We’ve been slightly obsessed with Deirdre McCloskey on TVHE for quite a while now but only just got around to reading her book, ‘The Rhetoric of Economics’, recently. The central premise of the book is that economists write to persuade, so we can use the theory of rhetoric to analyse economists’ writings and arguments. She describes many of the techniques that economists use to persuade others and what struck me is how different the dominant methods of persuasion are between academics and practitioners.
Academic economists tend to persuade largely through appeals to authority (references) and aesthetics (theory). They look down on computational work as insufficiently rigorous and too dependent on parameter selection. The lack of generality in computational results does not appeal to their tastes and they find it unpersuasive. I acknowledge that you can point to the entire field of econometrics as evidence to the contrary; however, there is a reason economists quip so often about the disregard they have for empirics that contradict the theory.
In contrast, economists outside academia tend to make far fewer appeals to theory or authority because those techniques are not persuasive to a lay audience. Unless one spend years marinating in economic theory its pronouncements carry little weight. Most people want to hear facts backed up with anecdotes that appeal to their personal experience. That leads professional economists to rely heavily on estimation and simulation to persuade their audiences. Those techniques generate firm predictions with a veneer of scienciness that lends them authority. Anecdotes then provide a comprehensible story that allows people to interpret the predictions. Obviously these descriptions are extreme generalisations but I think they represent at least the relative tendencies of each group.
Practitioners and academics tend to have a mutual disdain for each other. They accuse the other of being either irrelevant to policy decisions, or insufficiently sophisticated in their arguments. Could it be that a large part of the divide between them is due more to their implied audience than the type or subject matter of their investigation?