superfreakonomists don’t understand how to stop free riding. does anyone else? [apologies to marwell and ames]

Superfreakonomics is out. Well, one chapter of it is out because it sort of escaped. Levitt and Dubner have turned their professional contrarianism on climate change and — it seems — they were not so careful as they might have been with some of their facts and interpretations. Holy hell has broken loose in the blogosphere. People are not happy. Dubner has now responded to some of it. I can’t say I’m very convinced by what he has to say. I’ve read the chapter, and it seems to me that I’ve noticed a problem beyond the zillions that have already been identified. You see, Levitt and Dubner close with what has been described as a “silly analogy” but when you really look at it, it becomes clear that the problem isn’t that it’s silly. It’s that they don’t understand the implications of their own analogy. And they don’t understand them, I think, because they ignore basic sociology.

For those of you who have not read the chapter, I suppose I should say what that analogy is. Levitt and Dubner tell the story of Semmelweis’s difficulty in convincing doctors that they should wash their hands. Semmelweis figured out in 1847 that if doctors washed their hands, they could save a lot of lives. They note that even recent studies have shown that doctors still do not wash their hands nearly so often as they should, even though they know that is very bad. Then, they tell the story about how Cedars-Sinai responded to a 1999 study showing that doctors just didn’t change their behavior. They had a hard time. They tried all sorts of things. In the end though, they managed to do it by culturing the hands of many of the top doctors in the hospital while they were at a lunch meeting, showing them how filthy their hands really were, and putting the cultures up as screensavers in the whole hospital to remind people all the time. It’s a nice story. It was hard, but in the end, they were able to get handwashing rates up to nearly 100 percent.

What lesson do Levitt and Dubner draw from that analogy? They argue that the reason it was so hard to get doctors to wash their hands, even though they knew it was important, was because of externalities. The doctors were getting other people sick, but their own lives weren’t in danger so it didn’t seem that important. That seems plausible. But then things get weird. They argue that the analogy supports their argument that we need to think seriously about geo-engineering — rather than trying to control CO2 — because we’ll never be able to get people to change their behavior. We just don’t pay a high enough individual price for our own bad deeds. Homo Economicus!

It’s here that I start to ask myself whether they read the words they write, or like memento-on-speed, they forget even their immediately previous paragraphs. Are they trying to be contrarian even to themselves? I mean, they are talking about something where the science is clear, it’s important, and at the end, Cedars-Sinai does manage to get people to do what they are supposed to do without changing their incentives. And the lesson they draw for the climate change debate is that we need geo-engineering?!?!?! Where does that come from? I look at the exact same analogy, and say we’ve got something where even they now say the science seems pretty clear — CO2 causes climate change — so we need to work really hard to get people to change their behavior, and, while difficult, it is possible. We just have to go to great lengths to make people really understand that it is really bad, and we cannot relent on the social pressure.

So what does the analogy suggest we should be doing, besides improving our own behavior? Well, for starters, when people suggest that we’ll just figure out a simple technological fix that will surely work forever, we make it very clear that they are treading in dangerous territory, and we especially do not let them get away with any untruths. The burden of proof is on those who advocate the unknown, not those who advocate trying to maintain or return to the known. Their own analogy suggests that people should be tearing into them in order to maintain awareness that carbon dioxide is the right villain, insofar as inanimate objects can be villains. It is important to maintain that social norm, to chastise those who might even unwittingly dismiss it. This doesn’t mean stop thinking. We can still be fallibilists. We can certainly explore geo-engineering, but we cannot make the case for it by pooh-poohing the very possibility of behavioral change. People are social animals. It’s just not true that we never solve collective action problems. Did Levitt and Dubner never hear of Elinor Ostrom? I think she might have recently gotten some sort of prize.

We can even turn back to their own analogy to make a case against geo-engineering. Why should we care if doctors wash their hands? We’ve invented penicillin. And that’ll surely work forever, right?

18 thoughts on “superfreakonomists don’t understand how to stop free riding. does anyone else? [apologies to marwell and ames]”

  1. Hey, let me add my voice to those who are delighted you’re here and not on orgth— no wait. That can’t be right.

    Did Levitt and Dubner never hear of Elinor Ostrom?

    I believe Levitt has said that he’d never heard of her prior to her Nobel.


  2. BTW, I read the chapter after reading the various shame-on-them posts from Krugman and Brad DeLong etc. and I didn’t think it was bad as how they made it sound.


    1. To call it not as bad as the critiques made it sound is to damn with pretty faint praise! I agree that the tone of critique is both very aggressive and not very charitable, but I am sympathetic to the problem that the people attacking them have. They are trying to get people to understand that. To paraphrase Krugman, this is not the ethics of sumo wrestling; it’s important! If they are right, it’s no place to be contrarian or cute. You have to be willing to sacrifice cuteness for rightness. To me, it makes a lot of sense that outrage would be (dare I say “should be”) a function not just of the sloppiness of the analysis and its presentation, but also the importance of the issue, and the size of the opponents platform. That’s why I find the analogy compelling. A seemingly small thing has big consequences and gets pooh-poohed. How do you get people to realize it has big consequences? Never let them forget that it is important to wash your hands/mitigate carbon. D&L take a huge platform, and try to build support for the “suprising” position — geo-engineering — by suggesting that it could be a cost-effective *substitute* for mitigation, not an oh-shit-we’re-already-beyond-thunderdome complement.

      I also think that while the critiques are often uncharitable, it seems pretty clear to me that D&L are no more (and maybe less) charitable in their interpretations of those they are arguing against. So first stones, gooses, ganders, and all that.


  3. I find the chapter pretty awful, but then again I find much of the stuff they spew pretty idiotic. I thought all of Freakonomics was basically a bunch of just-so stories, utterly unfalsifiable, that asserted that an untenable, unlikely, and unverified kind of morally evacuated subject (H. economicus) explained all kinds of phenomena. The only saving grace of this atrocious collection of fiction (which, by the way, was the favorite of one of my co-committee members on the Summer Reading Program committee last year) was that most of the phenomena didn’t need explaining in the first place. So what if people think economics can explain why their milk cartons are square? It can’t, of course, but where’s the harm in letting them believe it’s true?

    jdw has pointed out that, in this case, it’s far more important. The phenomenon does need explaining, and economists are definitely not the right people to do the explaining. Particularly economics evangelists whose meta-point is just that economics always explains everything.



    1. Yeah, I personally regard drug/gang policy, the cause of large changes in crime rates, and reasons for differences in educational outcomes as important social issues, so I don’t really get the complaint that Freakonomics was just trivia and only now with global warming have Levitt and Dubner engaged an issue that’s important and/or “does need explaining.”


      1. Sorry if that came across as pissy. I had *many* arguments of my own with Freakonomics, but I saw it as a rather different book. I didn’t think they were so much pushing a “homo economicus explains all” perspective as a perspective where any kind of logical/coherent explanation of social behavior was categorized as “economic.”


      2. No, not at all – your critique of my comment is right on. And I don’t think our interpretations of Freakonomics are so different either – many of the expectations they make just state that they’re the result of H. economicus even when this makes no sense.


      3. Funny, my reading is to think of Freakanomics as a smug way of telling people they don’t know what they’re talking about. That they are relying upon silly assumptions to generate knowledge. And that the economic approach they deploy helps reveal this (both the assumptions, and the fact that knowledge isn’t knowledge).

        What frustrates me about this is that sometimes people do know what they’re talking about, even when they are making assumptions. Such assumptions are always required. But that doesn’t mean the position is incorrect. And more importantly, what really annoys me is the framing that the freakanomics crew are somehow more “free” of this sin than others. The economic approach has many benefits. But a lack of assumptions is not one of them.


  4. Great post. As you all say, there’s a non-trivial amount of, you know, actual research about what works and doesn’t work for solving social dilemmas. Not to mention theories.


  5. Isn’t there is a difference between using social pressures to change the behavior of a relatively small number of individuals in a very closed off social group and changing the behavior of every person on earth? I think they did read their Ostrom but they realized it just didn’t apply here.


  6. Josh: I’d say that a short answer to your question is that you are apparently mis-analyzing the problem. Of course it is impossible to get people to change their behavior voluntarily one by one, but that isn’t how it gets done. The solutions are political or normative and often involve coercion or strong social sanctions. US waterways were cleaned up, for example, not by persuading people to stop dumping pollutants into water, but by requiring businesses and communities to clean their water and imposing fines if they did not do so. There are still some violators, but the water is a heck of a lot cleaner than it was 40 years ago. Ditto the air quality in Southern California.


    1. I don’t think I mis-analyzed it. There was a reference to Ostrom whose work focuses on overcoming collective action problems in smaller groups without resort to privatization or top-down regulations.


      1. Uh, no. You did. In context, referencing Ostrom was to make the point (albeit snidely) that social pressures can change behavior, which is part of the really big point that her work makes, and which is the point that D&L ignore. That’s OW’s point to you. “The solutions are political or normative and often involve coercion or strong social sanctions.” Ostrom is only one of the people writing on it, and yes, she does have a focus. But given her current prominence, she was convenient shorthand for the point that D&L are ignoring something that even their discipline does in fact know.


      2. Just to toss in a few more examples. (1) Why do people not urinate and defecate in the streets in most of the US (unless they are homeless and without options, or drunk)? Why do people sometimes endure quite a bit of discomfort to find an appropriate place for their bodily function instead of just doing it whenever they feel the need? There are laws against this, yes, but those laws are enforced by even more powerful social norms that make most people feel that it is disgusting to even contemplate such a thing. These norms affect very large groups. (2) Study of the “second order problem” of enforcing other people’s compliance with norms shows that there are lots of cases in which rational self-interested benefit-cost maximizers will act to enforce other people’s compliance with norms. (3) Political systems embody sunk costs for enforcement systems. Depending on the incentives for non-compliance and the degree of normative enforcement, the marginal cost of enforcing one more law can be close to trivial. This is easy stuff that is very well established in the sociology & political science literatures.


  7. To be clear, I certainly at least don’t think of myself as a knee-jerk opposer of Freakonomics. I won’t go so far as to say that I liked Book I, but I certainly did find things to like in it. I’ve taught Levitt’s paper with Venkatesh. I like it. And I certainly agree with Jeremy that they do attack important problems along with their less important problems. My general problem with them can be summed up in a phrase that I got from Shamus: they’d rather be clever than right.

    This doesn’t mean they are always wrong. Nor is it against cleverness. But it is against the insistence that the only interesting result is a contrarian one, one that cuts against the “conventional wisdom”. I think that when rightness and cleverness potentially cut against each other, D&L go with (what they think is) clever if there is any ambiguity at all. And in this case, I think they were wrong – and dangerously so.

    RE: Jeremy and the annoyingness of “any kind of logical/coherent explanation of social behavior is categorized as ‘economic.'” Hear hear. Yes, checking things with data. Only economists do that. Right. Of course. Argh.

    RE: Shamus and assumptions — somewhere in the freakonomics stuff, I remember reading the line “the data don’t lie.” Well, maybe not. But people sure as hell lie with data. The line I teach is “data speak, just not for themselves.”


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s