Overall, I like the (often counter-intuitive) insights that micro-economics and game theory provide. Steven Levitt and Stephen Dubner have become popular authors, in this area, with their book Freakonomics and more recently, SuperFreakonomics. Yet, this later book has come in for some strong criticism for how it tackles climate change, and how it uses statistics generally.
Even without these reviews, I was not inclined to read SuperFreakonomics after being rather disappointed by Freakonomics. (If you want an enjoyable and insightful read, I thoroughly recommend Game Theory at Work by James Miller.) Below comes from a review of Freaknomics that I wrote a couple of years ago:
We should also be very careful on how to interpret statistics as a guide to action. Specifically, I believe that the inappropriate use of averages to be the most common cause of analytical error in the social sciences.
To illustrate my argument, consider the following quote from the popular Freakonomics (page 2005) by Steven Levitt & Stephen Dubner (page 142).
“As it happens, economists have a curious habit of affixing numbers to complicated transactions. Consider the efforts to save the northern spotted owl from extinction. One economic study found that in order to protect roughly five thousand owls, the opportunity costs – that is, the income surrendered by the logging industry and others – would be $46 billion, or just over $9 million per owl.”
There are all sorts of reasons to question the principles and accuracy of this calculation, but I particularly want to query the meaningfulness of the average of “$9 million per owl”. The northern spotted owl is an endangered species and it is critical to know, approximately, what the minimum viable population is thought to be (and likely margin of error!). If the numbers of such owls is at risk of being threatened, we are not talking just about these 5,000 owls, but also their offspring, and their offspring after that, and so on – for eternity. Is that an infinite number of owls? Clearly it now becomes impossible to calculate a value per owl into the indefinite future at this “phase transition” number of owls. If the number of owls is much larger than the minimum viable population size, then it would not matter very much, from an ecological systems point of view, if the logging industry took some of their habitat. If the number of owls were very small, they would tragically be doomed to extinction anyway.
We should do all that we can to protect biodiversity which approaches this critical “phase transition”.
As a footnote, I have just come across this wonderful critical review of Freakonomics by John DiNardo (pdf).
As a footnote, I have just come across this wonderful critical review of Freakonomics by John DiNardo (pdf).
