Does high heat equals high crime? It's a lot more correlation than causation, but the age old Chicago logic of hot summers being violent ones does hold up under scrutiny. The higher the temperature, the worse the crime. The question is, why does high crime occur on hot days? What is the mechanism involved in this relationship?
Navy Pier and Grant Park weren't the only Chicago places with fireworks popping through the night. Chicago's South and West Sides had busy and bloody evenings, enjoying a firework symphony with gunshot accompaniment. It was the most violent 4th of July in the last 6 years, continuing a disturbing crime trend that has sent Chicago's homicide rates to the pages of the New York Times and both Forbes and Time Magazine. But Chicago's summer has also broken records in a much more neutral statistic: The weather. This week alone has seen 3 straight days of 100+ degree weather. Today might be the hottest day on record since the turn of the century, and it would set us up for an entire week in the 100s. Inquisitive minds might wonder if there is a connection between the two factors of temperature and shootings. Let's take a look and see if a hypothesis holds up.
|Date||# Wounded||# Killed||Temperature|
# Wounded from https://data.cityofchicago.org/Public-Safety/Crimes-2001-to-present/ijzp-q8t2
# Killed from http://homicides.redeyechicago.com/
Temperature from http://weathersource.com/past-weather/weather-history-reports/free)
Just looking at the data, it seems pretty clear to me that there is some correlation between the shootings and the temperature. In 2009, with a 69 degree Independence Day, there were a total of only 11 shootings. The hotter it got, the more shootings there were. 2012 saw 21 total with 92 degrees. 10 degrees hotter gets us to 2012 with 5 more incidents.
For those interested in the rudimentary statistics of it all, we can run an admittedly too-small-sample-sized Pearson's R Correlation test to figure out if there is actually a trend to the numbers. Yes, the sample size is small. Yes, there are other tests to use. Yes, I am enamored with statistics although my understanding of them often borders on the sophistic. But even so, when we run the test, we find that:
Correlation = .88
(A more extensive study of violence would want to look at every single temperature entry for every day since, say, 2002. You would then want to run the correlation test for every value of total shootings and high/average/low temperature. I suspect you would still find a >.4 correlation, but probably not one so strong as .88.)
If, and it's a big if, this were an appropriately sized sample (which it isn't in this case), that would be a gigantic correlation that is often unseen in social sciences. Sadly, our data pool is a bit small, but it does confirm our eyeballing of the data. Simply put, there is a correlation, a fairly strong one, between temperature and shootings on the 4th of July.
Given that this is the case, why do Garry McCarthy and various other criminologically inclined public figures claim that high temperatures do not lead to high crime? Are they just ignorant? Or is there something more nuanced to their claims?
It turns out that temperature is not the only factor at play in crime, although on the 4th of July it tends to overwhelm other parts of the equation. When the temperature rises, more people go outside. This is especially true if, like most impoverished minorities in Chicago, you do not have air conditioning in your oven-like apartment. Between breezes, beaches, pools, fire hydrants, cooling stations, etc. there is just more relief from the heat outside than there is inside.
This is particularly pronounced on July 4th, when people are inclined to go outside regardless of the weather. Parades, barbecues, block parties, and gatherings in the park are the events of the day, and this necessarily brings people into close proximity with one another. Add in sweltering temperatures and extant conflicts and disputes and you have a particularly explosive situation, especially once you consider that the parties involved are gang members, not the average working class guy with a nice day off.
Does a heat wave make a crime wave? Yes and no. It gets people outside, it keeps people outside, and it puts them in a sweaty and irritable mood where they are disposed to conflict. But unless the conflict is already present, the heat is unlikely to create it. The same population that commits crime in the cold does so in the heat. Warm weather doesn't turn UChicago students into a drove of murderous academics. It just worsens the situation in already distressed communities.
The real interesting bit of this comes when you consider its implications for possible interventions. In a rather Levittian (Steven Levitt, that is) turn of thinking, if you were to give people air conditioning units, or provide them access to air conditioned building with activities, you might lower crime. This would keep people off the hot streets and otherwise engaged. Similarly, summer programs and jobs would also divert young people away from the antisocial street alternatives on a hot summer day. These are only rough sketches of ideas, but all of them center around the notion that high temperature amplifies criminal conditions and worsens violent crime rates. The air conditioning idea, or at least the provision of air conditioned havens, is also fairly affordable and decidedly concrete. Obviously, the quick-fix, one-page solution is never enough, but it is definitely something that can be incorporated into extant strategies.