R&D spillovers and patent production of British, French, and German firms

This paper investigates effects of regional research and development (R&D) spillovers on firms’ patenting performances. The localized knowledge pool is well accepted as one crucial source of externalities contributing to firm innovativeness in agglomeration economies (Rosenthal and Strange, 2004). In practice, the pool can be measured by R&D investment of geographically or technologically proximate firms. The analytical framework proposed in Bloom et al. (2010) draws our attention to important causal relations between R&D spillovers and firm patent production.

Bloom et al. analyzed a two-stage game of R&D and patent production in which one firm’s R&D could bring up two countervailing spillover effects to other firms with different distances in technology and product market spaces. The R&D activities of firms in close technology space provide positive spillovers effects, while R&D by product market rivals lead to negative effects. Extending on these results, this paper proposes that for firms locating in regions with sufficiently beneficial spillovers, the product market rivalry effects can be overcome by the technological spillover effects, leading to positive impacts of R&D spillovers on firm patenting. Wilson (1977) found that two key technological dimensions, opportunity and complexity, affect the degree of product rivalry and patenting activities since firms will react to the environments by making R&D efforts and participating in innovation activities. This paper supposes that the two dimensions of technological environments are different across a range of industries. Consequently, the intensity of product rivalry and R&D spillovers on firm patent production should vary across different sectors.

Since Jaffe (1989) and many others, the empirical work on relations between knowledge spillovers and patenting has suggested that knowledge flows are geographically clustered, and that innovation activities are concentrated in space. Still, comprehensive analyses and micro-data evidence on geographical determinants of firm innovation are less jointly explored in the literature (Feldman and Kogler, 2010). Besides, limited empirical studies focus on cases of top R&D investing firms in European Union. According to the 2006 EU industrial R&D investment Scoreboard (EU-R&D Scoreboard), regions in the United Kingdom (UK), France (FR) and Germany (DE) are locations of most heavy corporate R&D investors which account for 72 percent of total R&D activities. To shed more lights, this paper aims at gauging how the localized knowledge pool and sector specific R&D spillovers influence innovation performances of UK, FR, and DE firms listed on the EU-R&D Scoreboard. The patent production function is established based on the theoretical framework proposed in Bloom et al.

Using panel data of UK, FR, and DE firms locating in the NUTS 2 regions, this paper applies a GMM panel count model to deal with the potential endogeneity problem arising from the presence of firm’s R&D investment. Results show that the R&D spillovers positively affect patenting for UK and DE firms on average, whilst these effects vary by sectors. The negative spillovers effects for firms in UK and FR bio-pharmaceutical industries and for those in DE telecommunications industry imply that the product rivalry may be intensive in these industries.