The clustering of innovative companies¡­creates self-reinforcing feedback loops that drive their continued growth

The recent announcement by AstraZeneca that it is to relocate its main research and development facility from Cheshire to Cambridge raised cheers because, unlike when Pfizer closed its Kent facility in 2011, UK jobs would largely be preserved. However, it also raised concerns that the move would reinforce the strength of the ¡°golden triangle¡± of London, Oxford and Cambridge, which already hosts several other multinational pharmaceutical giants, including GlaxoSmithKline and Eli Lilly.

What draws these companies? The answer, according to AstraZeneca, is geographical proximity to world-class research in the biosciences, greater opportunities for research collaboration and access to highly skilled researchers. These reasons already hint at the fact that such factors can be self-reinforcing. But until now there has been little hard evidence for this.

Our recent research suggests that moving large research facilities really does have a strong impact on the further geographical clustering of R&D activity. We studied the impact on the locality chosen to site the Diamond Light Source - a third-generation synchrotron (a circular particle accelerator). We discovered that Diamond - the largest single investment in basic research infrastructure in the modern history of the UK - created strong local effects that promoted the formation of a research cluster in close geographical proximity to the facility.

ÖйúAƬ

Quantifying the impact of large-scale scientific infrastructure is hard because the location is not selected at random. Instead, policymakers, like private companies, place the facility within an existing hub of expertise. In the case of Diamond, the Rutherford Appleton Laboratory on the Harwell Science and Innovation Campus near Oxford was chosen. This then poses the challenge of how to separate the effects of the existing cluster from any additional effects created by the new facility.

The controversy in 1998 over the siting of Diamond - and, in particular, the existence of a ¡°runner-up¡± location at the Daresbury Science and Innovation Campus in Cheshire - presented an opportunity to study this.

ÖйúAƬ

The government initially planned to build the synchrotron at Daresbury, which was home to its predecessor, the Synchrotron Radiation Source. But fierce debate led to a change of heart and a final decision in 2000 to locate the new facility at Harwell. This sparked a great deal of controversy as it was perceived to reinforce the North-South divide in investment in scientific research.

We analysed how scientific research, measured by the number of journal publications, was affected by the opening of Diamond in January 2007. Comparison of publications by researchers near Diamond with those near Daresbury between 2000 and 2010 shows that output from researchers in direct proximity to the synchrotron increased more than it would have done if the facility been located elsewhere.

This is the combined result of two effects, one direct and one indirect. The direct effect results simply from ease of access for nearby researchers who use the synchrotron. The indirect effect is perhaps more surprising. We discovered an increase in the number of papers even when that research made no direct use of Diamond. Our explanation is that these indirect effects arise as a result of the localised knowledge spillovers that occur when researchers learn from each other through personal contact or via seminars, presentations and workshops held in the vicinity.

Similar mechanisms have been found in numerous studies to drive the clustering of innovative companies and the co-location of universities and industry. It is also well known that such clusters create self-reinforcing feedback loops that propel their continued growth by attracting more and more private and public investment.

ÖйúAƬ

In short, if AstraZeneca¡¯s relocation generates similar effects to those of Diamond - and it seems reasonable to assume it will - then it is certainly possible that it will further strengthen the golden triangle. This may be good for UK plc, but it also means that the North-South divide will keep widening.