Introduction to Modern Economic Growth Proof Much of the proof is provided by the preceding analysis Exercise 14.15 asks you to verify that the average growth is given by g∗ = ηL∗R ln λ and that (14.27) is necessary for the above described equilibrium to exist and to satisfy the ¤ transversality condition Therefore, this analysis shows that the basic insights of the one-sector Schumpeterian model, as originally developed by Aghion and Howitt (1992), are very similar to the baseline model of competitive innovations presented in the previous section The main difference is that growth has an uneven flavor in the one-sector model, because it is driven by infrequent bursts of innovation, preceded and followed by periods of no growth 14.2.2 Uneven Growth and Endogenous Cycles* The analysis in the previous subsection showed how the basic one-sector Schumpeterian growth leads to an uneven pattern of economic growth This is driven by the discrete nature of innovations in continuous time There is another source of uneven growth in this basic model, which is more closely related to the process of creative destruction The nature of Schumpeterian growth implies that future growth reduces the value of current innovations, because it causes more rapid replacement of existing technologies This effect did not play a role in our analysis so far, because in the model with a continuum of sectors, growth takes a smooth form and as Proposition 14.2 showed, there is a unique equilibrium path with no transitional dynamics The one-sector growth model analyzed in this section allows these effects to manifest themselves To show the potential for these creative destruction effects, we now construct a variant of the model which exhibits endogenous growth cycles Throughout, we focus on an equilibrium path with such a cycle The only difference is that we now assume that the technology of R&D implies that LR workers in research leads to innovation at the rate η (LR ) LR , where η (·) is a strictly decreasing function, representing an externality in the research process When more firms try to discover the next generation of technology, there will be more crowding-out in the research process, making it less likely for 627