Violence is Less Violent…
Steve Chapman at Real Clear Politics has a post today about the violence in Chicago. It’s been getting a lot of play these days, since there have been several publicized shootings in the last couple months. But does the reality actually prove the rhetoric?
Right now, the city is getting national as well as local attention for outbreaks of bloodshed, which reinforce its reputation as the murder capital of America. In terms of total homicides, it may be. But that figure fails to account for population.
In the overall rate of violent crime, Chicago ranks 19th — slightly worse than Minneapolis and better than Kansas City, Indianapolis and Nashville. It has half as much violent crime, per capita, as Detroit or Oakland, Calif.
Even when it comes to homicide, Chicago is enjoying, relatively speaking, a golden age. In 1992, it had 943 murders — 2.6 per day. Last year, it had 415 — 1.1 per day. Two decades ago, such progress was the stuff of dreams.
In general, overall violence rates have been falling in the US in the last decade. Even the worst neighborhoods are getting less dangerous. So why is there so much news on violence, especially violence related to guns? Well, maybe the answer is that the media sees gun violence of any kinds as news that attracts attention, so it’s being reported more. Frankly, I’ve always been surprised at how little inner-city violence gets reported, and maybe paying more attention to it will make people more aware of it so it can be lessened even more.
What seems to be more prevalent, though, is how this violence correlates with other things that are going on, and that’s where I wander off the talking points. Those on the right would probably point out that Chicago is still highly violent, and that’s only because of its strict gun policies. But the data show that violence is decreasing. Ah, so that must be because overall gun ownership is going up in the nation… Um, probably not. Meanwhile, those on the left would say that increased violence in Chicago means we need to further restrict guns… except violence is actually going down…
And herein lies the difference between correlation and causation. As we look at all the data around gun ownership, violence, and the like, we often take data and tie it together to prove a point, not realizing that it doesn’t come close to proving a point. Just because two or three vectors are going in a direction one likes does not mean that one causes the other two, or that they’re even totally related. This is the problem with data science. You can state facts and then try to find the links, but not all facts are inextricably linked. I’m sure that there’s some type of link between gun ownership, violence per capita, and the like, but stating one doesn’t mean you’ve stated the right one.
So I’m still leery of a bunch of facts followed by a, “so there!” statement. Hopefully we continue to see a decrease in violence and (at least for me) an increase in gun ownership, but don’t assume one causes the other.