Capital and interest in horizontal innovation models

Horizontal innovation models have a common structure of three sequentially connected sectors. This structure-production of commodities by means of commodities-necessitates the compounding of interest on an input that goes through multiple production periods before the final good is produced. I argue that this aspect is missed (or deliberately assumed away) in typical horizontal innovation models and that this practice generates internal inconsistency in relation to the long run nature of the models. Though discussion is carried out in particular reference to Barro and Sala-i-Martin's 'lab-equipment' model, implications are general.