Internationalisation of pharmaceutical R&D: how globalised are Europe's largest multinational companies?

Abstract

The internationalisation of corporate research is an issue with a large impact on strategic decision-making within the research-intensive pharmaceutical industry. However, sector-wide comparative statistics on the whereabouts of corporate research activities and the geographic distribution of their research partners is scarce. This paper describes the international research profiles of Europe's top 10 largest multinational pharmaceutical companies based on publication counts and author address information drawn from the thousands of pharmaceutical research articles published in 2005–2006. The results provide relevant aggregate-level background intelligence for strategic analysis of the industry's internationalisation processes and research cooperation patterns. The overall statistics indicate that international research activity of European industry is significantly more oriented towards the US science base than vice versa. Company-level statistical data on the research cooperation involving each company's central research centres based in country of corporate headquarters, reveal quite distinctive geographical characteristics alongside differential preferences for in-house or external research partners.

Keywords:

• business and management studies
• technology and innovation studies
• bibliometrics
• quantitative
• pharmaceutical industry
• Europe

Acknowledgements

The author is grateful to Erik van Wijk, Thed van Leeuwen and Bert van de Wurff for data handling and the many valuable comments and suggestions by an anonymous TASM reviewer are also highly appreciated. This research was partially funded by the European Commission DG RTD, Sixth Framework Programme through the projects ‘Improving Human Potential and Socio-Economic Knowledge Base’ Programme (project ‘STI-NET’) and ‘Analysis and Studies and Statistics and Indicators on Science and Technology’ (project ‘Regular Collection of Bibliometric Indicators’). The auxiliary data on the output of research papers dealing with phase III clinical trials was derived from the author's study funded by the Netherlands Association of Research-based Pharmaceutical Industries (NEFARMA). An earlier version of this paper was presented at the EUI Workshop ‘The Political Economy of Research’, European University Institute, Fiesole (Italy), 18–19 May 2007.

Notes

According to OECD's Frascati Manual, Proposed Standard Practice for Surveys on Research and Experimental Development (2002), clinical trials are defined as ‘development’ rather than ‘research’.

However, these advantages may come at a cost in terms of disincentives to innovate domestically, a loss of R&D capabilities to other research affiliates, less focus on local networks, diminishing absorptive capacity, and increased co-ordination and transaction costs to manage research elsewhere within the company.

The results of the Thursby and Thursby survey indicate that decisions on where to locate R&D are complex and influenced by a variety of factors. Regardless of where companies locate R&D, the quality of R&D personnel, university collaboration, intellectual property protection and market potential, stand out. How these factors affected decisions varied, depending partially on whether the R&D sites were based in emerging economies or in developed economies.

Harmonisation of company names was done at the country level using the name of the local parent company. The findings in this paper are listed by the standardised name of the global parent company, or a sufficiently distinctive generalised name.

The sets of research articles in the study deal only with clinical research in phase III that involved ‘controlled randomised clinical trials’ that were extracted from the Medline database according to the query based on Medline's MeSH terms ‘clinical trial’ and/or (‘pharmaceutical preparations’ or ‘drug’ (Tijssen 2006). The 12% fraction is a lower bound of all articles reporting on clinical trials in phases I, II or III. The relatively low level of coverage of these articles within the journal literature indexed by the Thomson Reuters’ Web of Science database is largely explained by differences in knowledge dissemination practices where results of the preclinical ‘discovery’ research are more likely to be published in the peer-reviewed international journals whereas results of clinical trials are often made available though other publication outlets or stored in open access repositories and databases (specifically those for physicians and patients). Note that several specialised journals dealing within clinical trials are not indexed by the Web of Science database.