Her-icanes Not More Deadly; Insurance Analysts Refute University Research

June 16, 2014

A recent study of hurricane names reveals some interesting information about gender stereotyping but doesn’t support the conclusion that storms named after women are more deadly than those named after men, two insurance analysts say.

In fact, Alan Zimmermann and William Wilt of Assured Research found that the mean number of deaths associated with hurricanes with feminine names is 17, compared with a mean of 15 for hurricanes named after men that have occurred since 1979.

Those numbers are pretty close to each other but very different from the mean number of deaths for pre-1979 hurricanes of 27.

Zimmermann and Wilt published their analysis of hurricane names last week, reacting to the late-May publication of the research report from a team of academics at the University of Illinois at Urbana-Champaign, which received a lot of media attention.

Here’s what the authors of the University study, “Female hurricanes are deadlier than male hurricanes,” say in an abstract of their report:

“We use more than six decades of death rates from U.S. hurricanes to show that feminine-named hurricanes cause significantly more deaths than do masculine-named hurricanes. Laboratory experiments indicate that this is because hurricane names lead to gender-based expectations about severity and this, in turn, guides respondents’ preparedness to take protective action. This finding indicates an unfortunate and unintended consequence of the gendered naming of hurricanes, with important implications for policymakers, media practitioners and the general public concerning hurricane communication and preparedness.”

An announcement from the university describes the laboratory experiment, noting that “people who were asked to imagine being in the path of ‘Hurricane Alexandra’ (or ‘Christina’ or ‘Victoria’) rated the storm as less risky and intense compared to those asked to imagine being in the path of ‘Hurricane Alexander’ (or ‘Christopher’ or ‘Victor’).”

Wilt and Zimmermann agree that this is “an insightful analysis into gender stereotyping,” but they argue that the conclusion that female hurricanes are more deadly—or more precisely that they are more deadly because people take less precautions when they hear warnings about a storm carrying a woman’s name—are unfounded.

“There has been a secular decline in hurricane deaths over time,” the two insurance analysts say, referencing the difference between pre-1979 and post-1979 average deaths per storm—27 in the earlier period versus 16 in the later one.

Prior to 1979, most hurricanes had female names attached to them. In fact, between 1953 and 1979, all hurricanes had female names, Wilt and Zimmermann note.

According to the paper, an alternating male-female naming system was adopted in the late 1970s because of increased societal awareness of sexism.

Reviewing the six decades of storms from 1950-2012—the study period analyzed by the university researchers—the duo from Assured Research reveal 24 deaths, on average, for hurricanes named after women, and an average of 14 for storms named after men.

For that period, the University’s announcement says that its research team, led by Kiju Jung, a doctoral student in marketing, found that “changing a severe hurricane’s name from the masculine ‘Charley’ to the feminine ‘Eloise’ could nearly triple its death toll.”

A simple count of the number of deaths for feminine-named storms and masculine-named storms—1,473 vs. 427—in the 1950-2012 timeframe backs up that statement. But there were 62 hurricanes named after women and only 30 after men, the figures in the Assured Research analysis reveal.

By not accounting for the secular decline in hurricane deaths—which the insurance analysts attribute to better quality forecasting and preparedness on the part of the National Hurricane Service and local agencies—together with the change in naming convention, the University researchers “have overweighted the deaths associated with female-named storms and trumpeted, in our opinion, spurious conclusions,” Zimmermann and Wilt say.

The Assured Research report also includes comparative mean losses associated with hurricanes in the pre-1979, post-1979 and full 62-year period, showing that dollars of damage are higher for the masculine-named storms, on average.

The names of this year’s Atlantic storms, alternating between male and female names, will start with Arthur, Bertha, Cristobal, Dolly, Edouard, Fay and Gonzalo, according to the National Hurricane Center website.

Coming Soon:

More discussion of possible flaws in data analytics by casualty actuary Ira Robbin in his three-part article series,”Predictive Modeling Pitfalls: When To Be Cautious.”