An alternative approach towards the knowledge production function on a regional level - applications for the USA and Russia

A knowledge production function is used to account for the efficiency of an innovation systems; it provides a quantile regression estimation of the knowledge production function to account for a nonlinear relationship (efficiency) between knowledge inputs and knowledge output. Regional data for researchers, expenditures on R&D and patent grants for the USA and the Russian Federation is used in two different designs. In both countries, groups of regions exist with smaller sized research systems that report significantly different dynamics and thus knowledge production functions than regions with larger sized research systems. In the USA, regions with small research systems report decreasing returns to knowledge inputs while region with large research systems report increasing returns to knowledge inputs thus efficiency is achieved in fostering larger regional innovation systems. In the Russian Federation, the effect is exactly the opposite; it is the smaller regional innovation systems that are most productive.