When Staying Cool Seems Better Than Being Bad

By Shankar Vedantam
Washington Post Staff Writer
Monday, August 7, 2006

So the bad news is that it is hot and sticky and muggy. Your skin makes tearing sounds when you get up from a plastic chair. On the Metro, you start to tell people apart by how they smell.

The good news? You just might be safer than when it was nice and balmy.

A curious aspect of high temperature is that while crime and aggression rise with the heat, beyond a certain point, you start to see less violent crime.

No one knows exactly why baking heat prompts badfellas to turn meek. We can speculate, of course. When it gets to 100 degrees in the shade, who can wear a hooded sweatshirt? Perhaps a .22 revolver feels icky-sticky.

Nor are the criminally minded the only ones affected when the summer starts to swelter. As we weigh whether to slug the guy who spills beer on us in a bar, at a certain temperature we apparently conclude there is something to be said for having an ice-cold beverage trickle down our torsos.

It wasn't always this way. Theories about the connection between temperature and aggression have been around for centuries. The Belgian social statistician Adolphe Quetelet concluded in the 19th century that people in hotter countries were more violent than people in colder countries.

The English language and the law have long assumed a linear relationship between heat and crime. Killing someone in "cold blood" is seen as worse than murders committed in the "heat" of passion, because we know that being hot or hot-blooded makes people lose control. People often say that when provoked their blood boils.

All this heat and bloodlust makes intuitive sense, but the relationship apparently breaks down when the temperature gets too hot.

When scientists tracked the connection between the temperature and 911 calls for violent assault in Minneapolis and Dallas, they found a curvilinear relationship -- crime rose up to a point, then fell -- making the shape of an inverted "U."

Is that why you see less crime at 2 o'clock in the afternoon than at 11 o'clock at night? Not quite. It is important to remember that temperature is not the only factor that affects fluctuations in the crime rate. The time of day, day of the week, population density, even whether a major holiday is underway, all have effects on crime.

The temperature effect is overlaid on all these other factors, said Ellen G. Cohn, a criminologist at Florida International University and a former Washington area resident who is familiar with our annual summer swamp-fest.

Air-conditioning plays a role, too. If you remove the effects of sky-high heat in homes and office buildings, you bring people back into the range where they may be prone to slug each other. (Memo to Chief Ramsey: Keep an eye on those K Street lobbyists.)

So does alcohol. If people drink a lot of beer because it is super hot, you may not see the decrease in crime at very high temperatures because alcohol raises the risk of assaults.

There are times of the day when the inverted "U" is more curvy, and times when it is nearly straight. Cohn said the curious relationship between heat and assaults is most pronounced in the evenings, say between 6 and 9 p.m. If you compare 911 calls during that time on balmy, hot and very hot days, violent crime seems to rise at first, then fall when it gets very hot.

The theory has its critics. University of Michigan psychologist Brad J. Bushman says the relationship between hot weather and crime is linear -- as it gets hotter, people get more aggressive, period.

But laboratory experiments suggest Cohn may be onto something. When you get volunteers into a room and ask them to administer a painful shock to someone in response to an insult, the hotter you make the room, the stronger the shock that people are willing to administer. But when the room gets too hot, you start to see the same curious effect -- people administer weaker shocks as they become very uncomfortable. (The experiments are set up so people think they are administering a shock to someone; in reality, no one is hurt.)

At a certain point, people prefer to leave rather than fight, said psychologist Paul Bell at Colorado State University.

When Ronald Reagan was renominated for president at the Republican Convention in Dallas in 1984, an array of protesters showed up. People thought the stage was set for violent riots similar to those in the 1960s, when hot summers were thought to be partly responsible.

But then the temperature in Dallas hit 110 degrees. Many people simply packed up and left.

Shankar Vedantam will be online to discuss this story and his weekly series about the sociology and psychology behind the news at 2 p.m. Go tohttp://washingtonpost.com/science.

View all comments that have been posted about this article.

© 2006 The Washington Post Company