May 10, 2023Liked by Noah Smith

Just seems to make sense that violence requires proximity and most people's social circles (and neighborhoods) are more racially homogenous than expected by random distribution.

More illuminating data (to me) would be rates of violence across class lines. Even the Cash App founder's murder was by someone in his social circle.

Expand full comment

So who shoves it in our faces and why?

Must be somebody shoving and must be for s reason.

Expand full comment

Social media people and some traditional media people. For attention. It scares people, and fear = attention.

Expand full comment

First time I have noticed the red tick. Do you have to pay for it or is it performance related I.e. based on number of subscribers?

Expand full comment

It's based on number of subscribers.

Expand full comment

Rupert Murdoch media outlets like the NY Post constantly feature stories of Black-on-White violence deliberately to stoke the right wing

Expand full comment

They couldn't feature them if they weren't happening. Or do you think they just make it up?

Expand full comment
May 10, 2023·edited May 10, 2023

I didn’t imply that. In a massive country, everything happens all the time. You can fill the front page of a newspaper with whatever stories you want. Most people are subject to availability bias and assume whatever is put in front of them is more common than it is. To put it in perspective using a non-political example, every single shark attack is reported in the news because it’s so sensational, even though it’s rarer than anything you can imagine; consequently, people overestimate their likelihood of being killed by a shark while swimming in the ocean.

Different audiences have different concepts of the sensational or salient. News providers target them accordingly. Reporting does not follow reliable statistical regularities but develops and feeds reinforcing narratives. Conservative and liberal news outlets both are constantly delivering content that skew people’s perceptions of the prevalence of all kinds of trends and omitting stories with. countervailing narrative. I do believe news bias is real and, at reputable outlets, mostly issued through what stories are even covered.

Expand full comment

I don't disagree with you. But I turn it around a different way. Press coverage is about meaning, not prevalence. The media chooses to pull stories and narratives out of the mass of facts and events. This is what news judgment is. This is how humans understand the world. They need not apologize for this.

A while ago I did an experiment looking for stories about police violence on Google News, which features a mix of local and national news outlets of all stripes. Now, statistics say that more white people are killed by police in a given year than black people. And yet I had to go all the way back to 1996 to find a story about a white guy getting killed by the police (and this was about a guy who barricaded himself behind a car and shot it out with the cops). Many of the other stories on this subject achieved large-scale, multi-day coverage stoked by activist performances, unchallenged by other reporting. At this point, the event isn't the police killing anymore, it's the "community protests" over the police killing. The activists achieved their goal (which they are good at) and the media lost the plot.

This is a meaningful subject to black people especially, but also to the wider community, because of our national history. And I think if I was barraged by stories about police officers killing people like me, I would be worried about that in my daily life. And the black people I know do in fact worry about that. At the same time, I recognize that others are worried about other things (such as soft-on-crime prosecutors) and they pay attention to the narratives that are meaningful to them, which papers like the Post give to them. All these stories are based on facts, they are not made up, but is the first thing false, because the second thing is true, or vice-versa? Both are true, and the narratives reflect that.

I added the New York Post to my reading during the 2020 riots, when I found that I literally could not find out basic facts about what was going on in my city in my paper of preference, the New York Times. It came out that this information was actively suppressed and distorted by this media institution that I had felt to be as reliable and as balanced as any human could make it. Thus, when the Times was giving editorial space to columnists arguing that there is a tide of anti-Asian violence that is due to "white supremacy", I could turn to the Post and see surveillance photographs that indicated otherwise. So the Post, this right-wing biased rag, provided in this case some necessary balance. I think I can read it from time to time without being brainwashed.

A healthy variety of such stories and narratives, and an open mind, is the only guide to all of this.

Expand full comment
May 10, 2023·edited May 10, 2023

This is more balanced than your first comment. Having watched CNN or MNCBC you do get the liberal side of bias which you touch on in your last comment.

But in general it is pretty deplorable.

Expand full comment
May 10, 2023·edited May 11, 2023

During the recall of the San Francisco DA (Chesa Boudin) people were shoving Black on Asian attacks in both social and mainstream media all the time. Sometimes you would see the same exact attack every night for a week on KTVU. There *was* an increase in hate crimes against Asians (almost all non-violent) which is of course terrible. There is no evidence that it was fueled by Blacks, in fact what evidence we have is that it was mostly Whites doing this. The recall was heavily funded by billionaires, with much of the money from out of state as part of the GOP campaign to use "get tough on crime" as a wedge issue.

Since the successful recall, I have only seen one Black on Asian violent crime reported on KTVU and it is mostly disappeared from Twitter.

Expand full comment

This reminds me of when we were constantly hearing stories about illegal immigrants committing crimes, with Trump actually inviting victims’ family members to the State of the Union. Tucker Carlson also got on the bandwagon. This was to support the “illegal immigrants are criminals” narrative Trump was then pushing, to the complete opposite of statistical reality.

In recent months, there have been an improbable number of stories in the NY Post about transgender individuals committing crimes. I believe these stories are being deliberately selected for clicks as the transgender populace has become somewhat of a right-wing boogeyman.

Expand full comment

I assume Kron is some cable network on the right.

Lot of assumptions though.

In the UK the prevailing media broadcast from the USA would be of race baiters. Because there is a massive upsurge in the UK slums, especially London, of murder and knife crime, which is largely gang black against gang black with some gang asians thrown in and very few whites involved at all, everybody or most are cynical.

There is little violence against gang whites but such is the demographic change now London is an immigrant city and the whites have moved out. So there is little of border disputes.

Expand full comment

KRON is one of the local news channels in the San Francisco Bay Area. It is not on the right by US standards, as there is no demographic for this in the region but it is probably a bit more conservative than the other local news stations. My wife likes it.

Expand full comment

Noah, I am curious what you think about Kareem’s points on age adjusting the data.

Given that Black folks are younger, and crimes are more likely to be committed by the young, and they are poorer, and many crimes are driven by poverty.

Thanks for doing this, btw. Jordan Neely’s death was tragic, but the way it’s being characterized and the reaction is beyond belief.

Expand full comment

A little younger, but not a lot younger, especially in the statistics where Hispanics are included within the White and Black populations.

My least favorite talking wrt crime is "many crimes are driven by poverty". I think people badly misunderstand how poverty and crime interact. Once you remove ethnicity from the equation there is very little correlation between poverty and crime. The poorest large metro in the US is El Paso, for example, which also happens to have just about the lowest murder rate among large US metros.

2nd - most violet crime is not economic - more typically people get into arguments or fights over girlfriends or gangs engage in cycles of revenge killings. Sure money is sometimes involved, but outside of high-level drug dealers we're not talking about life-changing sums or people stealing food to survive. It's more like, "hey punk, I want that iphone". During the last several economic recessions crime, in fact, fell and then rose during expansions that followed.

And levels of violent crime haven't really fallen over the entire past century despite drastic expansions in the social safety net. If crime was in any way a rational reaction to poverty you'd think it would drastically fall in times of lower unemployment because it's so much more rational to get a job, even a min wage job, than to risk your life and jail time robbing people.

Expand full comment


Source: I made it up

Expand full comment

Lolz? That's your best counter-argument?

I do have sources on almost all of this. And on the off-chance that you're not an unserious person and you have some capacity to learn and grow and be challenged by new views that are outside of your preconceptions I'll share them here.

Consider the chart "Violent crime rates (per 100,000) in the United States (1960-2018)" here: https://en.wikipedia.org/wiki/Crime_in_the_United_States#Crime_over_time

All categories of violent crime are up from where they were in the early 60s, before the War on Poverty and vast expansion of the welfare state that occurred at that time. Because of differences in reporting and changes in legal definitions I find it best to stick to homicide rates when comparing places and times. And those rates, at least, have not risen quite as much. But considering advances in emergency medicine and that we are a much older country nowadays (the % 65+ has doubled, for example) I'm not sure that's much to celebrate.

Now looks when the biggest rises in violent crime happened - the biggest increase occurred in the late 60s, a time when welfare expanded, the economy was booming and poverty was being rapidly reduced. If poverty caused crime we would expect a decrease at that time, right? Here's a link to % in poverty to show you that the late-60s drop in poverty was very real: https://en.wikipedia.org/wiki/File:Number_in_Poverty_and_Poverty_Rate_1959_to_2011._United_States..PNG

I don't want to type in the details of every recession and corresponding crime changes, but I think the great recession is worth mentioning because it was a severe recessions that did long-lasting damage to employment. And yet...you can see for yourself in that source that crime fell in that period.

Here's another good source on motivations for violence. Read the summary findings on page 9 and notice poverty is never mentioned as a risk factor and material need is never mentioned as a reason for violence: https://cjcc.dc.gov/sites/default/files/dc/sites/cjcc/release_content/attachments/DC%20Gun%20Violence%20Problem%20Analysis%20Summary%20Report.pdf


"In Washington, DC, most gun violence is tightly concentrated on a small number of very high risk young Black male adults that share a common set of risk factors, including: involvement in street crews/groups;significant criminal justice history including prior or active community supervision; often prior victimization; and a connection to a recent shooting (within the past 12 months).

While the majority of people involved in shootings, as victim or suspect, are members or associates of street groups/gangs, the motive for the shooting may not be a traditional gang war. Often shootings are precipitated by a petty conflict over a young woman, a simple argument, or the now ubiquitous social media slight. "

To be fair earlier in the report they mention that 7.9% of homicide have robbery as a partial circumstance and 2.9% involve drug robberies. But these aren't guys shooting each other to feed their families.

Sources on El Paso, Texas. It's 12th safest by homicide rate on this list of the 100 biggest US cities:


and 50th worst in terms of GDP per capita on this list of 380 US metros. But there are many other examples of safe but poor cities, mostly in Appalachia and along the Mexican border.

Expand full comment


Good article as usual. However, if one is looking at "risk" all of this is way down in the chart from other risks such as

1) Cronic Health... largely driven by diet choices "programmed" by food industry

2) Automobile related accidents

3) Suicides

An article which outlines these risks relative to crime would be interesting I think.


Expand full comment

While I agree with your analysis about the rareness of interracial crime and how anxiety about such occurrences is blown out of proportion, I think you’re glossing over one of the central arguments of the debate being discussed. It is about how information is being presented. Using the above data, on the basis of race, who is more likely to commit interracial crime on a per capita basis? Why can’t we discuss such an answer comfortably in an open forum? That is the question being presented to liberal America by the right. I don’t know how to answer such a question, but I think that it is something we as a country might have to work through.

Expand full comment

I think the interracial part is pretty uninteresting. The most shocking part of it is how rare it is relative to murder in general.

In a truly colorblind society where skin color is as socially unimportant as say, blood type, you'd expect Black people, when they commit homicide, to murder Black people only 15% of the time or so. The fact that it's closer to 80% means two things: 1) there's still a lot of social segregation in the country and 2) Despite media reports, there's very little actual meaningful racial hatred out there.

It's already clear from the broader data that violent crime is a bigger problem among Black Americans than for other census categories (both as victims and as perpetrators), and many media outlets do avoid that topic. But the interracial aspect doesn't really add much, except perhaps a somewhat hard to grasp data point on how good racial relations actually are.

Expand full comment
Comment deleted
Expand full comment

"despite black on white violence bring an order of magnitude larger"

I don't think this is quite right. The amount is closer to double. To get to an order of magnitude difference you have to rate adjust it, but that makes the statement "black on white violence" more muddled in its meaning.

The perpetrator rate of black on white violence is around 10x the w on b perpetrator rate. But from a victim perspective the rate of black victims to white crimes is 2.5 times larger than the reverse.

2x total B on W violence, 10x B on W perp rate and 2.5x B victim rate are all correct in themselves, but I think people are muddling them all together with imprecise language (no offense).

At the end of the day, though, I should probably emphasize what I said earlier - it's all pretty uninteresting and media and tweeters on either the L or R side trying to emphasize it are missing the forest for the trees.

Expand full comment
Comment deleted
Expand full comment

There are other levels of concern besides just individual risk assessments. People care and discuss and argue about group victimhood rates all the time, in which case the "blacks are 2.5 times more likely than whites to be a victim of interracial violence" framing is relevant. In a full-on inter-communal warfare mode we might even only care about total numbers, ie the 2 to 1.

Expand full comment

I don’t know why you claim that we can’t discuss this comfortably in an open forum. There is a whole cottage industry that barely talks about anything else. No one is stopping you from talking about it right here and right now. What did you want to talk about specifically?

Expand full comment
Comment deleted
Expand full comment

Most new of interracial violence is white people assaulting black people? That is not what I see at all. What kind of news do you consume?

Expand full comment

I think your risk of interracial murder is wrong.

Your numerator is based on the only 40% of total murders. Assuming that the other 60% of murders have the same ratio of black/white white/black murders then your numerators should be the 615 and 1415 respectively.

This would make the actual risk of a white person being murded by a black person of 0.54 per 100K and of a black person being murdered by a white person at 1.28 per 100K. Same 2.3 ratio, but different individual rates.

Still pretty low, but not as low.

Your violence table at 84% of reported crimes would also change I would imagine.

But one thing is for sure, for a modern country it's a crime that we don't have better data. Why not have a standardized reporting system for everything.

Regardless of stats, I use my own personal experience. I don't know anybody that has been murdered by anybody else, so I am going with overall risk of being murdered is low.

I will however be staying away from South Korea just in case.

Expand full comment

My comparison would be hate crimes. I would guess hate crimes are a vanishingly small number of overall crimes, but they punch above their weight because of their intent. Activists are equating every time interracial violence occurs as a hate crime, would be my hypothesis.

Expand full comment
May 10, 2023·edited May 10, 2023

I think that by basing this argument on aggregate nationwide statistics you are entirely missing the point. The black population in the US is not only smaller, but highly concentrated. There are very few blacks in many parts of the country, and relatively few remaining white inhabitants of central cities. This has to affect the risk calculation, simply because of relative opportunity for contact at the border where they meet.

Consequently, I'd like to see the numbers for central cities and suburbs of large metropolitan areas, where black populations are larger and actually-resident whites (especially moderate-income whites) are fewer. This is what the crime reporting focuses on. It's cold comfort to a New York subway rider that there are almost no black people in rural Kentucky, or even the more prosperous parts of New Jersey, to do or suffer violence. Many suburbanites segregate themselves specifically because they want to avoid elevated personal risk that they, or people they know, or their families in the past, previously experienced. They are safe at home, and will fight to preserve that. It is only human.

It would be interesting to do the same calculation for the Seventies, when the attitudes and media tropes were formed. Crime was much higher then, more white people lived in cities proximate to black people, and racial tension was higher. Under those conditions, it would not be inaccurate to describe "white flight" as forced exile. New York, for instance, lost more than a quarter of its white population in this period. Note that this is not ancient history. It was experienced by the parents and grandparents of people who are now alive.

People who feel this way have not been brainwashed by the media and it's insulting to them to argue this is so. Maybe their perceptions are less accurate under current conditions, but the mechanism for thinking otherwise needs to be acknowledged.

Expand full comment

Crime reporting focuses on sensationalism and foments racial bias. Are you surprised by this?

I guess we can both agree that racial bias should be acknowledged.

Expand full comment

I'm not sure we do agree, because I'm not sure that we mean the same thing by "bias." To me, "bias" implies deliberate bad faith. The biased party knows, or should know, some facts, but either chooses not to know them, or actively suppresses them to achieve a purpose. I don't think we have that going on here. I think rather that we have a difference of values that leads to a difference of perception of what is important and newsworthy.

If you tend towards a systemic view of the society and its problems, tales of individual ill-doing are not going to seem as relevant as they do to people who tend to emphasize personal responsibility above all. Reporting that so many of those involved in crime are black (a matter of public record) strikes the former as prejudicial and socially harmful, and the latter as a social fact that is important, morally, to acknowledge in terms of individual right and wrong. Both sides have a point. It's not "biased" for this difference of opinion to exist. It is only biased if you are sure that you are right and the other party is acting in bad faith. Personally I try to avoid that kind of thinking.

I will allow that headlines are often sensational, in proportion to editorial perceptions of what will grab the attention of their audience. This is not limited to Fox News and other conservative outlets, or to coverage of crime. Since Trump and especially since the 2020 riots, the august New York Times has been relentless in this game. They are no longer too dignified to use the word "lie" in their headlines, and the social justice angle of such crime coverage as they consider worth reporting, is played up in big, black letters. Whether this is bias or not is a matter of opinion, but whatever this is, it is certainly attention-grabbing and emotionally satisfying to their particular audience.

Expand full comment
May 11, 2023·edited May 11, 2023

Most criminals are white. I don't see on the news a majority of white criminals, so I can't figure out what you are talking about. There is a pretty strong bias for Black people to be overrepresented in crime reporting over their rate of criminality. You claim that this has to be due to an intent to deceive or manipulate. Since we can't divine motive, I think that definition is pointless. What matters are the outcome.

I share your observation about the New York Times. It's not an unbiased news source. For objective reporting I look to other sources. I find myself reading it less and less in favor of the Wall Street Journal news reporting.

Expand full comment

It depends how you define criminal. If you include all jaywalkers, people who have ever littered or sped on the highway or smoked a joint quietly in their own abode then everyone is a criminal. So it's more reasonable to think of criminals as "people who have committed serious crimes" however you define serious.

One proxy is murder, since it's the most reliably reported crime and arguably the most serious. Blacks are a slight majority of murderers with Whites and Hispanics roughly a quarter each. Another proxy is incarceration, where Blacks are a slight plurality at 40%.

Which categories are overrepresented is another question entirely and I have no idea how to answer it. YMMV depending on your news source, of course. But I tend to find that many left-of-center people underestimate the extent of the disproportionality of serious crime among Black Americans and right-of-center people often overestimate it.

In any case, most serious crimes have a victim who tends to be of the same ethnicity, so it's sometimes odd to hear simultaneous complaints that a given media over-represents X race as offenders but underrepresent them as victims.

Expand full comment

Plenty of white criminals on cop reality shows. If you look at Fox News as consisting of morality tales that reflect the concerns of their viewers, the point that is really being made is not about black people at all, it is about liberals, who they think are soft on black criminals because they have contempt for ordinary white people. Which is uncomfortably close to the truth.

Expand full comment

How come Asians are never included in these statistics?

Expand full comment

Asians are in 2018 stats but the FBI did not produce a summary table in 2019. I don’t know why. They have not produced this information in an easily digestible way since then. The data is there but it’s too much work to crunch it.

Expand full comment

Even if it doesn’t spark actual riots and pogroms, it could lead to the breakdown of social cooperation between people of different racial groups, and paralyze the government via racial polarization of politics.


Not to rain on your parade but that's already a massive problem in US politics with many people opposing programs that would help them on the fear that it might help "those people" (meaning Black and Brown people) more...

In other words, in order to succeed as a diverse society, we have to find some way to remind people that interracial violence is actually rare, and that the incidents that get shoved in their face on social media are being selected in order to fan their outrage and fears.


I don't know what you guys study in middle school but, in France, French (and language) teachers will spend a fair bit of time looking at media sources/articles etc. on common topics (on racism, immigration, integration, gender etc etc) and try to make the students think/build up a critical mind capable of analysis. The results are so-so but I'd say Le Pen has to be a notch better than Trump to be successful in France. Sure, a fringe needs nothing more than naked racist appeals, but many more voters need it to be at least half competently dressed up before they allow themselves to go for it.

And, personally, I think one massive failing of the French is that we study all kind of pretty useless maths (like geometry) but don't do much probability/statistics. It's changing (my kids are doing/did way more than I did) but it needs to be ramped up further. Stats is the only way you defeat appealing narratives.

Expand full comment

"Not to rain on your parade but that's already a massive problem in US politics with many people opposing programs that would help them on the fear that it might help "those people" (meaning Black and Brown people) more."

Which programs and what evidence do you have that fear of helping black and brown people is the true underlying motivation?

Expand full comment

It's a pretty well established fact.

But one study picked pretty randomly: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8853750/

"Higher levels of symbolic racism was a consistently strong predictor of lower social safety net policy support across health, housing, and employment policies"

Expand full comment

It's well-established in NYTs style bubbles. That's very different from well-established as an actual fact, from from biases and distortions.

In any case, this study is pretty tepid support for that thesis. I believe X and I support policy Y are separate questions. If X and Y are correlated it does not follow that X is a driver of Y. That may be the case, but the complexity of human thought and reasoning makes it very difficult to tease out the drivers and the study is very far from clearing that tricky bad.

On top of that: how many of these are "racism" or even "racial resentment"? Some of them are so closely adjacent to welfare support / skepticism that it's redundant to use them as explainers of welfare in general. I'd be more interested to see what the answers are if Black was replaced with "low income whites" or some other way to remove race from the question.


Irish, Italians, Jewish, and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors.


Generations of slavery and discrimination have created conditions that make it difficult for Blacks to work their way out of the lower class.


Over the past few years, Blacks have gotten less than they deserve.


It's really a matter of some people not trying hard enough, if Blacks would only try harder they could be just as well off as Whites.

For some of them I'm not even sure what the "racist" answer is, particularly 2 and 3. But all of them are things plenty of people think about their own racial or ethnic or national group all the time. It doesn't even take resentment or racist ideas to agree with some of these ideas.

Expand full comment

Good morning Noah, I think that this post is the most important thing that I’ve read recently. Why is this? I believe that no one wants to change their perceptions of the causes of the tension that we all feel. Personally, I want to forward this post to my small (a couple of dozen) circle friends who still read anything longer than an iPhone message or Twitter posting. When I do this, too often I get a two word reply stating “too long,” which at first offends me and makes me wonder why I bother with the forwarded email or why I don’t improve the attention of my circle of friends. But then, I persist because not doing so feeds into the problem. News is sound-bite journalism even including my standard bearers of PBS and MSNBC. Broadcast news is ratings based not thought based. It’s circular and frustrating. Keep,doing what you are doing. It must feel like shoveling analyzed data against a tide of ignorant bullshit. Sorry.

Expand full comment

It was widely reported in October of last year by AP, NPR, Reuters, etc. that only about 60% of the crime stats had been submitted to the FBI for 2021. Notable states and cities included FL, CA, NYC, LA. I am not aware of the stats for 2022 and whether they have even been submitted.

Expand full comment
May 10, 2023·edited May 24, 2023

"We pay a vastly inordinate amount of attention to these incidents of interracial violence, relative to how frequently they actually occur."

This is definitely true. It's important for people to have perspective about these things, and this logic also applies to all sorts of things: White Supremacist violence, mass shootings, police shootings etc. The media should in general report less on those acts of violence given the statistics underlying those events.

Expand full comment

That sort of misinterpretation of statistics and other numerical data is all too common. I highly recommend Innumeracy by John Allen Paulos.

Expand full comment

Very good, thanks Noah.

As @KetamineCal mentioned below, I'd love to see a breakout of violent crime by class/income strata. I feel that would be much more interesting although I'd bet we'd find outliers (like first gen immigrant communities that are poor but tightly-knit with relatively low crime). Good point on proximity, too.

Expand full comment

I'm sure there's some correlation. I've done similar analysis in the past on the state and county levels and there's correlation. It's much weaker if you control for ethnic groups, but still present.

But that said - my strong belief is that the causation runs in the direction of crime causes poverty much more than poverty causing crime. If I listed the many ways that crime mechanically causes poverty you'd be unlikely to disagree with them: they drain court resources, health resources, prison resources, police resources, they cause disinvestment, physical harms, psychological harms, disinvestment in positive things and investment in barbed wire, alarms and pit bulls. A single shooting is likely a 6 figure or more drain on the local community.

Maybe poverty causes crime a bit, too, but I think those linkages are harder to prove and more speculative.

Expand full comment

I wonder if there is anything in the white/Hispanic interracial violence that is caused by the social integration of Hispanics into white America, which I think Noah has discussed before?

I think what you’d see there is mutually particularly high interracial violence, since integration mean proximity (if nothing else, interracial relationships would drive interracial murders of women, not because interracial couples are more violent but just mechanically).

Expand full comment

It's ironic, but in a truly colorblind, fully desegregated nation levels of interracial violence would be much, much higher. If that were today's world we'd expect Black murder victims, for example, to be only 15% Black instead of 80% or so. The high percentages of intra-racial violence speak to both segregation and a low level of meaningful hatred.

Rates of interracial marriage have been surging over the last 2 decades between all ethnic groups, but especially White and Hispanic, and that is a huge driver of integration. It's somewhat missed in the data and barely reported on b/c the continual influx of immigrants (who are often already coupled and not as assimilate) keep the headline numbers down.

Expand full comment