Regular visitors to the Monkey Cage will probably remember the discussion/dismantling of a recent paper claiming a strong causal relationship between how masculine or feminine a hurricane’s name is and its death toll. As the press release puts it:
An analysis of more than six decades of death rates from U.S. hurricanes shows that severe hurricanes with a more feminine name result in a greater death toll, simply because a storm with a feminine name is seen as less foreboding than one with a more masculine name. As a result, people in the path of these severe storms may take fewer protective measures, leaving them more vulnerable to harm.
The finding indicates an unfortunate and unintended consequence of the gendered naming of hurricanes, which has important implications for policymakers, meteorologists, the news media and the public regarding hurricane communication and preparedness, the researchers say. . . .
The authors found that for highly damaging storms, the more feminine the storm’s name, the more people it killed. The team’s analysis suggests that changing a severe hurricane’s name from the masculine “Charley” to the feminine “Eloise” could nearly triple its death toll.
Most of the discussion of the study has centered on the paper’s perceived flaws. There is a lot to criticize here, such as the decision to include data from a period where hurricanes all had feminine names, the mechanism chosen for ranking masculinity and femininity of names (the unisex “Sandy” was given a surprisingly high femininity score), and questionable modeling decisions and jumps in causal reasoning.
What has been largely lost in the conversation is the importance of the underlying question itself. Does the way we brand disasters, both natural and man-made, affect the public perception of risk? Can making an event sound more ominous give people a heightened sense of danger and therefore save lives? Is this an effective way to promote public safety? Is it an ethical one?
There is a useful antecedent for these questions. We have an interesting case study of what happened when public officials, faced with the problem of needing to persuade people to avoid an activity over a certain period, chose to describe the event in the most dramatic and even ominous language possible.
Martin Wachs of the Rand Corp. and Brian D. Taylor of UCLA’s Luskin School of Public Affairs spelled out the details in this LA Daily News op-ed:
In 2010, Los Angeles began planning for the weekend closure of a 10-mile stretch of the 405 Freeway linking the Westside to the San Fernando Valley in July 2011. During the closure, contractors demolished the southern half of the Mulholland Drive Bridge in the Sepulveda Pass.
The closure was widely publicized and public officials sounded messages of hope that Angelenos would pull together to minimize the negative effects. . . .
Concerned that motorists might fail to heed upbeat calls for behavior change, some elected officials delivered more ominous messages. County Supervisor Zev Yaroslavsky coined the term “Carmageddon.” Local, national, and international media picked up and repeated news of the impending closure, emphasizing the potentially dramatic consequences of motorists failing to avoid the area. . . .
Obviously, the potential toll of even a massive traffic jam is far less than that of a hurricane, but otherwise the analogy is fairly close. Public officials want to present a disaster in a way that persuades the most people to avoid or leave the trouble spots, thus minimizing the human cost and the drain on resources.
If we judge the branding of the first 405 closure by the coverage it generated, Carmageddon was a stunning success. As this clip from “The Colbert Report” shows, the coverage was extensive (not to mention overheated and occasionally borderline hysterical).
There is also reason to believe that the branding was remarkably good at driving down traffic levels. Again, from Wachs and Taylor:
So how did the roughly 300,000 travelers who traverse the affected stretch of the San Diego Freeway each day respond? There was a 61 percent decrease in traffic volume heading northbound on the 405 toward the West L.A. closure on Carmageddon Saturday, compared with a typical summer Saturday. Southbound traffic on the 405 in the San Fernando Valley dropped even more: 73 percent.
Traffic did not simply take detours around the closure. In virtually every location we examined, traffic levels were way down along potential Carmageddon detours. On the Ventura Freeway just north of the closure, westbound traffic volumes were down by 43 percent, and eastbound by 45 percent. Likewise, traffic leaving the 101/405 interchange was down 34 percent in the west/northbound direction and 31 percent east/southbound.
Was this an effective way to approach the problem? There, the question gets murkier. The publicity apparently created a major response but did a poor job of targeting it. Traffic was down in Los Angeles but Wachs and Taylor point out that “statistically significant reductions in traffic flows occurred as far as 80 miles from the closure.” In other words, the goal of reducing traffic on the 405 was achieved but at great unnecessary cost in terms of inconvenience and lost revenue for people who would have been unaffected otherwise.
There is also the possibility of a cry-wolf aspect. We expect over-the-top dramatic language to lose some of its effectiveness with repeated applications, particularly if, as in this case, the original crisis turns out not to be that bad. This was an especially sharp concern with Carmageddon because it was only the first in a series of closures. As Wachs and Taylor put it, “It is unlikely that dramatic messages of fear of a traffic nightmare will have the same effect a second time.”
Finally, there is the ethical question. The government has a clear obligation to present disasters in a way that minimizes injury and damage, but it also has an obligation not to mislead. When these considerations come into conflict, which one takes priority? Carmageddon may not tell us the best way to brand natural disasters, but it’s a good precedent to consider when framing the questions.
Read more here: http://feeds.washingtonpost.com/c/34656/f/666713/s/3c7c14a4/sc/1/l/0L0Swashingtonpost0N0Cblogs0Cmonkey0Ecage0Cwp0C20A140C0A70C140Chimmicanes0Eand0Ecarmageddon0C/story01.htm by Andrew Gelman Originally posted on http://www.washingtonpost.com/blogs/monkey-cage