Wednesday, November 11, 2009


Just finished Superfreakonomics. It got so much bad publicity beforehand that I started it expecting it to be awful.

I was not disappointed.

Freakonomics was fun, largely because some of Levitt's early research was fun. But there isn't any serious research in this book - just a bunch of over-egged contrarian or counter-intuitive stories, purporting to show how economics can come up with surprising and irreverent insights.

So prostitution can be studied using the tools of microeconomics. Well, well, well. Hardly surprising, since prostitutes have one thing in common - they're doing it for money. If economics had nothing to say about the dynamics of prostitution, that would be shocking. Far more shocking than a chapter in this kind of book dedicated to paid sex.

The story of Dr Semmelweis is a nice one - although rather boring if you'd already read Celine's rather better written account of it - but doesn't have much to do with economics. Oh, it does have a rather dramatic story about the use of the scientific method - Semmelweis performs an informal ANOVA with two hospital wards - but perhaps Levitt and Dubner shouldn't claim a patent on that for the moment.

The big question for me is: what is the point of Freakonomics, assuming it means anything? Well, for L & D, their approach is based on three ideas: 1) people respond to incentives, 2) behaviour change is difficult, and 3) cheap and easy solutions are often around but ignored because of people's reluctance to believe they're possible.

1) is not that worrisome, until  - you start to wonder what kind of incentives we are talking about. Writing a sequel to Freakonomics is easy to explain - the incentive is the big pile of money to be made building on a successful franchise. People sometimes respond to other kinds of incentives, which may be inconsistent with material gain - the desire to impress others, or even oneself. This is accepted by the authors, but in the end their critique of experimental economics suggests that they do not believe that altruistic behaviour is that common. The proof? People don't give up their kidneys to save complete strangers.

2) Sure, behaviour change is difficult. We know this, because we tend to get entrenched in behaviour we know we should change - in my case, never returning library books on time, thus enriching my local libraries with completely unnecessary fines (maybe that's an altruistic incentive?). But if people respond to incentives, why don't they change their behaviour when the incentives change? Mmmm.

3) People respond to incentives, but they ignore cheap and easy solutions in favour of difficult and expensive ones. Why would they do that? Maybe because we know there is no such thing as a free lunch? Or because we don't have enough faith in incredibly smart people like L & D's friends at Intellectual Ventures? Maybe responding to incentives doesn't explain everything we would like it to.

Don't take my word for it, there are plenty of critical reviews out there: the best so far is this one.