With the heat wave continuing its stranglehold over the U.S., Wired magazine takes a timely look at the literature linking heat to escalated violence. The most detailed study is apparently one conducted by psychologists Ellen Cohn and James Rotton of Florida State University, who looked at violent crime over a two-year period in Minneapolis, Minnesota. The study noted that violent crime rises with temperature -- but only up to a point.

Around 80 degrees Fahrenheit, assault rates started to fall, a trend that dovetailed with a hypothetical explanation for heat-induced violence in which being uncomfortable provokes competing tendencies of both aggression and escape. At low to moderate levels of discomfort, people lash out, but at high levels they just want to flee.

So once the heat escalates to a certain point (as in, current temperatures), violence drops, perhaps because heat-induced lethargy sets in. But as plausible as that hypothesis seems, other psychologists, such as Craig Anderson at Iowa State University, contested those results for not including time of day effects. Anderson was able to  produce a linear relationship between heat and violence, with assault rates peaking at the highest temperatures. Wired summarizes the physiological findings behind this.

In hot weather, the body exhibits changes — increased heart rate, blood circulation and sweating, and metabolic changes — associated with sympathetic nervous system activity, which in turn is linked to fight-or-flight responses. Hot weather also increases testosterone production, tilting that equation towards fight.

In a blog for the Boston Globe on "Crime and Punishment," James Alan Fox provides another reason for our deeply-held belief that crime and temperature are linked:

Rates of violent crime are higher in our nation’s warmer states, although there are many other factors that contribute to the “Southerness Effect.” Countries situated closer to the equator also tend to be global leaders in criminality, although here too many other social, economic and political factors are partially to blame.

Nonetheless, Fox himself looks at daily counts of violent crime along with daily temperature data for Columbus, Ohio, between January 1, 2007 and December 31, 2007, and notes the same relationship between heat and crime. His findings are particularly telling because they show that "temperature has some effect on violence in the home, but a much stronger impact on violence in outdoor or commercial settings." But he concludes, as did Cohn and Rotton at Florida State, that violence "tends to decline when temperatures reach the 90s," both inside and outside the home. The heat wave, therefore, may have has this silver lining: it's actually chilling violent crime, which peaks in the 80 degree range.