Learning, scale effects, and (very) long-run growth

We develop a tractable model of endogenous growth that emphasises learning and the relevance of assimilated knowledge (as opposed to frontier knowledge) in research. The model is able to limit the scope of (level) scale effects, without degenerating into a model of "semi-endogenous" growth. We reassert the long-run potency of government policies that promote thrift, training, R&D and enhancement of productivity. We also show the relevance of learning for growth in the very long run, and its potential in accounting for the transition from stagnation to the secular growth of the industrial era.