Economic growth, a golden rule of thumb, and learning by doing

A simple rule of thumb which has been successfully used in the basic neoclassical growth model as an alternative to the unstable dynamic optimization solution is shown to be more generally applicable in a non-scale growth model with learning by doing. The model is formulated in accordance with empirical regularities about learning by doing and shown to generate the stylized facts about economic growth reasonably well. The external effects of learning by doing can be internalized by the steady state tax cum subsidy policy implied by a decentralized dynamic optimization approach.