1. ¹University of Rome 'La Sapienza', Rome, Italy
2. ²Katholieke Universiteit Leuven, Leuven, Belgium

Correspondence: Reinhilde Veugelers, EC-BEPA, KULeuven and CEPR-Fellow, Naamsestraat 69, Leuven 3000, Belgium. Tel: +32 16 32 69 08; Fax: +32 16 32 67 32; E-mail: Reinhilde.Veugelers@econ.kuleuven.be

Received 22 August 2003; Revised 8 June 2006; Accepted 5 August 2006; Published online 18 January 2007.

Abstract

This paper provides a theoretical model of the trade-offs that an MNE faces when organising its R&D as decentralised or centralised. R&D decentralisation avoids having to adapt centrally developed innovations to local markets, being able to use the specific know-how of the subsidiary. In addition R&D subsidiaries can be used to source locally available external know-how. At the same time, however, R&D internationalisation intensifies the spillover of valuable know-how to competitors located in the foreign markets. The analysis demonstrates the importance of the intensity of competition in the local market in determining the size of both the benefits and costs of R&D decentralisation. It shows that when R&D is undertaken abroad in association with production, the local knowledge base is not unequivocally a pulling factor attracting R&D investments by foreign MNEs, depending on the level of local competition. The paper also shows that efficiency in reverse intra-company technology transfers is a critical factor in benefiting from technology sourcing. The results thus illustrate the complementarity of efficient internal and external knowledge management systems. In addition the model suggests that, with a fall in the cost of intra-company technology transfers, relative market size loses importance as a locational factor for R&D decentralisation.