Journal of Economic Geography Advance Access originally published online on July 30, 2007
Journal of Economic Geography 2007 7(6):737-752; doi:10.1093/jeg/lbm027
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Rethinking the regional knowledge production function
*Corresponding author: School of Geography Sciences, Arizona State University, Tempe, AZ, 85287-0104, USA. email < ohuallachain{at}asu.edu>
**Department of Geography, George Mason University, Fairfax, VA, 22030, USA. email < tleslie{at}gmu.edu>
JEL classifications:: O18, 031, R11
The knowledge production function is a questionable device for identifying the determinants of commercial patenting. Adapted from firm level analysis, it implicitly assumes that some mechanism exists to transfer research and development inputs across institutions to maximize invention output. Augmenting the approach with regional structure variables risks confounding causes and effects, because these conditions determine research and development in the first place. We contrast a knowledge production function and a regional structure model to identify the determinants of commercial patenting in US states in the period 2002–2004. Results show that human capital, specialized knowledge flows, urbanization and industry dominance in technical advance drive commercial patenting.
Keywords: invention, spillovers, regional economic development
Date submitted: 15 February 2007
Date accepted: 29 June 2007