Lobby journalists and MPs have reacted with bewilderment at the Treasury’s use of a number of complex mathematical equations to underpin their argument for Brexit. This much vaunted analysis was based on a “gravity model”, a device used to model “trade flows between two countries as a function of economic variables such as GDP, geographic..and cultural variables”.
Well, the Treasury’s gravity model has been discounted numerous times for its use in predicting the impact that large scale economic changes will have on the economy. In 2005 Patrick Minford, Vidya Mahambare and Eric Nowell first looked at how useful gravity models were in predicting the effects of Brexit. Their conclusion was that they were far too simplistic:
“The gravity model may work numerically, and be more accurate in detail, for quite general changes in conditions, like a general drop in transport costs, mirroring globalisation, which is what Costinot and Rodiguez-Clare use it for. The problem with using it for a shock to trade structure like the UK leaving a customs union is that the responses will certainly not be the same as for a general globalisation shock; indeed such a shock changes the UK’s internal structure substantially, in a way not assumed in a gravity model.”
Essentially, the gravity model can’t take into account the numerous complexities that would arise in the event of Brexit. These sentiments are reiterated by Alan Deardorff, Professor of International Economics at the University of Michigan. Deardforff remarks in this paper on gravity models, that their sheer scope mean their “use for empirical tests of any of them is suspect”. So it looks like Osborne’s been too clever by half. His model has been long discredited. You can’t use a few trade equations to predict the future, they are only good for baffling hacks…