There is one school of thought that does not buy any of these explanations. Instead, it points towards a much simpler idea - the South is warmer than the rest of the country. Could it be that hot weather can lead people to anger easily, become violent quickly, and more readily kill each other? Supporters of the heat hypothesis think so. The heat hypothesis is a simple yet powerful idea: the more uncomfortably hot the temperature, the more likely people become aggressive.
There's a bit more to the heat hypothesis than that. It's not simply aggressive behavior that is increased, but affective aggressive behavior. Affective aggression is the sort that happens in the heat of the moment, so to speak. The primary result of affective aggression is to harm another person; other goals are secondary. For example, harming someone during a bank robbery would not be affective aggression, whereas getting into a bar room brawl after discovering a lying, dirty, cheating bastard would.
Getting back to the South, there were some pretty hot results from studies done on the temperature vs. crime rates. Initial studies between the North and South showed a positive correlation between heat and violent crime rates. Of course, there could be many other factors contributing to this effect, such as the inherent differences between cities and cultures of the two areas. Further studies were conducted not between geographic locations but over time. By studying the same cities over a few year time span, researchers were able to account for most outside factors that may come into play from geographic location. These studies show a positive correlation between violent crimes and unpleasant temperatures.
While controlling as many other factors as possible, the heat hypothesis does seem to show that violent crime rates increase with the heat. However, this is not to say that scorching weather causes all types of crime to occur more often. There was little or no correlation found between heat and occurrences of rape, robbery, or property crime. Since violence is (typically) not the primary goal of these crimes, the weather had little effect on them. Besides, it's just bad form to slap around someone from whose house you're already stealing.
There have been some smaller scale studies run as well. Laboratory tests seem to support these results, depending on the care taken in the study (some studies were not that effective since participants suspected something due to kerosene lamps being in the room). In one particularly amusing field study, an experimenter purposely stopped his car in front of green lights on days of varying temperature and recorded the number of honks he received. Besides showing that psychologists will be jerks to further their studies, this particular experiment further supported the heat hypothesis - the hotter the day, the more honks he received. However, this readiness of horn use was decreased if the car had air conditioning.
Of course, the heat hypothesis is just that - a hypothesis. It could hardly be used in a court of law as a defense - "I swear, I did not mean to kill him! It was 110 degrees that day!" Still, the implications are scary, especially with so many reports on global warming being released. It's bad enough that my ice cream will not last as long outside; now I'll have to deal with people trying to hurt me when I don't share it.
Article suggested by Nora P.