Abstract
The impact of regional R&D on productivity, as well as the comparative productivity effects of R&D from the business, government and higher education sectors remains understudied. The objectives of this research are to: (i) compare productivity effects of own R&D and regional knowledge from business, government, and higher education research at the NUTS-2 region levels; (ii) Establish whether business R&D, government R&D and higher education research funding are complements or substitutes.
I estimate a standard Cobb-Douglas production function augmented with own-R&D capital stock from the Business (BERD), Higher education (HERD) and the Government sectors (GOVRD). I estimate both a static and dynamic model of the Cobb-Douglas function. The static model is estimated using OLS with time and regional dummies and the dynamic model is estimated using the system generalised method of moment (System-GMM). The study reports several findings. First, the static results indicate that total R&D has a positive and significant effect on regional productivity. However, with the R&D components (BERD, HERD, and GOVRD), only government R&D was positive and significant. Second, the dynamic results confirm the findings of the static results that only government R&D has a positive and significant effect on productivity. I also find that: (i) regions with higher R&D intensity enjoy higher productivity levels; (ii) the R&D types can be considered as substitutes as opposed to complements.