One-sector R&D-based models predict scale effects, which is empirically inconsistent. This is due to the "knife-edge" assumption that new ideas created are linear in the stock of knowledge. If this assumption is dropped to make a one-sector R&D-based model consistent with data, growth becomes...