Agglomeration and automation in the twenty-first century: Prospects for regional research

Abstract

This paper discusses the future of technology and its implications for regional agglomeration and related research. In regions with strong agglomeration economies, positive externalities generated by spatial proximity lead to higher productivity. Urban scale economies raise the productivity of people residing in dynamic cities. Firms benefit from regional clusters of similar firms. Now we are entering a second machine age, with potentially exponential growth in productivity based on pervasive, interconnected computing power, massive digital information, machine learning, artificial intelligence, and robotics. Productivity may depend more on the “cloud”-based advantage of machines, rather than “crowd”-based advantages of humans. Employment may fall with the automation of routine tasks. However, non-routine tasks requiring tacit knowledge may rise, given comparative advantages with machines. Machine learning and robotics will both substitute and complement employment. Agglomeration benefits may then accrue to regional employment based on non-routine tasks. Even so, the size and stability of agglomeration benefits will require further inquiry. With better measures of agglomeration, future research will potentially delve deeper into the three main sources of regional productivity from agglomeration: (1) better labor skill matching and local pooling; (2) more efficient input sharing along deeper and specialized supplier relationships; and (3) enhanced localized learning and knowledge flows.