Published online by Cambridge University Press: 19 September 2017
Does innovation proceed from the top down or the bottom up? This is a crucial question for those who think about the sources of economic growth and especially for those who think about policies and institutions to promote innovation. The answer lies in part with the structure of the existing system of production and the array of assets that an innovation would displace, especially on the extent of complementarity and modularity in that structure. But it also depends on institutions. This paper argues for the centrality of decision rights to the process of innovation. Especially if it takes place in a systemic, non-modular way, innovation may require unified decision rights, often implying integrated control of complementary stages of production, in order to overcome the dynamic transaction costs of change. But the processes of subdivision, differentiation, and learning – the processes of fission, forking, and fine tuning – may also require changes in decision rights in order to overcome dynamic transaction costs. I illustrate these points with a case study of three generations of an American family of inventor-entrepreneurs in electricity and electronics.