Review by Tim Harford
Published: February 23 2008 00:17 | Last updated: February 23 2008 00:17

Not long ago three professors, Daniel Ariely, Elie Ofek and Marco Bertini, set up a stall to hand out free cups of coffee at the Massachusetts Institute of Technology (MIT). In exchange, they asked patrons to tell them whether they liked the roast.

Ariely and his colleagues set up a table of condiments – milk and sugar, but also obscure offerings such as cloves and orange peel. Nobody ever sampled the unusual options, but they turned out to matter a great deal. Some days, the cloves and orange peel were presented in glass containers on a brushed-metal tray, on other days they were dumped in Styrofoam cups with hand-scrawled labels. The presentation of the “condiments-not-taken” turned out to make a big difference as to how MIT students thought the coffee tasted.

This is a typical anecdote from Ariely’s book, Predictably Irrational: inventive, lovingly described, verging on the trivial, and yet something I immediately wanted not only to tell people about but to try myself. In future, I vowed, I would be sure to present wine in a carafe, and dinner on fancy plates.

Ariely’s book is an accessible account of his own research programme, drawing occasionally on the insights of others. His aim is to show that our choices – usually, but not always, in commercial settings – are irrational but predictable. He does this by conducting psychological experiments, sometimes carried out in the laboratory, often in more real settings.

The book has a lot to recommend it. Ariely is a more than capable storyteller, and he sticks close to his own research so his writing is full of colour and detail. He also has a knack for conveying the rigour of the experiments without brandishing too many technicalities. And although he is pursuing a consistent theme throughout, there is a fresh insight in every chapter.

I could scarcely imagine a better introduction to “behavioural economics”, a discipline of growing influence that sits on the boundary between economics and psychology. But opinions differ among economists as to whether behavioural economics seriously challenges the long-held basic assumption of economics that we make rational choices, or whether it merely illuminates some fascinating but relatively minor human foibles.

Ariely’s research shows that our perceptions of a good deal can be hugely influenced by marketing tricks; that sexual arousal changes the way we make decisions, and that we are not good at anticipating this; that we can be confused by the fear of losing options; that the placebo effect is partially dependent on our perception of price. There is much else of interest. And there is plenty there for the economic traditionalists – I am one of them – to chew on.

Yet a question remains over how much we can generalise from these experiments. I have long been persuaded that the evidence shows we are fundamentally rational creatures when it comes to most of the decisions that really matter. Predictably Irrational did not change my mind about that, partly because it tended to steer clear of the bigger questions.

That may have been wise. When Ariely attempts to generalise from his experiments, his conclusions are far less satisfactory than the clever experiments themselves. One example is his work on sexual arousal: having painted a memorable picture of undergraduates answering survey questions while masturbating over a laptop encased in cling-film, he recommends first that teenagers should carry condoms, and second that they should steer clear of situations where they may become aroused. Wise advice, perhaps, but something of a stretch from the experiment itself.

Another example is his work on how choices confuse us. Again, this is based on a laptop experiment. Ariely explains that when his subjects were paid pennies to click on computer icons, they clicked – irrationally – on unprofitable icons, because doing so kept (useless) options open. That suggests a broader message, but when Ariely draws conclusions about how we should choose careers, or boyfriends, he is putting more weight on his own experiment than it can bear.

Fortunately, Ariely does not spill too much ink on heroic conclusions, preferring to describe with charm his relentlessly creative experiments. For anyone interested in marketing – either as practitioner or victim – this is unmissable reading. Other readers will be engaged and looking forward to a sequel. If only more researchers could write like this, the world would be a better place.