Yesterday marked the beginning of hurricane season, but the first big piece of hurricane news isn't about a looming storm. Instead, it's about how severe hurricanes with female names kill more people, on average, than hurricanes with males names. And the reason they're so deadly has more to do with the way society perceives women than the strength of the storms themselves.

"Our experiments suggest that a storm with a feminine name is seen as less threatening and risky than one with a more masculine name," said Sharon Shavitt, a marketing researcher at the University of Illinois at Urbana-Champaign, in an email to The Verge. Her study, published today in Proceedings of the National Academy of Sciences, shows that people are less motivated to evacuate when a serious storm is named "Victoria" than when the same storm is named "Victor."

To assess the effect of a storm's "gender" on its death toll, Shavitt and her team used data from the past six decades. Because the World Meteorological Society selects hurricane appellations from six pre-established lists of alphabetically sorted names with alternating genders, the naming system is free of gender bias. But to make sure the analysis wouldn't be tainted by outliers, the researchers excluded two particularly lethal hurricanes from the analysis: 2005's Hurricane Katrina and 1957's Hurricane Audrey.

The researchers found that although there was no difference between fatalities caused by small hurricanes with male names compared to those with female names, there was a substantial difference when the storms grew in intensity. On average, the researchers write in the study, a severe hurricane with a masculine name causes 15 deaths, whereas a strong hurricane with a feminine name causes 42 deaths.

But the researchers also needed to verify people's threat perception, so they asked a group of 346 participants to imagine being in a certain hurricane's path. As expected, participants thought that hurricanes with names like Arthur would be more intense than those named Bertha or Dolly — all names from the official 2014 Atlantic Hurricane list. And a second experiment, in which the researchers compared participants' perception of Hurricane Alexander with their perception of Hurricane Alexandra, confirmed that participants think storms with male names are more dangerous. "These kinds of implicit biases routinely affect the way actual men and women are judged in society," Shavitt said, "but now it appears that these gender stereotypes can have deadly consequences."

Hurricanes kill more than 200 people in the US each year, so Shavitt hopes this study will make people less likely to judge the riskiness of future storms based on the gendered meanings of their names. But regardless of the study's impact, it might still make sense to "move away from human names," she said. The problem with that idea, however, is that any naming system can be tainted by cultural or gender biases. If you decide to name three storms "delta," "lion," and "begonia," Hurricane Lion is probably going to get more people to move out of its way. That's why labels should be pre-tested beforehand, Shavitt said, if only "to make sure the meanings associated with them are appropriate."

Update June 2nd, 4:35PM: As Ed Yong points out, there is one factor that might temper the effect found by the researchers: hurricanes all had female names before 1979. It's therefore possible that an imbalance in the data could have caused the researchers to conclude that female hurricanes are significantly deadlier than male hurricanes. This is especially important because hurricanes kill less people now than they did 60 years ago.