Imagine you’re a politician with a chunk of money to spend on improving a local park. You have a choice – you could spend the money fighting petty crime or you could spend it dealing with deer overpopulation. Your risk analyst warns you that the risk of deer overpopulation to human health, property, and the environment exceeds the threat posed by crime. Where do you allocate most of your money – deer or crime?

At this point in the story, Joseph Arvai of the Michigan State University pauses. “Believe it or not, people get attacked by deer and can end up in the hospital,” he says.

Nevertheless, you’ll probably elect to spend more money fighting crime, Arvai said in a press briefing before a symposium titled “Numbers and Nerves: Affect and Meaning in Risk Information.” Crime evokes a more emotional response than deer attacks, he said, and emotion triumphs over numbers.

When Arvai tested this experimentally, he gave study participants a quantitative estimate of the risk posed by both problems. Not until he had ratcheted up the risk from deer to nearly twice that of crime did participants take the problem as seriously. Arvai tested other scenarios as well — avian flu or Anthrax bioterrorism vs. food poisoning, for example. The trend was always the same. To counteract that emotional first instinct, Arvai is designing programs that will crunch the numbers for politicians making difficult decisions and present the results in ways that will make more of an impact.

Think you’re immune to that emotional pull? Odds are you aren’t. Arvai and his colleagues tested scientists and other well-educated folk as well. “It didn’t matter how smart or educated — everyone was biased,” he said. “Including scientists.”