How Random Are All Homicides?

One of the more searched for items on my blog is:  How Dangerous Is Compton? 

I don't think homicides are all that random.

Shankar Vedantam from NPR tweeted some research about how homicides tend to be concentrated within certain networks.

You know, we might be missing the wood for the trees, Steve. So take Chicago, for example, in the example you just gave about the ZIP codes. If you visit the website of a newspaper, like The Chicago Tribune, it will tell you that you have a high risk of becoming a victim of violent crime if you live in a neighborhood such as Washington Park or Fuller Park. But not everyone in these neighborhoods is actually equally at risk for becoming a victim of violent crime. I spoke with Andrew Papachristos. He's a sociologist at Yale, and along with Christopher Wildeman, he found the real risk doesn't lie at the level of neighborhoods, but at the level of a network with in the neighborhood.

Gun violence is much more like a blood-borne pathogen. It tends to be very specific behaviors - risky behaviors - that put you in these networks. And in some ways, it becomes much more like the spread of diseases through needle sharing or unprotected sex, rather than catching a bullet from somebody sneezing.

The LA Times Homicide Blog's mission is not so much to make those connections but is there to put out the names, dates, times, and places.

Reading it is all at once sad, engulfing, enraging. Sad because of what has happened.  Engulfing because of the stories behind what happened.  Enraging because we often don't know why what happened happened, and were left to thinking about the big, basic questions of why.

They report victims, location, age, race, a blurb about the incident, perhaps some background information about a victim.  Occasionally, they'll come up with blog entries about trends in a city.

Recently, they noted the "rising" homicide rate in Santa Clarita from 2 in 2011 to 6 this year.

The headline reads "Quiet Santa Clarita adjusts to recent jump in violence" as if that increase over a 62 square mile area with over 213,000 residents was one place full of carefully manicured soccer fields, acreage for people who choose to live that life, wide streets, was basically devolving into one big biker's bar.

To their credit, the writer did note that most of the cases were based on familial/intimate domestic disputes. 

However, a handful of the commenters, still wrote in search of deeper societal and/or demographic causes as if the homicides were "random" and wanted to identify those root causes to theoretically root out homicide.

I think those commenters represent what seems to be a common way of thinking about how crime happens:  it happens more in certain locations, with certain peoples of a certain age, sex, race, and it is either random or because people are in a gang.  Sometimes people conflate those factors of age, sex, race, or a location with being a cause for being murdered (i.e., you're black and in Compton, of course you're going to get shot!).

I think that if people learned more about the importance of networks in our crime discourse, those factors might matter less.  In the absence of information, people would attempt to look at a person's network rather than their age, sex, race, location before voicing a judgement.