The dominant practice in economics is to choose the mathematical specification of model relations on the basis of convenience, without much theoretical support. This paper discusses how quantitative model specifications can, in some cases, be given a more formal scientific underpinning in the...
