In all the March's between 2004 and 2011, Chicago only saw an average of 122 shootings. This March 2012, we experienced 162. This was a statistically significant increase that suggests a broader change in Chicago crime; something has happened to increase the violence.
A few posts back, I promised an in-depth look at Chicago shootings in March. Homicides were definitely up, but the questioned remained as to whether shootings were also on the rise. 7 days have now passed since the month’s end, so the statistics are now live for public consumption. The numbers, and the conclusions we can draw from them, bode ill for the summer; March shootings were way up from the past few years.
You won't find these reports in the news (I certainly haven't). Of course, for most people, especially readers of this blog, the rise in shootings won't directly affect safety and well being. But in our lovely city of districts, we must know our home! Moreover, for the Warriors amongst us, we must know the dangers that our town faces. Here is a snapshot of the reality in this Chicago March Madness.
Statistical tests are admittedly not the best tool we have at our disposal in analyzing crime. That said, they can give us an acceptable, general sense of a problem. I'm no math or statistics major, but numbers remain interesting and important. They offer a clearly quantifiable portrait of a qualitative problem. So please, as usual, bear with my simplistic statistical analysis and look at the bigger picture that it suggests.
March 2012 saw 162 aggravated batteries with a handgun or firearm. The question stands: Is 162 a lot of shootings for a March? Did something change in Chicago to cause a spike? To figure this out, I took a look at the number of March shootings from the past decade. Here is what I found:
|Year||# of Shootings|
The first thing to notice is that we had a huge plummet in shootings between 2003 and 2004. There are a lot of reasons for that which we don't need to get into right now. For this analysis, it just means that we should not view 2001-2003 as part of our current 'pattern' of violence. Those dates were remnants of an earlier era in Chicago violence, the leftover shooters and shootings from the wild 1990s. In considering the data, we are going to only look beyond 2004, after the city supposedly calmed down (for all those reasons which we can talk about another day).
That leaves us with the following dataset:
|Year||# of Shootings|
We need to see if the 2012 shooting number, 162, is an outlier given the other data. That is to say, if 2004-2011 represented a pattern of violence, was that pattern still true in 2012? To do so, I am going to conduct a one-tailed z-test. Now, before all my mathematically oriented readers pounce, I am going to re-acknowledge that this is a crude mechanism. It assumes a normal distribution of crime (not a safe assumption!), and it ignores a lot of variables. That said, it is a quick and dirty way of quantifying a problem that is very much qualitative.
The population mean (average) between 2004 and 2011 is 121.88. The population standard deviation is 18.77 (again, we are assuming a normal distribution). To conduct our z-test, we subtract the mean from the tested value (162), and then divide by the standard deviation. This provides us with a sigma score of 2.137. How do we interpret the sigma? It means that our value of 162 shootings is 2.137 standard deviations above the average. We can then check this 2.137 value on a normal distribution to find the area under the normal distribution curve. We find that area to be just over 98. Based on that, here is what we conclude.
If the 2004-2011 dataset mean of 121.88 shootings per March was still true in 2012, we would only expect to see 162 shootings in 1.63% of all cases. 98% of the time, we would never see 162 shootings. If, however, there was a NEW mean for our March shootings (that is to say, if shootings just went up), we would expect to see 162 attacks. For this year, shootings were way up.
Based just on this crude test, we would reason that something changed in Chicago this month to cause shootings to increase so dramatically. March 2004-March 2011 all saw between 100 and 130 shootings, with the average sitting at 122. 162 is an outrageous increase from this dataset, representing a major shift in Chicago for this current March 2012. That leaves us with just one question: What changed?
To some extent, temperature is to blame. It was record-shatteringly hot in Chicago this last month, which might account for the record-shattering number of shootings. Like most social scientists, however, I am leery of offering only one answer. Other factors are likely at play. Without delving too much into this topic, I hypothesize that recent Chicago Police Department efforts at targeted area policing are, in large part, to blame for this increase. By saturating certain areas with police officers, the CPD has displaced extant drug markets and gang/clique corners. Once displaced, the criminals do not just give up their guns and narcotics. They pick up shop and move somewhere else, probably to a neighborhood that already has its own established territory and markets. Increased competition in the criminal underworld almost invariably leads to violence.
No matter the explanation, it is clear that March 2012 saw a statistically significant increase in shootings from the past 8 years. Some other city-wide change, or neighborhood specific change, had to accompany this uptick in shootings. For now, it is enough to know that violence is on the rise in our city.