Technical Change Theory and Learning Curves: Patterns of Progress in Energy Technologies

Better understanding of the role of learning by doing and learning by research at different stages of technical progress is important for development of a theory of innovation and designing effective technology policies. This paper presents a comparative analysis of energy technology learning and progress within the framework of Schumpeter’s invention-innovation-diffusion paradigm. We estimate learning by doing and research rates for a range of energy technologies in four stages of technical progress. We find learning patterns that are broadly in line with the perceived view of the process of technical progress. Emerging technologies respond slowly to research and development (R&D) and capacity expansion. Evolving technologies exhibit high learning by doing and research rates. Reviving technologies exhibit considerable response to learning by research although they do not face significant market constraints. Mature technologies exhibit similar learning characteristics to emerging technologies. The results point to relative importance of R&D in the process of technological progress. We generally find higher learning by doing than learning by research rates but do not find any development stage where learning by doing alone is the dominant driver of technical change. Also, high capital intensity and market constraints appear to slow down the pace of progress of emerging and evolving technologies. We find little scope for potential substitution between learning by doing and learning by research across the technologies and different stages of their development path. It remains crucial to understand how a technology can be supported in its transition from the emergence stage to a self-sustaining growth path.