Hurricanes with Female Names Kill More Deadly

Whether identifying changes in public policy that could do exponential good or revealing internal assumptions about the way the world works, mental models are the best construct I’ve ever come across to categorize, analyze, study, document, and reflect upon the universe.  Yesterday, The Washington Post published an example of cognitive bias when it detailed how how subconscious beliefs can have life or death consequences for people as small inputs change the way they behave.  While the math leaves a lot to be desired and is less than convincing, the follow-up experiments testing the hypothesis indicate people do treat male and female names differently.  This is not exactly a new phenomenon.  More than a decade ago, researchers found that even with identical credentials, resumes with “black sounding” names were far less likely to get an employment offer.  Words matter.  Names matter.  The brain makes split-second, often irrational, decisions based on internal prejudices, cultural constructs, and a host of other variables.

According to the new study published in the Proceedings of the National Academy of Sciences, researchers at Arizona State University and the University of Illinois ran the figures on hurricane deaths between 1950 and 2012.  They excluded Hurricanes Katrina and Audrey as outliers that had the potential to make the typical hurricane look worse than it was given the power of those two storms.  They concluded that even throwing out the deadliest hurricanes, when looking at the 47 remaining most destructive hurricanes, those given female names resulted in mean deaths of 45 people per storms, while those given male names resulted in mean deaths of 23 people per storm.

Hurricanes with Female Names Kill More Deadly

In other words, they believe that hurricanes with female names are killing, on average, nearly twice the number of people as preparation plans differ based on subconscious bias triggered by mere association; hurricane Samantha sounds less scary than Hurricane Kyle.  Given that hurricane names are chosen long before the weather system forms (here is the list up through 2017), and the fact the death rate among the genders is even more extreme as men and women respond differently to differently named hurricanes, they argue this could indicate that a simple paperwork change in the weather naming convention could save lives.

[mainbodyad]To test their hypothesis – one potential flaw in the model was that hurricanes only had female names up through 1979 – the researchers set up half-a-dozen laboratory experiments where they posed questions to test subjects.  The individual experiments had between 100 and 346 people in them and revealed the same pattern.  People answering questions about how they would prepare for a hurricane showed a lower likelihood of seeking shelter if the hurricane in the scenario had a female name.  They also predicted higher destruction for the hurricanes with male names.

Besides potential changes in the naming conventions of hurricanes, this is certainly going to be of interest to sociologists.  Is this a case of soft prejudice against men; a form of internalized misandry?  As Sharon Shavitt, one of the co-authors of the study, tells The Washington Post, “The stereotypes that underlie these judgments are subtle and not necessarily hostile toward women – they may involve viewing women as warmer and less aggressive than men.”  Is this a case of a soft prejudice against women; a form of internalized misogyny?  The “little-lady-that’s-cute-you’re-trying-to-fight-me” sentiment?  Both?  Neither?

Regardless of the sociological questions, if the conclusions can be replicated by other researchers in labs (the hurricane data itself is not conclusive but merely raises the question), it proves a very important point.  Namely, that each of us needs to be aware of the second and third order consequences of our decisions, or at the very least accept that we can’t possibly predict everything with even the most complicated models.  When the meteorological community switched to human names for tropical storms and hurricanes, the intention was to make them easier to remember and discuss.  Yet, that seemingly helpful shift might have ended up killing a lot of people due to quirks in the evolutionary wiring of the human brain.  Or, it might not.  But this is the fun of trying to constantly improve your understanding of the world, to fit it together on a framework of bigger ideas that interact with each other.

Reader Comments (17)

Comments are presented chronologically, with replies indented beneath the comments to which they respond.

Rob

June 3, 2014

Interesting study and I'll have to look into it further, but at first glance this screams of a correlation / causation problem. As someone who has lived most of their life on the east coast, including in a beach house, when we created a response plan for hurricanes the name of the storm never came up (you can argue that it is a subconscious bias and perhaps so, although not to a extent significant enough to alter our planning). It was always based on more quantitative factors such as how much damage the storm had done before reaching us, did it stay over land or go back over the water and increase power, will it be high or low tide when it is expected to hit, etc. I believe mental models are extremely important, although this specific example/study feels like a reach.

Joshua Kennon

June 3, 2014

Replying to Rob

I believe mental models are extremely important, although this specific example/study feels like a reach.

That's where I was due to the chronological problem (1950-2012 data, yet female names only beginning in 1979). I think it was intelligent of them to try to proactively address the correlation/causation question through the six laboratory experiments, which demonstrate something is likely happening that warrants further experimentation.

Personally, I'd want to see it replicated in a few other studies before I took it seriously enough to change the current naming convention but it sounds like exactly the sort of thing Dr. Cialdini talks about in his groundbreaking book as well as the reason for the success of Clotaire Rapaille, who is paid a lot of money to discover that subconscious signals can change human behavior in drastic ways (e.g., exponentially increase sales of very expensive Hummers by putting stars on the car to subconsciously trigger associations with military power). It also seems plausible based on some of Noam Chomsky's work about the way language influences the perceptions of a person in how they interpret and interact with the world; e.g,. two identical people, one English and one German, are going to think very differently because of how the language itself is structured.

It was always based on more quantitative factors such as how much damage the storm had done before reaching us

That raises an interesting question on its own. Do the results weaken as you go further up the East coast? Presumably, your response would be different than those who didn't have that data. Those in a state like Virginia or North Carolina would have had much better advance warning in many cases, allowing them time to adjust the plans based on what had already happened compared to a first-landfall area. The few hurricanes I experienced in years living in New Jersey were that way. The response was predicated on what had happened down South, which is why everyone hunkered down for Isabel, which was the first I ever lived through.

That would be my next experiment if this were my field of study; to isolate states like New Jersey and compare it to Florida. The problem is, it may be impossible to find an answer because the storms weaken as they get further up the coast and the same raw subconscious tendency might be there as part of the human condition so you'd get this paradox of sorts where people were just as likely to engage in the behavior yet it had no ill effects the further North you went due to the way the coast is structured as their rational mind kicked in and adjusted based on the damage reported on the news.

As a midwesterner, it's always fascinating to me to watch people on the coast nonchalantly go about their preparations when a hurricane is coming. If you grow up in tornado country, you learn that, with very little warning, the sky could open up and start coming for you. Even as a little kid, elementary school tornado drills were part of the routine. Grab your textbook, cover your head, line up in the hall, duck, cover, and pray.

I mean, this morning we woke up to discover that conditions have arisen that could make today a nightmare weather day, as has been all over the news for the past few hours. If we actually knew tornado paths ahead of time like those in hurricanes do, I think most of the folks in this part of the country would look like Captain Jack Sparrow. There'd be none of this, "We're gonna ride it out and see what happens." At least not in my neighborhood.

Marc KS

June 3, 2014

http://www.slate.com/blogs/future_tense/2014/06/03/are_hurricanes_named_after_women_more_dangerous_not_so_fast.html

Looks like someone wanting to make something out of nothing

Joshua Kennon

June 3, 2014

Thanks for the link; that was a good read.

Sometimes, I wonder what people are thinking when they analyze data, or if they just go on autopilot. For example, the opening objection in this story, quoted from National Geographic, is dealt with in six follow-up experiments. Whether or not the original premise was proven in the data (it's not terribly convincing to me), the result of the controlled experiments on test subjects indicates that there is some sort of cognitive bias occurring that warrants further investigation.

And yet, the most important part of the study, again, at least to me - the presence of this bias in the test subjects, which has nothing to do with the hurricane data for the past half century - is ignored. Why? Probably because of the stupidity of one section of the study itself, which tries to model the increase in fatality based on the femininity of the storm name, which is rubbish. The math simply isn't there to support it. (It reminds me of some of the wishful thinking I see in Wall Street valuation models where people think adding more complexity to an equation increases the certainty of a good outcome.)

Maybe it's that we live in a news snippet culture where nuance is lost. "This is possible" or "They raise an interesting question that needs to be further studied" becomes "So-and-so said this and here is why they're wrong and let me serve you advertisements as you listen to my righteous indignation."

I'm of the school this falls into the camp of, "Something is probably happening because people demonstrate the behavior in studies, but trying to quantify it is impossible so it goes on the 'too hard' pile or, to be safe, we can change the naming convention since it's not that much work."

Then again, like that article says, perhaps I'm just excited at the idea of the newscasters announcing the approach of Hurricane Skeletor.

Andrew

June 5, 2014

Replying to Joshua Kennon

I admit, I'd be terrified if I heard "Hurricane Thor is heading your way". I'd leave the potential area ASAP regardless of facts.

Would you treat it differently?

John Tate

June 3, 2014

I thought the Freakonomics guy had a section of his book devoted to the "black sounding names get less job interviews" study you linked to and concluded it had to do with the structure of the names and not inherent racism? Its been years since I read that book so I may not be remembering it correctly.

Joshua Kennon

June 3, 2014

Replying to John Tate

It's been awhile since I read any of the books but the last thing I remember looking at in-depth was that same year when they concluded in The Causes and Consequences of Distinctively Black Names, Working Paper 9938 that the names were a consequence, rather than cause of socioeconomic status. That is, the name doesn't seem to influence the person's life at all because they ended up exactly where you expected.

They later built on this in a podcast, saying:

"In other words, it’s not the name your parents give you; it’s the kind of parents you have in the first place. And different kinds of parents of course choose different kinds of names. So let’s say two similar families, both African American, each have a baby girl. One is called Molly – which, it happens, is one of the whitest girls’ names in America. And the other is called Latanya, which is a distinctively black name. Now, if, decades later, Molly becomes, let’s say, a professor at Harvard, and Latanya is just barely scraping by – well, the reason won’t be because Latanya’s parents named her Latanya"

So, in other words, the Latanyas of the world are less successful because they have a certain sociodemographic mix of parents that would name her Latanya in the first place which lends itself to be stuck in a cycle of poverty, etc.

It doesn't refuse, nor satisfactorily explain, like this one and this one, which even the podcast itself references. There is a very real statistical bias on the lower end of the skill spectrum where name matters.

Levitt reconciled this by saying (link to transcript):

I think that both could be right. There are ways to reconcile them. So let’s start with the audit studies. [snip] So in the audit studies what researchers do is take identical resumes and just change the first name so that one name is distinctively black and another name isn’t. And they send those out to employers and see whether there’s a callback. And what they find every time is that if you have a distinctively black name you’re less likely to get a callback. So how can that be reconciled with the fact that in our data, in real life data, how people actually lived, the names didn’t seem to matter? I think the answer comes in a couple different ways. The first is that just because you get a callback doesn’t mean that you’re likely to get a job. So to the extent that there are discriminatory employers out there and those discriminatory employers are using your name to figure out whether or not you’re black, then indeed the worst thing you could possibly do would be to show up for an interview if you are black with a white name and have wasted all day trundling downtown to do the interview for a discriminatory employer who’s not going to hire you anyway. That’s one possibility. The other possibility is that there are two different kinds of labor markets. There’s a sort of formal labor market that involves resumes and applying, and really hardly anybody gets jobs that way, that’s not the typical way people get jobs. And your black name might hurt you in that segment, but it might actually help you in other areas. So you could certainly imagine that within the black community having a distinctively black name would help you get along better with people, signal that you’re part of the community, and might work in your favor in all sorts of informal networks that aren’t captured in these audit data.

Richard Garand

June 4, 2014

Replying to Joshua Kennon

Steven Pinker's How The Mind Works has a great analysis of similar patterns in fashion. His explanation is that we start off with certain distinctive fashions among the poor, middle-class, and rich. The middle-class will avoid the "poor" fashions because they risk being mistaken for a different (lower) class, while trying to copy the "rich" fashions so that they may be mistaken for a different (higher) class. The rich want to stand out so they avoid things that the middle-class are wearing. Instead they imitate the "poor" fashions because there is very little risk that they will be mistaken for a different class, and in fact this sets them apart from the middle class who are afraid of those same fashions. Over time the cycle keeps repeating itself.

It's the kind of thing that is understandable, but even if everyone knows about it is unlikely to change because the basic motivations will always be there.

Engineer7006

June 4, 2014

Replying to Joshua Kennon

Paul Fussell's book "Class" goes into those various topics quite a bit, but is a bit dated since it came out in the 80's (for example some of the speach patterns and manner's of dress are outdated). It's a fairly comprehensive list of status markers for different classes in terms of speach, dress, etc and how various peple attempt to "jump class" by emulating those aspects of classes above them.

If you are exposed to those markers at an early age, its pretty obvious what many of those markers are and how they enable people to tell if you are "one of them." I would say that this is a bit less important today as there is so much new money instead of inheritied money, but I can remember stories from friends parents who made money are were spending it greatly in their later years and lacked said markers.

Joshua Kennon

June 4, 2014

Replying to Engineer7006

Great comment! I have that book in my library and have been meaning to read it. I put it off a couple of years ago because, like you say, a quick perusal showed it was very dated and no longer accurate, though certainly useful in an abstract sense. Maybe I'll go pull it out and add it to this week's list.

My objections, if I recall, are this: Throughout history the term "class", when referring to socioeconomic status, has largely referred to the relative access to a combination of social, economic, and political capital. It has its roots in biology and survival. In some societies, one was more important than the other (e.g., the merchant class was among the wealthiest in Japan yet had very little social capital because they were thought to not create anything), but for the most part, it was always tied to the animal. In times of war, it was the great fighters who ascended to dominate the class structure. In times of peace and prosperity, it was the great musicians and artists. In times of famine and struggle, it was the landowners who owned the fields that could help people survive. There are some exceptions, but generally this true; a sort of great, unifying thread throughout the eons across different contents, cultures, races, and political structures.

In a society like the United States, class is nothing like the author argues, and it hasn't been for going on a century. It's simply too dated. Kayne West, as much as it pains me to say it, is to mass of modern society what the Vanderbilts or Prince of Wales were to the society of old. The same goes for Sam Walton, etc. In the United States, if you were to line up Bill Gates, Warren Buffett, the founders of Google, Steve Jobs (assuming he was still alive), Angelina Jolie, the rest of Hollywood's "A" list, etc., on one side of the room, and the Kings, Queens, and Sultans of the world on the other, the former is going to have a lot more clout and be revered much more highly. They are, to put it bluntly, of a higher social class. To the American public, what Steve Jobs pulled off for the world can't be inherited. He was inherently superior to the typical person by virtue of his accomplishments and contribution.

Part of this, I suspect, is the spiritual roots of the country, which began with a rebellion and contains a general distaste for non-earned accomplishments. Few people have a problem with Mark Zuckerberg making billions in his 20's, yet there's something culturally unseemly about a person the same age inheriting a trust fund worth 1/1,000th the net worth and talking about it. We're an odd lot. It's partly what makes me love this country so much.

Engineer7006

June 4, 2014

Replying to Joshua Kennon

Joshua, I think that is a pretty fair assessment of Fussell's work (I personally wouldn't put it on the same level as an academic study) as to how it may no longer apply the modern upper classes. Economic class and social class are no longer conjoined (and your points about how the nouveau riche's children could become upper class socially are true).

I do think some of his assessments of the middle and lower classes are still valid and rather humorous (I recall a section bathroom decor and use of various cleaning agents). For those who have not read it, Fussell does not make a value judgement on the various classes; there is no glorification of the rich nor of the poor. Rather it can be a form of entertainment to try and discern people's backgrounds based off the various markers covered in the book, though may still be useful to some in the corporate sales profession.

preston nelson

June 4, 2014

Replying to Joshua Kennon

If I even think about calling my family my speech patterns change.

I went see if I could find that interview and found this skit by key & Peele instead searching for "code switching". You know it's true when the comedians are making jokes about it.

And it is apparently old news in linguistic circles. See "Code Swtiching" in Sociocultural Linguistics.

Joshua Kennon

June 4, 2014

Replying to preston nelson

This is so true about family, haha! In my case, it manifests as slowing my vowels and reflexively sentences like, "Listen, y'ain't gunna get anywhere by doin' it that way..." come out of my mouth. There are also quite a few more religious allusions and I begin using the idiomatic "have got" form for everything (e.g., "I've got to go to the store" instead of the correct "I need to go to the store.").

If this is happening between social classes, geography, and race, I wonder if it also explains "surfer dude" voice, "gay" voice, "sorority girl" voice, "preacher" voice, etc. Is it all signaling theory and code switching? Probably. I doubt physiology plays any role.

I'm printing the study in your link now and can't wait to read it. I get the feeling this is going to be my obsession for the next couple of weeks ... I need to understand it. I'll bet it's a subclass of signaling theory and tribalism; a way for tribes to identify other members.

... this is all I'm going to be able to think about now. I need to start a spreadsheet and start putting together a reading list. Thanks again for the study; it looks like a great start!

Preston Nelson

June 4, 2014

Replying to Joshua Kennon

I would say that code switching is pure signaling theory (conformity). And that voices are a conforming signal, some of which have been created in relatively more recent times, particularly "gay" voice. I mean why bother with subtle eye shifts as indicators of interest with a large (possibly unknown) population who may react with hostility when you can declare to everyone your preference at one time?

I think the real question is: How much influence does the voice have on a person's mental processes in the moment? Something akin to smiling to make yourself happy? I highly doubt that talking in "gay" voice or "hetero" voice will change orientation, but can it change their behavior from what it otherwise would have been? (Separate and apart from any outside influences) I bet it does.

I think that is why "fake it 'til you make it" such a powerful idea. You just have to "fake" the right things. And I think that aligns with people trying to look rich without thinking rich and staying poor. Poor is a mindset and has its own voice...and for me it sounds like this: https://www.youtube.com/watch?v=0UtzJRKU8JI

"We become what we think about all day long." --Ralph Waldo Emerson

P.S. After listening to Haystack I really need to listen to some music with a wealth (not rich) voice. Any suggestions?

Tarquin Mahoney

April 25, 2015

wheres the origin of orangutan theory? who told you this? i had no idea but its my theory...what a wonderful clrification of the creative process of intellectual pursuit, ideas are like wheels , testing the bicycle with an imaginary orangutan is always safest though.

Joshua Kennon

April 25, 2015

Replying to Tarquin Mahoney

The idea itself has been around forever but that particular iteration - talking about an orangutan - seems to be credited to Charles Munger. It came to light in the mainstream nearly 20 years ago when Warren Buffett mentioned in a book about Katharine Graham, the Washington Post CEO, that he was a subscriber to Charlie's theory that a person could improve his or her thinking by going into an empty room with an orangutan, who did nothing but sit there and eat a banana, because the process of being forced to organize your thought caused to to be more likely to become aware of errors or weaknesses in your line of reason. By that point, he and Charlie had known each other for 30 or 40 years so no telling when he formulated it.

Tarquin Mahoney

April 26, 2015

Replying to Joshua Kennon

warren buckets, i care not for the man tbh, hes famous for being a moneygrubber - i can imagine him just shouting ut the oranngutan telling it to give him money.aha therein lies the rub.
youve been had, quite dramatically by munger and buckets.
forgive the pun but i do hope you will understand as i give the "orangutan" his credit as it is due,

you only get out what you get in from the theory i suspect.
its a take on the shakespeare and typewriters quote in reality, a sneaky psychopathic plagiarism that gives itself away on purpose actually, monkeys tent orangutan, red bearded bard?
and the quote about a monkey?
theres a famous quote about monkeys and shakespeare i seem to remember which i believe may well of been bufets moivation for liking it, hes a plagiarist after all, encourage ideas and you can steal and profit off them.
you really soured the pudding there, charlie mungers a mystery to me too ut his reputation is now thoroughly tarnished by being friends with warren "i ate the whole" buffet.