The great virtue of mathematical reasoning in economics is that by its precise account of assumptions it becomes crystal clear that application to the ‘real’ world could at best be provisional…

This view which I have always held has however been severely strained by relatively recent developments... For instance a recent macro-text starts with the mathematics … for a single agent which seems distinctly peculiar for the macro-enterprise which is to distill something useful out of economic analysis. Of course these macro-economists are not mathematical economists. They use mathematics to be found in texts but are quite unrigorous in their analysis of what would have to be the case for their exercises to be applicable. Nonetheless it looks as if the scientific air of mathematical reasoning has misled them into believing that they are saying something scientific.