Social Networks and Risk of Homicide Victimization in an African American Community

That is a new paper from Andrew V. Papachristos and Christopher Wildeman, here is the abstract:

This study estimates the association of an individual’s position in a social network with their risk of homicide victimization across a high crime African American community in Chicago. Data are drawn from five years of arrest and victimization incidents from the Chicago Police Department. Results indicate that the risk of homicide is highly concentrated within the study community: 41 percent of all gun homicides in the study community occurred within a social network containing less than 4 percent of the neighborhood’s population. Logistic regression models demonstrate that network-level indicators reduce the association between individual-level risk factors and the risk of homicide victimization, as well as improve overall prediction of individual victimization. In particular, social distance to a homicide victim is negatively and strongly associated with individual victimization: each social tie removed from a homicide victim decreases one’s odds of being a homicide victim by approximately 57 percent. Findings suggest that understanding the social networks of offenders can allow researchers to more precisely predict individual homicide victimization within high crime communities.

Some of those sentences could be framed for their importance.  For the pointer I thank DP.

Here are other papers by Andrew Papachristos.  Here is a paper by Christopher Wildeman.


"He that lieth down with Dogs, shall rise up with Fleas."

Wow - almost as if the experience of police statistics in DC during the crack years is a complete surprise. Relying on memory, during a year (late 80/early 90s) there were six hundred murders.

Of those murders, 37 were white people, and something like 95% were in the clear geographic boundaries of DC's largest open air drug markets (6 areas or so - not being in one was the simplest way not to get murdered in DC during its most famous period of homicide).

The conclusion drawn at the time by Metro Police? - most people murdered in DC were involved in the drug trade. Admittedly, some of the deaths involved people not belonging to the same social networks at all - they just shared the same profession.

Actually, I also seem to recall there was some HBO series on this involving a neighboring city - I've never seen it.

“The righteous should CHOOSE HIS FRIENDS CAREFULLY, for the way of the wicked leads them astray.” (Proverbs 12:26)

Reminds me of the now-famous Forest Troop story about baboons that was publicized in 2004 (e.g. at; the violent, aggressive, alpha-male baboons ate some bad food, got sick, and died, at which point the the rest of that baboon community began to flourish. The females took on a greater role in general, and the not-so-aggressive male baboons (who inevitably would have been drawn into more aggression had the violent alphas not disappeared) remained unaggressive, and the whole community benefited as a result. In other words, get rid of the bad apples (e.g. via imprisonment) and everyone else wins.

So you're saying "One bad apple spoils the barrel."

This is fun.

More or less, yes. Maybe not "one" bad apple, but certainly a small handful of bad apples in close proximity can cause a disproportionate amount of damage.

Doesn't quite roll off the tongue the same way as the original

i wonder what happens to one's chances of being a high wealth american if we remove one social tie to another high wealth american? and would Tyler frame the result?

Crime gangs and criminal families exist. Is that their point?

The point is that a surprising amount of the bad aspects of our society is purely voluntary.

So why should I care?

I don't know that your social network is entirely voluntary. You might be able to pick your friends, but you certainly can't pick your family. And you can only pick your frends from a limited pool of people, controlled primarily by outside forces.

I need Steve Sailer to interpet this for me.

The difference is that this is an attempt to blame the victim. SS blames the victim's ancestors.

"Homicide victimization" !?! It's almost like academics are trying to make their prose as ugly. We might call it Ugly Prose Chic. Any normal person would call it "being murdered."

My favorite is "negatively and strongly associated"

But, profiling...

"One death is a tragedy, a million deaths is a statistic." - Joseph Stalin

Majoring in the minors is more than clever diversion, and it won't even solve the problem either.

Love the mug shot.

This reminds me a bit of the paradox that the St. Louis police department found. Among its murder victims, a high proportion had rap sheets as long as their murders. Thus, at a time when St. Louis had the highest murder rate in the country, you could live there and not sense it unless you were withing certain social circles.

This does suggest a strategy for the police because just saying "scum killing scum" ignores the real danger to innocents (plenty of innocent children getting shot in Chicago) and to the degradation of order. Its something a decent database among the beat cops (not just of those with records, but of all contacts and their connections) could help. It also takes engagement, something that police are not good at in Chicago. (and frankly community leaders don't encourage)

The solution is simple: Sales of ammunition are free in the city, but restricted to lightweight, ultra-high velocity. frangible stuff with zero penetration.
Furthermore, there is a huge penalty reduction if the bodies are disposed of thoroughly.

The obvious way to parse this (reading the summary/excerpt, but not the paper yet) is that it reflects the distribution of dangerous people. That is, there is a smallish number of dangerous people in the world--people who are subject to kill those around them on small provocation. Those people pose a risk to all around them. A world in which most or all of them were in prison, dead, followed around by a policeman keeping them out of trouble, shipped to Coventry, etc., would be enormously safer.

An alternative way to parse it, though, is that the social networks are associated with a cluster of behaviors and such that raise the likelihood of being murdered if you're close to it. An obvious example is that drunks and drug addicts have behaviors that lead them to both be more likely to murder someone, and more likely to be murdered by someone. In that case, finding the few especially dangerous people to lock up won't help much, as the cluster of behaviors will continue.

Still another way to parse it is that social networks reflect shared personal characteristics. If violent people tend to associate with violent people, or people tolerant of violence, you could see this kind of pattern. (For example, in my social circle, smacking your wife around is the sort of thing that would quickly get you ostracized.) My understanding is that people tend to associate with people similar to them in other ways--IQ, education, income, race--so it wouldn't be shocking if this were also true of tendencies to violence, or lack of impulse control, or mental illnesses inclining people to violence. Anecdotally, this seems like something I've seen a lot in the world--people mostly seem to find their own level in mates--messed up people often end up together.

Each of these is plausible, and indeed all three seem likely to be at least somewhat true. And yet, they lead to different responses. If the social network/violence connection is mainly about a few super violent people, the solution is likely locking the super violent people in jail. If it's mainly about people inclined to violent crime clustering together for various reasons, it's not so clear what the solution to that would be.

The homicide rate has fallen a lot since its early 1990s Crack Years peak. There are various reasons, but massive incarceration is obviously one of them. It takes a lot of iterations, though. You lock up the scariest crack dealers or see them shoot each other, and then slightly less scary personalities replace them as crack dealers. Rinse and repeat and eventually you are left with guys on the streets who aren't as scary by nature and/or are more amenable to nurture (nurture in this case being the threat of years in prison or of murder by business colleagues).

One interesting point that hasn't been investigated systematically yet is that a lot of scary dudes got themselves removed not just from the streets (or from life itself), but also from the breeding pool. The young kids today are more likely to be fathered by the less scary guys who weren't locked up in the 1990s. In contrast, the cohort of young males that went homicidally nuts in the early 1990s tended to more so to be sons of guys who were on the streets in the 1960s and 1970s only because of the dominant Soft on Crime policies of the Warren Court era.

I was reading an interesting book by Geoffrey Canada the other day - Fist, Stick, Knife, Gun: a personal history of violence - and it was striking how an absence of Fathers worked out in these boys lives. You have a whole lot of scared young boys. They go outside, and they get victimized by slightly older boys who beat the crap out of them. They in turn victimize younger boys. But there are no adults to supervise them. The streets are really like the Lord of the Flies. Presumably Fathers are not present to begin with, but of those that are, a lot are in jail. Most places, two boys beating the living crap out of each other would have to stop because some adult would step in. Not in the South Bronx.

So I don't think it is precisely who their Father is. It is the fact literally none of them, or close to it, are around.

But it's been like that for a few generations in the black slums by now: no fathers around. Obviously, that's bad for nurture.

But, on the nature side, I'm starting to get hopeful: the scariest guys have been killing each other off or going to prison in such large numbers for so long that the generation born in the slums from, say, the late 1990s onward is likely to have fathers who tended to be more peaceable and reasonable than the previous generation, who were fathered during the Soft on Crime era when scary guys were mostly left on the streets to reproduce in large numbers.

Or maybe not. But it would make an interesting study to do. There are a lot of longitudinal databases out there, such as the NLSY79 and the NLSY97 for anybody who is interested.

But is there any reason to think that the rate of violent men killing each-other is higher now than during the colonial times? Or the middle ages? Or the Bronze age? Why would the evolution just start to kick in now?

But that can only be Nurture too - Fathers who are less inclined to solve problems violently raise children who are less likely to solve problems violently. A cultural thing. If, of course, Fathers have much significant influence on their sons at all.

The time frame is just too short for a genetic change to have taken place.

Besides, I dimly remember a study that said men in jail fathered slightly over the average number of children. Some women just love them some thugs.

A genetic change can happen quite quickly if the genes are already spread around and then you go wipe everyone at an extreme of a trait. If you killed every man over 6'2" and woman over 5'9" the next generation would be markedly shorter.

I believe that steve is suggesting we've locked up so many that we stuffed a large portion of the most violent away from females for much of their breeding periods.

If the behavior is mostly genetic, that sounds like it could cause a change in a generation

Careless, I am dubious that any genetic change could happen that quickly. For one thing, is the extreme edge of the trait being wiped out? If teenage boys shot each other in large numbers in the 1990s and declining numbers ever since, you are taking out a tiny percentage of the population over a two decades or so. People with the gene would be breeding at the older end - men have children into their seventies. And people too young to shoot anyone but with the gene would be growing up.

No one gets locked up for that long in the US. Especially not in their prime reproducing years. Let's be brutally frank and say that probably goes from 14 to 34 for African American males. What is more few women have many children. So those young men may spend the years from age 22 to 72 in prison, but they have time to have as many children as any suburb father.

I just don't see a genetic change working through this population in such a short time. Try it over a few hundred years perhaps.

As a matter of what did happen, I think you're right and Sailer is not. I'm just saying that it could, given the right and strong enough selective pressures and the right genetic diversity and sufficient genetic influence on the trait.

in my social circle, smacking your wife around is the sort of thing that would quickly get you ostracized

In my social circle, smacking your wife around is the sort of thing that would quickly get you arrested.

I'm interested social pressure that's not imposed by the police and courts. My guess is that social acceptability of violence varies from social group to social group in big and important ways. In communities where violence and threats of violence are more acceptable, you will presumably find more violent people, because they can exist there without being shunned by everyone. In my social circle, violence and threats of violence would get the cops called on you, but also would get you permanently labeled as unwelcome among your neighbors.

I once worked overseas on a project where everyone basically lived in the same company-provided accommodation except the very senior management. The accommodation was cheap and hence not particularly sound proof. Which meant that everyone heard bumps and thumps from some people's rooms as well as every argument every couple ever had. But we all ignored it. After all, there was no proof. We couldn't be sure. So what are you supposed to do? You have to work with these people after all. Well I didn't but other people did.

Maybe that was moral cowardice. It is certainly easy to condemn now. But perhaps that is where we are now - we have evolved to condemn this sort of thing in public, but in reality we continue to ignore quite a lot. It may have been different if any spouse had complained or carried any visible sign of violence. Maybe not.

I think it is interesting to consider this matter of social topology, rather than geographic topography (which do, of course, overlap). It's not really about bad neighborhoods, it's about bad groups of people.

Yes, I think the overlap question is key. Even without the research presented by this article, it's easy enough to believe that your probability of being murdered is affected by who you hang around with. It's also uncontroversial that your probability of being murdered is affected by where you hang out. And of course there are your individual demographic and behavioral characteristics (are you male, uneducated, and an abuser of mood-altering substances with poor impulse control?). The research question is (or should be): how important are those factors, relative to each other and in absolute terms?

There's a lot of collinearity there so it may not be easy to measure the importance of these variables simultaneously.

Um, have you considered looking at the paper?

What is the identification strategy? Disentangling the causal effects of social networks from those of unobserved covariates or selection is really difficult.

This is astonishingly apropos given the Reader’s recent coverage of Chief Keef (Keith Cozart), Chicago rapper, who is implicated in the assassination of a rival rapper, Lil Jo Jo, because of Twitter…

Cozart has actively invited and engaged controversy on his rise to fame thanks to his violent and anarchic songs. On his breakout hit, "I Don't Like," he boasts that he's "pistol toting and I'm shooting on sight" and "I done got indicted selling all white." The backlash against Cozart reached new heights after his flippant and reckless (alleged) response to the death of 18-year-old Joseph "Lil JoJo" Coleman, who was gunned down earlier this month while riding double on a bike with a friend on 69th and Princeton. This was hours after Coleman publicized his whereabouts on Twitter—and months after he released "3hunna K," a track that disses Cozart and his GBE crew. The next day, a series of tweets in response to Coleman's tragic demise showed up on Cozart's feed (he claims he didn't type them), sparking a police investigation into a possible connection between Cozart's crew and Coleman's death, jump-starting rumors of an intense rivalry between the pair, raising speculation that Interscope might drop Cozart, and launching a new wave of anti-Cozart fervor. One of the tweets on Cozart's feed particularly captured the public's attention: "Its Sad Cuz Dat N*gga Jojo Wanted to Be Jus Like Us #LMAO."

Like I always say, the biggest problem with being poor in modern America is not that you can't buy enough stuff, but that you can't afford to get yourself (and, more importantly, your children) away from other poor people.

I reviewed all 1,259 homicides involving young males in Los Angeles County in 2007-2009 using the L.A. Times' Homicide Report. The fraction who clearly weren't involved in anything illegal when they were killed were tiny -- e.g., an engineer who chased after thieves who had stolen his iPod.

My analysis of the implications of the data is here:

The US locks up a really large fraction of our population--higher than anyone else in the world--and a stunningly large fraction of black males. Intuitively, that calls the model of a few super violent people into question. At any given time, I think something like 4% of black men are in prison or jail somewhere in the US. It sure seems like the super violent black men inclined to kill at the drop of a hat would manage to mostly be swept up by the time you'd locked up 4% of the black men in the whole population. This article has the 4% number, but it's consistent with other stuff I've read.

The idea that there is a subpopulation of people inclined to both murder and be murdered (by nature, nurture, environment and incentives, or what-have-you) is also supported, to some extent, by the pattern of murder victimization and conviction rates. If I recall correctly, blacks are something like 7.5 times as likely to commit murder as whites, and something like 6 times as likely to be murdered as whites. That broadly fits the idea of a definable group where murder is in general more common, for whatever reason--more poverty, more drugs, worse law enforcement, worse mental health services, poorer impulse control, lower IQ, more violent culture, more gangs, whatever.

I have been exploring for a little for any high quality articles or blog posts on
this kind of house . Exploring in Yahoo I ultimately stumbled upon this website.
Reading this information So i'm satisfied to express that I've an incredibly good uncanny feeling I found out
just what I needed. I most surely will make certain to don?
t put out of your mind this site and give it a look regularly.

Exploring in Yahoo I ultimately stumbled upon this website.

Looks like "Exploring in Yahoo" makes you drunk.

That message is comment spam. The payload is the link in the name.

Comments for this post are closed