I was 10 minutes into my interview with Ryan Harp before he showed me the most shocking piece of data I have ever had to emotionally process. For three years, Harp has been investigating the connection between climate and crime. At one point, Harp had to call the FBI– he was asking for some statistics about crime that the FBI collects from most police departments in the United States. Ten minutes into our chat, he was holding up a jewel case containing that data.
Hold up, I thought, and then vocalized. “Are you telling me that the FBI sent you three decades worth of detailed statistics about violent crime… on a CD?”
Did you know that the FBI employs over 35,000 people and has an annual budget of $8.7 billion? It has its headquarters in Washington, D.C. and has 56 field offices throughout the 50 states. Did you also know that the FBI still sends out 3GB of data on CDs?
Don’t worry, I stayed professional while processing this bombshell. Harp went on to explain why he needed that CD’s worth of data. Harp is a graduate student in the Cooperative Institute for Research in Environmental Sciences (CIRES), and the Department of Atmospheric and Oceanic Sciences (ATOC) at CU Boulder. He’s currently working in the lab of Kristopher B. Karnauskas, studying the connection between local weather and crime. When he was sent this data, he was trying to correlate two very different datasets. The first dataset was from the National Oceanic and Atmospheric Administration (NOAA), and described the local weather all over the United States from 1979 to the present (sensibly delivered in a downloadable digital format). The second dataset described violent crime from the same time period (the CD).
Scientists have long known that crime rates increase during the summer and with warmer temperatures. There are two different theories about why this might be true. The first is that warm weather makes people more aggressive:
“When it’s really hot outside, your body gets physically agitated like you’re experiencing heat stress,” said Harp. “You’re just irritated and a little more likely to do something that you might not do. You’re cranky.”
The second theory, called the “routine activities” theory, is built around the idea that a crime needs three elements: a motivated offender (someone who wants to commit a crime), a target, and the lack of an obstacle to the crime, like a police officer.
“During pleasant weather, people are more likely to leave their house, go to restaurants or parks,” said Harp. “You’re just increasing the likelihood that those three things are going to come together.” This, in turn, makes crime more likely.
Harp was interested in two questions: first, which of these two hypotheses is correct, and can we prove it? And second: will a warming climate lead to more violent crime incidents?
This is where the two databases described above come in. Harp first split the United States into five different regions: the Northeast, Southeast, South Central, West, and Midwest. Next he categorized all the data points for climate and violent crime he had by region. He then compared temperature and crime to ask: if the temperature in a given month is higher than average, does violent crime also increase?
The answer is yes. In a publication for the journal GeoHealth, Harp showed that when the temperature is above average for any particular month, violent crime is also higher. This is true in all five regions, and the correlation is especially powerful in the winter months. This implies that it’s not just warm weather making people cranky—higher than usual temperatures in the winter are still not usually very high.
“This was our first piece of evidence that the second theory—the routine activities theory—was the stronger driver,” said Harp. “So when it’s warmer in the winter, people are out interacting more, and that leads to more crime.”
But Harp has even more evidence for the second hypothesis—that unusually warm weather brings perpetrator, target, and motive into closer proximity than usual. As well as looking at violent crime, he also looked at property crime. If the first hypothesis was true—that warm weather makes people cranky—then you wouldn’t expect property crime to increase with higher than average temperatures. When people get aggro in the heat, it’s usually because someone’s actions are more annoying than usual. But you wouldn’t expect a thief to be more likely to rob a house just because it’s hot… right?
But sure enough, property crime does increase with higher than average temperatures. This suggests that all crime increases when temperatures are higher than normal, not just a crime that revolves around people getting cranky. This lends more support to the second hypothesis: that warmer temperatures bring people who would commit crime out of their warm, cozy houses.
Harp’s conclusions are important for multiple reasons. This report accounts for seasonality—meaning Harp split up the results by month and showed that the correlation between warmer temperatures and a rise in crime is very high. Reports that don’t split up the data by month just show that crime and temperature are both seasonal, which is true but not a novel finding. Interestingly, there is still a seasonal trend—the correlation between an unusually high temperature and a rise in violent crime is stronger during winter months.
Harp also used relatively simple statistics, which he thinks speaks to the power of the trend that they’re finding.
“We created these regions, which allowed us to get past the random fluctuations from year to year and pull out the climate signal a lot more easily,” said Harp. “We were able to use really basic statistics and were able to detect the relationship without using a lot of complicated models.”
Harp, a climate scientist, thinks it’s important that this story is framed in the context of how climate could shape the health of future generations. Many people have researched how a warming climate could influence the spread of disease, but Harp thinks there could be health consequences to global warming we haven’t predicted. “When a violent crime occurs, the victim experiences physical and mental trauma, and their health is very negatively impacted,” said Harp. “We’re adding this in to the overall climate and health focus.”
Harp is interested in the next big question: will a warming climate lead to more violent crime incidents? This seems like an easy connection to make based on the results of this study, but it needs to be directly investigated before any conclusions are drawn.
A warming climate could have many impacts on human society and health. Questions about the future increasingly keep scientists, myself included, up at night. Will disease spread faster as disease vectors migrate into new regions of the world? Will rising oceans displace people from coastlines? Harp adds two more questions to my nightly worried thoughts: will violent crime increase as temperatures rise? And—most importantly—will the FBI ever learn about flash drives?
By Alison Gilchrist