in society, end up committing more acts-that-are-legally-defined-as-crimes, and this eventually leads to a greater number of arrests.
4. Actual racist ideology, which is generally vile and I will not repeat it here.
The reality is probably a combination of explanations 1-3, and there is surely mutual causality among them.
Any serious attempt to understand the problem must attempt to disentangle these explanations from each other.
> assumes that white people commit fewer crimes with no data to back it up
I said that we specifically don't have good data, but it's not safe to assume that all demographics commit crimes at the same rate. See above.
> Sorry but that is gaslighting and victim blaming. Tell the parent of a black teenager to not be overly fixated on police reform.
This is a bad faith misrepresentation of what I wrote.
If anything, the media getting stuff kinda-wrong but still raising awareness is a net positive.
Also it should be obvious that police reform is necessary, and obviously people will care the most about the issues that affect them personally.
Moreover, nothing ever seems to get done without single-minded people focusing intently on a single problem. We need people working on all aspects of the racism issue, including racism in policing and law.
But you simply cannot look at the arrest numbers and assume that the discrepancy is definitely and entirely because of police racism. It would be fallacious to do so, and if you intend to put forth policy based on data then you should try to avoid logical fallacies in the process.
>In 2008, the [homicide] offending rate for blacks (24.7 offenders per 100,000) was 7 times higher than the rate for whites (3.4 offenders per 100,000)
> In fact, crime rates among poor whites are virtually indistinguishable from those among poor blacks.
This is untrue. From my prior research, virtually every analysis concludes the opposite - that income alone cannot explain the difference in crime rate between black and white Americans. Most studies show income as 30-60% of that regression, which is huge, but when comparing similar income brackets, African Americans have a higher crime rate than whites in almost all of them under $100k. To cover the full difference, it's necessary to bring in other factors (in particular, the rate of single parent households). Please cite a source for the claim you have made or retract it.
> So poor blacks kill each other in huge numbers, and poor whites don't. What's your point exactly?
90% of black murder victims are killed by black perpetrators, 84% of white murder victims are killed by white perpetrators, so I'm not sure why you suggested that poor white people don't kill each other. My point is that the higher black on black homicide rate is what we'd expect to see for a population that is disproportionately concentrated into densely populated low income regions, which will push up their representation in overall homicide stats.
> Also, maybe you missed the black on white numbers which are off the charts compared to the inverse
What exactly does off the charts mean?
13.6% of white murders are committed by black people, 7.6% of black murders are committed by white people, so yes, blacks murder whites at a higher rate, that is clear, but what is your point?
> you have people on HN saying pretty much that most of the black people who get shot and killed by police deserved it and that black people are stupid violent criminals so it's no wonder they get killed so often
I haven't seen anyone say anything so nasty, but I have seen the FBI's crime statistics:
Blacks make up 13.2% of the population, yet are arrested for 49.4% of the murders (3.7x their share), 32.5% of the forcible rapes (2.4x), 54.9% of the robberies (4.1x), 34.1% of the aggravated assaults (2.5x), 38.5% of the violent arson (2.9x) and 31.9% of other assaults (2.4x). The numbers for whites (77.1% of the population) are 48.2% of the murders (.6x),
Now, it's important to note that the explanation is almost certainly due to poverty and class; class in America being unfortunately highly correlated with race. It's also important to note there are almost certainly structural factors which protect some whites from being fairly arrested and lead to some blacks being unfairly arrested.
Given those violent crime statistics, is it not understandable that people in a potentially violent situation fear more for their lives when facing a random black, who is approximately 6.1 times more likely to commit murder than a random white?
And in fact, adjusted for the racial disparity in homicides, the police are 1.7 times more likely to kill whites than blacks; adjusted for the racial disparity in cop-killing, the police are 1.3 times more likely to kill whites than blacks[1].
'Black people' are not 'stupid violent criminals,' and anyone who says so is a racist idiot. But a higher percentage of black people are violent criminals, and thus it's no wonder that a higher percentage (but fewer absolutely, and fewer adjusted for the higher rate of crime!) get killed for violent crime.
Now, how does one reduce the racial disparity (by reducing the rate of black criminality to that of white criminality, not vice versa, I hope)? That's really tough to answer. Maybe it has something to do with ending the War on Some Drugs, which takes fathers away from their families; maybe it involves basic income, which would eliminate poverty; maybe it involves school choice and/or vouchers, to enable blacks to escape terrible schools and get good educations; maybe it involves using social persuasion to reduce the culture of violence (fat chance when that culture is so highly profitable!).
> This is why homicide rates are useful. For most cases, it is pretty obvious a crime of some sort occurred.
Tldr; homicide conviction rates are subject to other artifacts and can't support either side very well without additional data.
Law enforcement is discretionary not just in deciding what counts as a crime but in who they decide to investigate and other tactics employed. Not that this hugely affects any relative statistics, but the murder false imprisonment rate is estimated at 4-11%, and false imprisonment (as far as we can tell from subsequent DNA exhoneration and whatnot -- this might not be representative of all murders) disproportionately affects black people. Not that this necessarily says anything about broad crime racial statistics, but it could throw any direct comparisons off.
More importantly though, 30-40% of murders go unsolved, and it's probably not reasonable to assume that distribution to reflect the distribution of successfully convicted murders (if for no other reason than the race<->income correlation). That's such a huge monkey wrench in any comparative statistics that murder conviction rates without any additional information don't strongly support either side of things.
reply