Model robustness in "old" and "new" growth theory

In economics, models, rather than theories, seem to be the fundamental units of appraisal and practitioners seem to hold in high esteem the criterion of 'robustness'. In this paper we shall try to explicate the multifarious notion of robustness, and articulate it on three different dimensions. In order to show their relevance to concrete economic practice we shall apply these notions of robustness to a particular case: the "old" vs. "new" growth theory. Special attention will be paid to the robustness of the implied causal mechanisms, due to its substantial role concerning the possibility of deriving sound policy prescriptions from models