Responding to journalist Elizabeth Kolbert’s negative review of Freakonomics 2 in the New Yorker, Stephen Dubner writes, that, although they do not have any training in climate science, it’s also the case that:
Neither of us [Levitt and Dubner] were Ku Klux Klan members either, or sumo wrestlers or Realtors or abortion providers or schoolteachers or even pimps. And yet somehow we managed to write about all that without any horse dung (well, not much at least) flying our way.
But Levitt is a schoolteacher (at the University of Chicago)! And, of course, you don’t have to be a sumo wrestler to be (some kind of an) expert on sumo wrestling, nor do you have to teach in the K-12 system to be an expert in education, nor do you have to provide abortions to be an expert on abortion, etc. And Levitt has had quite a bit of horse dung thrown at him for the abortion research. The connection is that abortion and climate change matter to a lot of people, while sumo wrestling and pimps and teachers who cheat are more like feature-story material.
Kolbert writes that Levitt and Dubner treat climate change as “mainly as an opportunity to show how clever they are.” But I don’t think that’s right. Freaknomics 1 was all about Levitt’s cleverness, but Freakonomics 2 is much more about the cleverness of other people such as John List and that super-rich Microsoft guy.
To follow up on one of my earlier thoughts, I think Levitt really has changed his career trajectory in a big way with Freakonomics 2. Before Freakonomics 1, Levitt was a very successful professor: well paid, with the opportunity to work with excellent students, lots of invitations to speak in interesting places, the assurance that people would notice his articles when they came out, etc. After Freakonomics 1, he had all this, plus riches and fame. And I have the impression that he worked hard to keep up with his academic duties, doing research, editing journals, etc. After this new book, though, I think there’s no going back. A lot of people just aren’t going to take his stuff seriously anymore–and the people who do like it, might very well like it for the wrong reasons. Also, Levitt’s gotta be careful now about who he pisses off. He might feel on top of the world now, as an equal-opportunity offender who’s riled conservatives on abortion and race, punctured liberal myths on climate change, and lived to tell the tale. At this point, though, further bold stands might well be subtractive, chipping away at the proportion of the audience that can trust him.
Or maybe I’m wrong on this. Maybe Levitt can have two parallel careers: one as a serious researcher and one as an author of popular books with less well-supported claims. I wouldn’t think it’s possible, but maybe he can keep both balls in the air.
Getting back to the climate change issue, I think Levitt and Dubner are on to something, not about global warming being a “religion”–that just seems like the currently-popular insult, to label things you disagree with as being a religion–but about the idea that going against the grain on this one bothers people, a lot. Again, Freakonomics 1 had some highly-questionable abortion research, and abortion is a hot topic, but this didn’t seem to hurt the reception of the book or Levitt’s standing as scholar. And Freakonomics 2 also features the offensive-to-many claim that drunk people should drive instead of walk and that prostitution is a good career opportunity. But I doubt either of these would’ve derailed the juggernaut (to mix metaphors). It took global warming. Beyond the political issues, one key difference is that there really is a research consensus that Levitt and Dubner are opposing on global warming. For the abortion, drunk driving, and prostitution examples, all that they’re battling are the majority views on morality and common sense. It’s highly plausible that a hard-nosed researcher can bring his quantitative skills to bear on a problem and reveal hidden truths that are counter to conventional morality and common sense. It’s not so plausible that an outsider can demolish the findings of scientists such as Raymond Pierrehumbert (see below) who are themselves quantitative researchers.
Whatever you say about the merits of the case, I admire Levitt for taking his stance on climate change. Freaknomics 1 made lots of money, and Chicago has a pretty low cost of living (I’m comparing to New York, remember), so it’s not like he really had any need for the sequel. He’s already famous and can get in the newspaper whenever he completes a new research article, and he also has a blog where he can share his ideas with many thousands of people and also promote others’ work that he likes. Rocking the boat with Freakonomics 2, though, that changes everything. He gets more of what he doesn’t need more of (money and fame) but maybe loses big in other ways. If he really cares about the geoeongineering thing, though, then it could be worth it. Throwing away (part of) his reputation for a cause–that’s admirably selfless, I think.
When his University of Chicago colleague Raymond Pierrehumbert (in the Geophysical Sciences department) slammed his discussion of geoengineering and invited him to come by his office (by a somewhat circuitous route) to discuss the topic, Levitt responded by saying that he enjoyed Pierrehumbert’s “intentional misreading” of the chapter and saying, of Levitt’s own writings on the subject, “I’m not sure why that is blasphemy.” I’m not sure on this latter point either–I did a quick search and couldn’t find the place where Pierrehumbert described Levitt’s writings in those terms.
But I do feel for Levitt on the general issue. When I worked at Berkeley, I had a couple of colleagues who intentionally misread my work, for example describing a nonlinear differential equation model as “linear.” (Actually, I have a feeling that they were (a) too lazy to read my articles in detail and (b) too stupid to understand them, but I’ll just be charitable and describe them as intentionally misreading.) I had thoughts of going over to their offices to explain their misunderstandings, but I was so angry that I couldn’t bring myself to do it. I suspect Levitt feels similarly about Pierrehumbert. It’s hard for me to imagine exactly what conversation the two of them would have in that office in the Hinds Geophysical Laboratory, but probably nothing much would come of it.
P.S. I really hope I can use some of this in my next book. Otherwise I’ve been blogging my ass off on this topic and I’m getting nothing out of it.
I think as an intellectual approaches rock star status it is much harder to continue with the scientific method.
Tom Schelling avoided this, in my opinion. But that was in an early time and his fine essays in Choice and Consequence never became #1 in popularity.
ooh my favorite topic!
Why did people get most upset about the climate chapter in FC2? I think your right that the fact that they went against years of established research played a role. Some other reasons:
I think the only topic that Levitt has addressed that's similarly hot button is abortion. I think there are two important differences here:
a) Levitt was very clear in that he didn't advocate a particular policy. And I think that's actually plausible: Most people who are against abortion(rights) think that abortion is in itself a very bad thing, some equate it to murder. If you believe that, then some positive externalities shouldn't affect your stance on abortion. And that has actually been the reply of many anti-abortion activists.
b) Although it looks like Levitt and Donahue may have made a couple of mistakes in their article, it is hard to deny that they thought hard about it and came up with a clever identification strategy for the effect of abortion on crime. There is nothing particularly new or clever about the climate chapter.
Finally, I think it's a timing question – FC2 co-incided with a well-orchestrated campaign against global warming science, and FC2 actually used a whole bunch of the "skeptics" half-truths (e.g. the alleged global cooling cooling consensus from the 70s, the (non-existent) temperature decline over the last 10 years, etc.).
I think anti-abortion types are less well represented in academic departments and journalism. Not to say that Levitt took a "pro-abortion" position either.
Most of the stories Levitt paints in his research are very simple: someone is willing to cheat if their expected benefit is high enough. But climate science is probably the most complex of all fields of study: there is tremendous uncertainty and even chaos in climate systems.
So the standard economics expected utility maximization seems highly delinquent. There's a place for economists in the climate change realm, but leave the risk assessments to engineers (in particular, engineers are familiar with the idea of "unknown unknowns"–adding something new to a system might have consequences we didn't think of. That's something an economist wouldn't consider.).
The problem is that they're trying to perform climate science without being climate scientists. They'd get similar poo flung if they tried to perform sumo wrestling, or realty, or abortions. I'm not sure why that's a difficult distinction for them to understand.
When I say "without being climate scientists" I mean they don't seem to know much about the subject. That's highly correlated with, but not the same thing as, having some kind of degree or publication record in the field.
Of course, there can be good contributions made by outsiders to the field, but any such outsider necessarily has to spend extra effort showing why people should believe them even without a solid reputation to back them up.
A good editor would correct the insult from "religious" to "dogmatic". Religion is about god(s). It's not religion they're objecting to, it's dogma.
Andrew,
Why do you think it is more charitable to say that your colleagues intentionally misread you than to say that they were lazy or stupid? I think I would rather misunderstand something because I am stupid than correctly understand something and then intentionally misrepresent it. The latter (but not the former) seems like a serious moral failing to me. (Maybe laziness is as well, but somehow, it doesn't seem as bad.) Do you disagree?
On the subject of the abortion paper and "Levitt's standing as scholar," he may still do fine in economics, but he's kind of a joke in criminology.
The abortion analysis was shown to be patently ludicrous bunk in a variety of ways. Levitt only responded to the most blatant failure that Foote and Goetz pointed out, i.e. that the paper lied about including had a coding error regarding state-year effects.
But worse, really, is the utter disregard for reality/sanity checks. If Roe v. Wade caused the crime drop, abortion must have risen substantially among the "at-risk population" they're talking about. But they didn't. If Roe v. Wade caused the crime drop, the drop in age-specific rates would have occurred in appropriate chronological order, for those cohorts directly following legalization. But they didn't. Why would someone making an cohort argument not test the rates for cohorts? It's all completely inane, and it's an embarrassment.
Levitt's other claim to fame in crime research, a 1997 article on the effect of police on crime rates (http://www.jstor.org/pss/2951346), was also shown to be junk econometrics (http://www.jstor.org/pss/3083311).
Fool us once…
1) "He gets more of what he doesn't need more of (money and fame)" How do you know that Levitt doesn't need, or think he needs, more money? Most rich people I know quickly adjust their spending to their wealth. Finance folks continue working ridiculous hard even after they are much richer than Levitt.
It is, however, common for people with wealth level X to think that anyone with wealth 10*X does not need more money.
2) Levitt related: Here is further commentary on his ludicrous claims about the birth month of World Cup players. Feel free to steal that graphic and put it in a new thread. I would love to get some feedback from your readers.
David: I guess it depends on how you define "need." I don't need all the money I have, but it doesn't stop me from wanting more, so I could well imagine Levitt feeling the same way. But, sure, it's possible that Levitt has some sort of "leak" and is spending his fortune as fast as it's coming in. I don't know the guy personally.
engineers are familiar with the idea of "unknown unknowns"–adding something new to a system might have consequences we didn't think of. That's something an economist wouldn't consider.
I suggest reading some economics. Googling "economics unintended consequences" should prove enlightening.
Why do you think it is more charitable to say that your colleagues intentionally misread you than to say that they were lazy or stupid?
This is a running debate in our household. Is stupidity or malice more common, dangerous, and/or blameworthy? "You can't fix stupid."
I can not follow the claim by Sam R. that Levitt is a joke because he did not test something which is (he feels) important. In micro-econometric research (empirics with data at the individual level) identification is "the" most important issue. Levitt rose to fame in an era in which the instrumental variable approach took off. He could publish (and he did) by just coming up with another clever instrument.
Later on, his abortion-crime these got crushed as people found out that there were some mistakes in it. E.g. in the panel regression he estimated, inappropriate errors were calculated.
That is were he hit the wall. He claimed to have found an effect, which is not there.
If I would answer a relevant question and someone from another discipline would say that I am a fool because I do not use the research approach that is common in the literature, then I would be inclined to consider the suggestions but I would not accept such a remark. We can't use every conceivable approach, each time we want to answer a question.
Empirical disciplines tend to look at questions through their own methodological lens. An applied economist has far harsher standards whent it comes to identification then a sociologist. At the other hand, the economist tends to regress everything he sees, ignoring straightforward approaches that might be more insightful.
I do think that one may add value when taking his methodological machinery to another disciplne, e.g. using the lens of an economitrician (pun intended) in the climate debate. Saying things like, an economist is unfit to this because he hasn't the machinery, strikes me as a statement which is -prima facie- wrong.
I would agree that Levitt's dissent from orthodoxy on global warming showed selflessness and courage.
I do not, however, see much evidence from a quick Google search that opposition to Levitt's view on climate change are motivated by his loss of credibility over the abortion-cut-crime fiasco. I don't see much evidence that his more than a few of his critics on climate change are even familiar with what a disaster that turned out to be for him. You are familiar with it, but I don't think many other people are at all.
As for his abortion-cut-crime theory, it was wildly popular because legalized abortion is popular with much of the elite, and a common argument for it that you hear in private (although seldom in public) is more or less eugenic: abortion eliminates people who are mistakes, so that's got to be good for society. In Levitt's 2001 paper, he attributed 39% of the effect of abortion in cutting homicides to the #x higher rate of abortion among black women combined with the 7x higher rate of homicide offending among blacks.
I pointed out to Levitt in our debate in Slate in 1999 that his abortion-cut-crime theory only looks true if you did what he did and look at crime statistics just in 1985 and 1997. If you look at the intervening years and you look at national homicides rates by age group, then you see the youth homicide rate of the first post-legalized abortion cohort being triple that of the last pre-legalized abortion cohort. His only reply was that he was an expert statistician who had looked at all the data by state, and that proved he was right.
http://www.slate.com/id/33569/entry/33571/
Between 1999 and 2005, a variety of academics published articles making my point in greater detail and rigor.
In late 2005, a half year after Freakonomics hit the bestseller list, Foote and Goetz showed that the result from his state-level analysis was caused by Levitt's programming errors.
So, I was right and Levitt was wrong.
But, how widely is that understood? Although the WSJ and The Economist reported this, the New York Times refused to print this story about their new star blogger.
How often do critics of his climate change theory mentions his abortion embarrassment? Not often, so far as I can tell.
Re: Cost of living in Hyde Park for a U. of Chicago faculty member.
The Obama family made a quarter of a million in 2004, up from about $160k in 1997, yet didn't appear to be saving a dime. Then, they got really rich off books and Michelle's diversity / political wife job with the U. of Chicago Hospitals starting in 2005, but didn't start putting money into a tax sheltered SEP until 2007, suggesting they spent every penny of the three million bucks they took in in 2005-2006.
Levitt has more kids than Obama.
That said, I think Levitt would make more money sticking to the orthodoxy on global warming. Look how rich Al Gore has gotten off it.
So, as I said, although I haven't been terribly impressed by Levitt's character in the past, I think this shows character.
To all:
All I can say is that, after Freakonomics 1, many people (including myself) thought of Levitt as a serious researcher and respected his opinion on social science research questions in general. If a question arose and Levitt expressed his thoughts, I'd listen. This is not to say that I would assume Levitt was always right (recall the abortion controversy), but I'd assume he was thoughtful and serious in his opinions.
After a few years of the blog and Freakonomics 2, I'm not so sure what to think. I suspect that his scholarly research, Levitt is as serious as ever, but when he writes about other things (including the research of his friends and colleagues), I don't really have the confidence that he stands by his words. I wouldn't say that my mind was changed by the climate-science story, or the missing-girls story, or the cars-don't-pollute story, or any other particular example; but all of these (and others), put together, give a general impression of "casual inference," one might say.