Althought I haven't read it, I've heard great things about the book Freakonomics, co-authored by (and about the work of) University of Chicago economist Steven Levitt. Levitt is supposed to be a true original thinker, and has really shaken up the somewhat traditionalist field of economics.
And now, Levitt has moved his popular blog--co written with Freakonomics co-author Stephen J. Dubner--to The New York Times online. This week he started off the new show with (among other things) a post on hurricanes and global warming. Alas, Levitt seems unaware of the history of this debate, and so pretty much makes a hash of things.
Let us detail the errors of interpretation:
1. Levitt links the 2006 World Meteorological Organization statement on hurricanes and global warming--signaling a lack of consensus as to precisely what's happening right now--and opines, "It is interesting...that you can get a bunch of scientists to basically agree that they don't even know whether global warming will cause hurricanes to increase or decrease. The conclusion is especially surprising because no one ever wants to look like they don't know the answers, and because scientists who work on hurricanes have strong incentives to convince everyone else of the important of their research." On the contrary, scientists are very cautious by nature, and when you pull together a group of scientists who disagree, this is precisely the kind of statement you get. Moreover, explain to me again how a statement saying scientists don't know everything is going to detract from the importance of research?Levitt also mischaracterizes the statement itself when he says scientists "don't even know whether global warming will cause hurricanes to increase or decrease." What the statement actually says is: 1) we don't know if hurricanes have yet measurably intensified on average; 2) we don't know how storm numbers, tracks, or areas of impact will change due to climate change; 3) but there are reasons to expect storms to intensify on average and rain more; 4) and sea level rise should worsen storm impacts.
2) Levitt then turns to the latest Holland and Webster study (PDF), extensively discussed on this blog here and here. Levitt says it "flies in the face of the scientific consensus," aka the WMO statement. Levitt doesn't appear aware that Holland was one of the chief people involved in overseeing the drafting of the WMO statement! (See this PDF for his name listing). That statement represents a temporary assessment of where the science stood at a given time--or at least, what could be agreed upon. But some of the scientists involved in drafting it, such Holland, thought even then that global warming was having a much more pronounced effect. Skeptics remain, however, and they too were brought into the WMO process, which is why the statement is pretty conservative and properly and responsibly highlights all the remaining uncertainties.
3) Levitt writes of the new study, "I don't think Holland and Webster have convinced many climatologists; in the four or five articles I read, there wasn't a single endorsement from another scientist." Actually, MIT's Kerry Emanuel (among others) would appear to broadly agree with the Holland/Webster findings and in fact was quoted by the AP commenting favorably thereon. Emanuel was also, like Holland, involved in overseeing the drafting of the WMO statement.
4) Levitt writes, "Two months into what was supposed to be a very active season, with 7-10 hurricanes predicted, so far not a single hurricane has appeared. By this time in 2005, there had already been three hurricanes." Well, 2005 shattered a huge number of records and nobody expected a repeat of that year this time around. Levitt seems unaware that 2005 notwithstanding, June and July are not typically very active hurricane months in the Atlantic. It really doesn't make a lot of sense to debunk this year's predictions outright when we haven't seen the peak of the hurricane season yet (although it may make some sense to downscale them somewhat). The forecasts may well be wrong--I don't invest a huge ton of faith in them, honestly--but last I checked, all the forecasters are still expecting a fair number of storms, even in updated forecasts released earlier this month.
5) Levitt also denigrates the concept of scientific consensus, which is a pretty unsophisticated thing to do.
All and all, I'm not very impressed. As Matt Nisbet might say, Levitt has shown himself to be a "cognitive miser" on the subject of hurricanes and global warming. Which is totally fine...except that he blogged about it.
- Log in to post comments
glad to see that someone's done some much-needed due diligence on levitt's post.
but while his misleading information made for an irresponsible post, i fear that #5 has the largest potential long-lasting impact on the scientific community. disregarding scientific consensus as illegitimate or incomplete not only reflects a poor understanding of scientific methodologies and interworkings of the scientific community, but represents an abysmal grasp on how underdeveloped scientific communication can affect the lay public's use of scientific information made available for purpose.
if he'd just firmly establish that his post is an opinion piece NOT rooted in scientific training (of which he has none), it'd help. but to the sleeping reader, even that acknowledgement might not be enough...
In terms of perceived cultural authority, economists are the new physicists. The media and the public construct them as having magical powers of observation and knowledge that others don't have. In Freakonomics and in the NYTimes blog post, the new movement in economics is to stray into areas where they lack subject expertise and apply econometrics and their field specific assumptions to the topic. Here's one leading example where it goes wrong.
I wonder if some confusion and disagreement comes from differing concepts of the word "consensus." In some fields consensus means "unanimous" and in others it has more of the connotation of "widespread agreement."
For example, in the United Nations, most committees operate by "consensus," meaning everyone has to agree (which Kofi Anan later implied was kind of a stupid way to get anything done). I think many conflict resolution or other decision-making bodies operate the same way (before leaving a meeting, the leader may ask "OK, do we have a consensus?" meaning "is there any objection").
I see a "consensus" on global warming because the "vast majority" of scientists agree on the basic concepts, and the science itself is convincing to me. But someone else may see a few scientists disagreeing, and therefore that there is no "consensus," and by their definition, they may be right.
Just one more thing to add to the confusion.
Thank you for this post Chris. I read that article by Levitt and was kinda awed. The real problem I have with it is Levitt uses his Freakonomics power to bring up other issues that the masses read as gospel. When really he is just another person with an opinion when not speaking about Economics.
I hope you don't mind but I linked this post in a comment I left back there. You may get some Freakonomics Fundies...
Thanks for the comments, folks.
As if to underscore my fourth point, the National Oceanic and Atmospheric Administration is out with its updated seasonal hurricane forecast for the Atlantic, and they have barely scaled back the predictions at all. They're calling for 13-16 named storms, 7-9 hurricanes, and 3-5 major hurricanes.
http://www.cpc.ncep.noaa.gov/products/outlooks/hurricane.shtml
Lat's see, so in terms of public statements I happen to have seen we only have Holland, Webster, Emanuel and Trenberth. Someone should point out to Levitt that a consensus among those four could constitute a fair working definition of "climatologists agree." I suspect Levitt would tend to give considerable deference to someone with a Nobel in economics, but does he even know what a Rossby Medal is?
Also, Chris, I still maintain that the WMO statement is best understood as an agreement to stop yelling at each other until the next round of papers, and to try to tone it down after that. So far the only truly intemperate remark seems to have been from Landsea, so arguably it's sort-of working.
Now I've read the Levitt post, and it's even shallower than billed. Two examples not mentioned above:
1) Levitt writes "The devastation from hurricanes like Katrina, Rita, and Wilma was powerful evidence that man-made global warming had triggered an onslaught of unforeseen consequences -- at least, that was the way the media tended to portray it. Maybe I am wrong, but I think the current focus on global warming in this country would be much weaker had those hurricanes not hit landfall, or had they hit Mexico instead of the U.S."
Funny, I seem to recall Wilma hitting Mexico in a rather major way and having a comparatively minor impact on the U.S.
2) He identifies Landsea as just "a researcher," which shows he doesn't even know who the parties in the debate are.
I read Freaknomics about a year ago and wasn't all that impressed. It seemed to catch waves because of its assertions, not the arguments backing them up, and ultimately I found myself unconvinced.
As Martin Gardner pointed out some 55 years ago in his ground breaking book, "Fads and Fallacies in Science," Mr. Levitt is a perfect example of what happens when an expert in one field (in this case economics) starts pontificating in another field (in this case climate studies) in which he has no expertise. Mr. Gardner went even further in his analysis (which at the time was centered on the theories of Wilhelm Reich) to suggest that the incompetence shown in the new field might well reflect incompetence in the experts own field.
A freelance writer and an economist debating global warming - no wonder this whole discussion has gotten so obtuse and off track.