The geography of invention in high- and low-technology industries: Evidence from the Second Industrial Revolution

Production in "technologically-mature" industries has recently relocated from more-developed to less-developed countries with lower labor costs. It is not clear, however, if these latter countries will realize corresponding increases in their generation of new technological knowledge. More generally, we do not fully understand the sources of geographic clustering in invention, or how prevalent and persistent such clusters are. To investigate these issues, this dissertation explores the geographic patterns of invention in the American shoe, textile and electric industries during the Second Industrial Revolution. The three industries offer intriguing contrasts: two traditional labor-intensive industries, one whose production migrates to a low-wage area and one that does not; as well as an industry based on a radically new technology. Using both U.S. patent records and inventor information from census manuscripts and city directories, I find that the location of invention does not appear so directly, or closely, related to the location of production. The shifts in production capacity were not followed by corresponding increases in invention. Even in shoes and textiles that were more craft-based, a significant number of inventors had no production experience, and were instead distinguished by possessing a high level of technical skills. There were also sharp contrasts in the geographic patterns of where invention was taking place, and in the characteristics of the inventors, between the traditional and the new technology industries. The spatial association between invention and production was generally weaker in electric than in either shoes or textiles. Moreover, electric inventors were far more educated, younger, and geographically mobile. The intriguing implication is that because individuals with the appropriate skills to contribute to new technology are often young and scarce in supply, they will tend to migrate to those areas where demand for the technology (and rents to their scarce human capital) is high and resources to support the R&D are available. The historical evidence appears to suggest that invention and production might not be clustered in the same location. This may be unwelcome news for developing countries that hope to emerge as centers of invention after having attracted shifts in manufacturing capacity from developed countries.