I have just finished Animal Spirits by Akerlof and Shiller. I am glad that I read it, but not for the reasons that I would have expected when I started it. I do not feel particularly enlightened about behavioural economics, but then I am not an economist and I didn’t really need to be. The book is interesting because it gives a good impression of a debate that needs to happen between economists.

The role of psychology in markets should be obvious to any financial professional who has done even a modest amount of thinking and reading. The notion that markets are efficient, and the related idea that they can operate without very much regulation, are so transparently daft to someone who follows the markets day-to-day as to appear straw men. Why then have two distinguished economists written a book to knock them down?
Systematic understanding requires models of the world. Macro-economists have for some years built their models with the assumption that people are motivated by economic considerations and act in their own rational self interest. This approach is not necessarily shown to be valueless by the fact that individuals that we encounter have different motivations and act in a variety of ways; it may be that there is enough truth in the modellers’ assumptions as they apply to our aggregate behaviour to allow their models to make reasonably accurate predictions about the economy, and therefore provide useful advice to fiscal and monetary decision-makers.
The point of this book is therefore not to show that individuals are not the ideal economic agents of macroeconomic models, but that the behaviour of human beings, individually and in aggregate, is sufficiently far from that ideal that it cannot be the basis of a sound model. In doing this the authors shed some light, for the non-specialist, on the arguments that economists need to have between themselves in order to improve the field of macroeconomic modelling.
At times I felt that the book was rather wooly, and I have had a stab at understanding why. Akerlof and Shiller quote various pieces of research to back up their views of human nature. But the research still only covers a few people, and it is a leap to generalise its conclusions to aggregate behaviour, even though the generalisations seem to untangle some of the knots that are created by models based on the rational-actor view (it is not established that there will not be greater problems with new models developed from their ideas). The authors are making an appeal to fellow economists to recognise that their existing models are not good enough, arguing that the reason for this is that those models are built on poor assumptions, and suggesting what seems at present to be a plausible alternative set of assumptions to be used in a new model-building effort. An exercise like this is more philosophy than science, and that is why the book feels wooly if you read it expecting the latter: it is a manifesto for a new paradigm, not a new paradigm in itself, because the work of building and testing those new models is yet to be done. After reading the book, I find the appeal persuasive, and hope that the it will be done.