Skip to content

My review of Freakonomics 2

The above title is a joke. I haven’t actually seen the book. As a big-time blogger, I get some books in the mail to review, but maybe this one is sitting in my NYC office. Anyway, the backlash has begun, so maybe this is the right time to buy low and be the first to offer the contrarian claim that, despite what everybody’s saying, the book is awesome.

From a short-term economics standpoint, the controversy has gotta be good for the book. So far, Levitt and Dubner have put the words “GLOBAL COOLING” on the cover of their book, they’ve endorsed a report saying that future trends are “virtually assuring us of about 30 years of global cooling,” and that “even if man is warming the planet, it is a small part compared with nature,” and they’ve written that “we believe that rising global temperatures are a man-made phenomenon and that global warming is an important issue to solve.” That last bit should do the job of ticking off anybody who was with them so far! (I was actually surprised when reading the comments on that last quote–where Levitt assures us that they do believe in global warming–that pretty much all the global-warming-skeptics among the commenters still seem to think that Levitt is on their side. I guess half a loaf is better than none at all, politically speaking, but I’m surprised that more of them didn’t get angry at Levitt for saying that.)

I don’t want to get into the substance of climate models, a subject on which I’ve worked on only a little bit. (The paper we wrote a few years ago never got published–actually, we never finished it enough to submit it anywhere–and our current work on the topic is still in the research-and-writing-up stage.) But I do want to speculate a bit on the political angle.

The interesting question to me is why is it that “pissing off liberals” is
delightfully transgressive and oh-so-fun, whereas “pissing off conservatives” is boring and earnest? Based on their writings in Freakonomics 1 and their blog, Levitt and Dubner strike me as open-minded political pragmatists, so it’s not that I think they have a big political agenda.

It’s possible to write things that piss off conservatives while still retaining an edgy, transgressive feeling–take a look at Nate Silver (or, to take a less analytical example, Michael Moore)–but I think it’s a little harder to do. Flouting liberal conventional wisdom is funner somehow. As I said, I think there’s something more general going on here but I don’t feel I have a full picture of this phenomenon.

Freakonomics 1 was based on Levitt’s previous research, which was all over the map, whereas in Freakonomics 2 the authors got to choose ahead of time what to be counterintuitive about. When it comes to writing, too much freedom can be a dangerous thing! Tennis without a net and all that. When Levitt popularizes (as in the disastrous Oster/Das Gupta episode) he can come off really badly; he’s much stronger when discussing his own work.

One problem with social science methodology–when it is not accompanied by rigorous
data collection and analysis–is that it is a loose cannon that can be turned in virtually any direction to yield a counterintuitive finding. And then the question arises: how do you choose which direction to turn your cannon?

I guess there really is something to this “political correctness” thing or whatever it should be called. In Freakonomics 1, Levitt and Dubner featured some questionable work on abortion and crime. Reviewers reacted to this analysis with skepticism but didn’t condemn the whole book. In Freakonomics 2, they question global warming and there’s a veritable poopstorm of controversy. If they’re not careful, this book will send them from the “popular science” to the “political punditry” category, with no turning back. Perhaps Freakonomics 3 will have a chapter explaining why evolution is just a theory, not actually proven at all?

And just wait till some kids gets killed by a drunk driver, Freakonomics 2 is found in the glove compartment, and their parents sue Levitt and Dubner for 50 million dollars. That’ll be publicity that money can’t buy!

P.S. Uh-oh. Just to be clear, I’m not endorsing such a lawsuit.

P.P.S. I should have put this earlier, but . . . let me say how much I admire much of Levitt’s work. For whatever reason, the social science work that my collaborators and I do tends to be understated to the point of invisibility, usually staying well clear of any policy recommendations. Our splashiest, most counterintuitive findings were probably that redistricting is good for democracy (this one got us mocked in the House of Commons, one of my proudest moments) and that elections are predictable even though the polls jump around a lot. Neither of these is really so exciting, and most of my applied work is much more technical, for example our analysis of who should measure their houses for home radon. True, we did go against the EPA on this one, but we never really got political traction. So I admire the willingness of Levitt, along with other applied micoeconomists out there, to go out on a limb and make some data-based causal inferences. In particular, the Freaknomics 1 book was great. It really gave a sense of the process of social science and the interplay between subject-matter ideas, data, and analysis. (We tried to get some of that feeling of excitement into Red State, Blue State, but we couldn’t quite do it. Our book is readable, I think, and interesting, but we didn’t capture that joy of discovery.) Between Freaknomics 1 and Freakonmics 2 came years of blogging, and I wonder if that dulled Levitt’s edge. In a blog it’s just too easy to throw out ideas without seriously evaluating them. And it probably didn’t help that, with their book’s immense popularity, they were getting hundreds of admiring comments on the blog each day. In my own blog, in contrast, the comments are few enough that when somebody disagrees with me, I notice it.

P.P.P.S. More here.


  1. Here's a serious mistake in Superfreakonomics that I heard from Paul Krugman's blog today, "…he [Levitt] quotes Weitzman, which cites his [small] probability of utter catastrophe as if it were a reason to be skeptical of the need to act." This is similar to the criticism that the large Stern study on the costs and benefits of global warming action used a low (but non-negative) discount rate.

    The replies seem to center on not valuing future generations less than current, but as primarily a finance person I focused on the risk-return tradeoff in a post on this two days ago:

    The risk-return tradeoff says that the higher the risk of an investment, the higher an average rate of return you will, or should, require. But it also says, conversely, that the lower the risk of an investment – or the more risk decreasing an investment – the lower an average rate of return you will happily accept.

    What average rate of return do people happily accept for fire insurance for their home? A negative one, not just a low one, a very negative one. People even accept a very negative return for insurance on their car.

    So what return would you accept for fire insurance on the planet you, and your children, and your grandchildren will live on? Scientists aren't that sure what exactly will happen with global warming. The feedback effects could get out of control and devastate the planet.

    Stern's study justifies large spending on global warming even using positive rates of return (or discount rates). But when you use a negative, insurance like, rate of return, then clearly it becomes far more than worth it to spend at least moderate sums combating global warming, sums much greater than anything that's currently being discussed.

  2. jonathan says:

    I've been reading the commentary and now the chapter is in pdf and it's an interesting reasonable magazine quality article, one that seems intent on picking one of the many iconoclastic views available in discussing global warming. It's not very good, meaning it isn't New Yorker standard but maybe might make GQ. It isn't a denial of science but rather a reducto ad absurdam; it takes the manyfold complexities of the science (and politics and economics) and picks a simplistic vision, complete with a few selected voices. That is like what you'd read in a Sunday newspaper and it is not what I expect from an academic of Levitt's caliber.

  3. Mike says:

    "the disastrous Oster/Das Gupta episode"

    I am guessing you are referring to the 'missing women' debate. But I am not certain what you mean exactly and how it relates to Steve Levitt. Care to elucidate?

  4. Phil says:

    I don't disagree with anything that you say here, but I will note that part of the reason for the near-invisibility of your policy recommendations may be that, at least at times, you work so hard to avoid anything controversial. I'm thinking in particular of the true story about the EPA rejecting one of our grant applications because, the grant officer said, they were worried that our policy recommendations would be at variance with theirs (which, in the end, they were). Last year, literally fifteen years after the event, you still wouldn't let your name be attached to this story on someone's (Seth's) blog, for fear of repercussions! People in general, and the press in particular, love controversy, and stories about the Big Guy trying to suppress the little guy, and so on, so if you're not willing to give them any of that, you're giving up a major avenue for attracting interest.

  5. Jim says:

    "why is it that "pissing off liberals" is
    delightfully transgressive and oh-so-fun, whereas "pissing off conservatives" is boring and earnest?"

    If you start from the possibly unfair position that conservatism is about acting in one's own self-interest and discounting costs to others, while liberalism (in the American sense) is about promoting norms of collective altruism, then a lot of comedy starts to look basically conservative, in that it's about saying one thing ("I'm a good, socially responsible person") while doing another ("I'm basically irredeemably selfish"). See Curb Your Enthusiasm, seasons 1 through 7, for example. It's very, very easy to make jokes in this kind of vein.

    Maybe there's a contrast to be drawn between the economic sphere, where Adam Smith says self-interest promotes the collective interest, and the social sphere, where our self interests are promoted by being seen to act in a socially responsible way.

  6. Andrew Gelman says:


    You're right. I want the fame and fortune but without the conflict. I like what I do, and I don't want to end up like Levitt, being in the awkward position of spewing out contradictory statements in a desperate attempt to defend my more outrageous pronouncements.

    But I suspect Levitt was the same way until Dubner came along. If a bigtime journalist came up to me and offered to write a book, "Freakostatistics: How a maverick statistician showed why presidential candidates don't matter, gerrymandering is good for democracy, and the EPA doesn't understand decision analysis," I'd probably take the offer in a minute.

    The only drawback is that to qualify as a "rogue" statistician, I'd probably have to take a job at the University of Chicago. It's a great place, and I love deep-dish pizza and Italian beef sandwiches, but I can't stand the weather.

    Mike: Search this blog for Oster for more on the story.

  7. David Kane says:


    Phil writes:

    Last year, literally fifteen years after the event, you still wouldn't let your name be attached to this story on someone's (Seth's) blog, for fear of repercussions!

    Is that a true statement? If so, some more context would be appreciated. What is the point of tenure if not to allow you to tell the truth, damn the "repercussions?"

  8. Andrew Gelman says:


    No, the statement isn't true. It happened less than 15 years ago. Also, tenure had nothing to do with it: pissing off people at EPA would have no effect, one way or another, on my tenure in a statistics department. It might affect my ability to get an EPA grant, but I've never actually been on an EPA grant (at least not that I remember, but something could've slipped my mind) so that's not such an issue either.

    I can't remember the details of how this was happening on Seth's blog, but I don't think I wanted my little story to be part of a larger narrative that I didn't necessarily agree with.

    For many years I've been planning to write something about incentives in environmental statistics. For most pollutants, there's a pro-pollution lobby (the coal industry, the oil industry, the semiconductor industry, whatever). But there's no pro-radon lobby: radon is a natural pollutant. As a result of this lack of counterpressure, there's a tendency to overstate the radon risk. To put it another way, when the EPA is pushing everybody to measure their house for radon, they're pushing at an open door. If Phil and I ever do get around to writing this up, we can include our story.

  9. Phil says:

    Andrew: There's a lot of space between making true controversial statements, and making controversial statements that may or may not be true just for the sake of publicity. Even if you think it's a slippery slope, you could start sliding down it now, and with a bit of restraint you wouldn't get to, um, Levittown before you retire.

    David: the events took place in July and August, 1996. So, Andrew is correct, that was just a bit over thirteen years ago, not fifteen years. There really isn't much more to the story than what I said about it above. Actually there is one additional fact, though it's not very relevant. We proposed some research and were initially told that we had been awarded the grant through a joint NSF-EPA program; we were told a few weeks later that they had decided not to fund it after all, because the work had already been done (by us, no less!); we explained and documented that was not true; the grant officer then admitted over the phone that the real reason that they were withdrawing the support was that the EPA thought we were going to suggest policies that differed from theirs. We (with colleagues) eventually did the work anyway; the resulting paper is on Andrew's website. For some reason, some people think this story is scandalous…but aren't we a little old to be shocked, SHOCKED, that funding decisions are affected by concerns about saving face?

  10. Andrew Gelman says:


    As you perhaps know, my family lived in Levittown for several years (but moved before I was born), so you're closer to the truth than you realize.

    As for the rest: (1) I don't think either of us was "shocked" that the funding decision was affected by concerns about saving face or whatever, but I recall that both of us were really pissed. Actually, I think I was more pissed than you were. (2) I have no problem telling the story, I just didn't want it as part of Seth's larger story, which I recall I had some issues with. It's a story I'd rather tell myself (or with the help of some Dubner-like facilitator).

  11. David Stern says:

    The economics in Chapter 5 of Superfreakonomics is just as bad as the climate science. At least in the discussion of externalities where they get at least two things wrong – talking about compensating the victims of pollution with the tax raised by taxing pollution and talking about a volcano generating "positive externalities".

  12. Andrew Burton says:

    The reason "pissing off liberals" is good fun is that it's part of the strategic objective on the right, while "pissing off conservatives" is seen as collateral damage on the left.

    Compare, for example, Ann Coulter and Rachel Maddow. For Coulter, writing books about how liberals are godless, treacherous and stupid is a business model. Maddow hosts a show in which conservative arguments are taken seriously, but then the host (or a guest) presents a Queensbury Rules refutation – ie, taking quotes out of context or simple invective are out of bounds.

  13. otherwayaround says:

    "…conservatism is about acting in one's own self-interest and discounting costs to others, while liberalism … is about promoting norms of collective altruism…"

    Isn't it the other way around?

  14. skylights says:

    "Isn't it the other way around?"


    I'll grant that there is not a perfect dichotomy. Conservatives are individualists –until people start doing things they think are immoral. Conservatives don't believe in government promoting the collective good, but they do believe in getting benefits for their particular "tribe" (community) if it benefits themselves.

    Liberals promote norms of collective altruism, but they also favor individual freedom and expression. Liberals generally oppose proscriptive mandates (like "don't do drugs," "don't have gay sex") but support prescriptive mandates (like "pay your taxes," "wear your seat belt").

    Just my quick take. Exceptions abound.

  15. Stuart Buck says:

    The interesting question to me is why is it that "pissing off liberals" is
    delightfully transgressive and oh-so-fun, whereas "pissing off conservatives" is boring and earnest?

    Because liberals predominate in universities and the media. It's the same phenomenon whereby a Catholic priest opposed to abortion is just the usual earnest thing but a Catholic priest bashing the Pope is transgressive. The concept of "transgressive" can exist only where there is otherwise assumed to be some sort of obeisance.

  16. Sarah Lucy says:

    As a fan of Freakonomics, my initial reaction to this book was something along the lines of: a friendly but wild read that keeps the mind flexing at a steady pace. Like in their previous book, it seems impossible that the authors can assemble so much diverse information into a seamless and absorbing narrative, and they keep it going in this book to a point.

    When their discussions turn to child safety seats and climate change, the book unexpectedly veers off course and essentially plunges off a cliff. These areas sour the book with blatant poor research. Their conclusion that child seats don't save childrens' lives any better than seatbelts is troublesome. That statement in itself would be fine, even useful, if it was based upon good research. However, the authors base this opinion on a study that looked at the past thirty years as a whole! Most everyone knows that child safety seats have come a long, long way, and all the data I've seen suggests that seats made in the last decade are far superior to seatbelts.

    Sequels disappoint; don't ask too much of this one. Perhaps you cannot blame authors for wanting to cash in on their popularity, but if Super Freakonomics had been written before Freakonomics, few people would have bought it. The authors are trying very hard to shock and amaze, but the organization is scattered and the research seems questionable.

    Their standard formula is to begin with a counterintuitive statement and before your very eyes show you how clever they are. I, for one, do not see how prostitutes are patriotic, and thought that the comparison between Al Gore and Mount Pinatubo forced and unconvincing.

    It appears the authors casually chose climate change as another topic to spout off on since it's a hot button that will help them sell books. Being fun and irreverent is great if the topic is whether or not sumo wrestlers cheat, or at least their research is sound. Levitt and Dubner cheapen their thinking by presenting a few dozen pages that are littered with obvious mistakes and even the perpetuation of myths, such as the scientific consensus in the 1970s predicted global cooling. Anyway, the "expert" they cite has already publicly renounced them for misrepresenting his ideas, and the whole thing is a shame.