Coruscation

The abstract models of economic theory are intended to represent things like the labour supply under various tax regimes, for example. But abstract models, by their very nature, require idealization and simplification in order to be manageable and useful. Every good map must leave off many features of the terrain. But this raises the possibility that one has left off the wrong features, mistaking the essential for the inessential. A "general theory" will not make headway against real economic problems if we have overgeneralized by leaving out complexities that really do matter.

So what to do? If an abstract model is supposed to be a model of the real world (and not just a mathematical construct), then we should test its assumptions and predictions against the real world. But the real world, in all its boisterous glory, is far from the serene desert landscape of an abstract model. So we must cleverly search out "natural experiments"--real-world circumstances that happen to conform to enough assumptions of a model to provide a relatively direct test of its predictions--and then apply advanced statistical techniques to isolate the effect of the variables we want to study. The result of such tests can lead to tweaks in abstract theories that improve their predictive power and utility.

Especially "cute" or "freaky" analyses, such as many of Levitt's, may not directly challenge or confirm aspects of general economic theories. But they may usefully advance the techniques of testing, which will eventually help improve general theories. If different economists specialize in different but complementary jobs along the assembly line of economic knowledge, we are more likely to get more, and better, economic knowledge. We are also likely to get better pop economics books.

An earlier generation of these books, like Steven Landsburg's The Armchair Economist and David Friedman's Hidden Order, tackle the economic puzzles of everyday life by applying good old-fashioned price theory to novel situations. Many of the new spate of pop-econ page-turners reflect the maturation of economics as an increasingly empirical science.

Freakonomics is the bellwether of this shift. But Cowen's new book, which may seem superficially similar to old-style pop-econ, in fact is something different. It integrates a great many of the insights of Levitt-style work, as well as insights from behavioral and experimental economics (which Lozado, confusingly, opposes to Freakonomics-style work at the conclusion of his review). Cowen's synthesis of these new insights adds up to a level of psychological realism heretofore unseen in the pop-econ genre. If Cowen succeeds in offering excellent cute-o-nomic advice, and I think he often does, it's because economics as a whole is now generating a more empirically adequate picture of the world. For those of us weird enough to love economics, that's better than cute: that's beautiful.