Hurricane with female names are more destructive than those with male ones

It used to be the case that all hurricane names from 1953 to 1978 were female because hurricanes were ‘unpredictable’ just like women (!) but that kind of sexist stereotyping became unacceptable and they decided to alternate between male and female names.

The report that found that hurricanes with female names cause more death and destruction received a lot of attention in the media. Not only that, it appears that more feminine names cause more damage than less feminine names. How they assessed the degree of femininity of names was not specified (maybe they did surveys?) but the example given was of Laura versus Bertha. The researchers needed this additional feature to establish their result because the number of hurricanes after 1979 was too small to give significant results and they needed more data. This worries me because while the gender of names can be quite unambiguous, the degree of femininity is not and could well vary by culture and geography, and this might well be a problem for the robustness of the conclusions.

The proffered reason for this correlation is that people unconsciously think of ‘female’ hurricanes as being kinder, gentler ones and are less likely to take precautions or evacuate.

My first reaction to this was that all hurricanes should be given scary names from made up gender-neutral fictional monster names and I was going to post some suggestions but Stephen Colbert beat me to it.

(This clip aired on June 4, 2014. To get suggestions on how to view clips of The Daily Show and The Colbert Report outside the US, please see this earlier post. If the videos autoplay, please see here for a diagnosis and possible solutions.)


  1. maudell says

    I read the study (it’s only 6 pages) and I’m not impressed. In my opinion, the statistical model is flawed. Most importantly, there are no controls in the observational part of the study. The experimental phase was done by using Amazon mechanical Turk, and it was a pretty small sample. It also appears to me that the authors have not justified their causal statement at all.

    I don’t think the hypothesis is that far-fetched, but this is a sloppy study and I think their conclusions should not be accepted. It’s frustrating how many poor studies are touted in the media as groundbreaking.

  2. cafink says

    I, too, am rather unimpressed with this study. As a friend pointed out, it used death tolls from 1950-2012, but the early part of that time period included only female hurricane names, and also had worse technology for predicting hurricane paths and warning people than we do today. That alone skews the numbers, and they don’t appear to have controlled for it.

Leave a Reply

Your email address will not be published. Required fields are marked *