On the Fallacy of Forward Linkages : A Note in the Light of Recent Results

Following on from de Mesnard's (2009) radical criticism of the Ghosh supply-driven model, this paper draws the dramatic consequences for the widespread use of forward linkages in input-output analysis applied to regional science: the practice must be abandoned. The arguments are based on three points: (i) it is impossible simultaneously to choose the Leontief model for the backward effects and the Ghosh model for the forward effects; (ii) it is impossible simultaneously to consider a production function of complementary inputs (Leontief) and a production function of perfectly substitutable inputs (Ghosh); and most importantly (iii) price effects and output effects remain inextricably mixed in the Ghosh model, so precluding its use for analyzing forward effects on quantities or prices