Superclusters – the UK may need to think big in its Silicon Valley

Superclusters – the UK may need to think big in its Silicon Valley

By Alice Frost, Director of Knowledge Exchange, Research England .

9 October 2020

A universities-Research England group is currently exploring whether London is an “ecosystem” of the likes of Silicon Valley or Boston in the USA. We published in February our initial report with our model of an entrepreneurial ecosystem:

Ecosystem model


Simplified ecosystem model


Notwithstanding the challenges and disruption of Covid-19, we believe that this topic remains vitally important as part of economic recovery, and we are seeking ecosystem builders elsewhere to compare approaches. Although America is home to a number of the most prominent ecosystems, other US cities seek ways to join this route to economic development[1] and, the American government[2] is currently exploring how to support these further.

We publish today a study from SQW with US and UK university-city-ecosystems matching detail. There are limitations to this work - notably any method is limited to metrics available across countries. Given this was a pilot, the approach was minimalist, and clearly attention to a multitude of dimensions would improve match-making. Our intention is to build on the study through conversations with comparators, including our initial target New York, to test for the most insightful comparators, including how to pick the “right” footprint of any ecosystem. Our study highlights the sensitivity of this - as example, picking San Francisco over the west coast Bay area.  We note that American counterparts seem to pick their ecosystems differently for different technologies, for example a larger footprint for biotech with its need for R&D space.

Subject to these caveats, what leaps out in the report is the scale of America (Table 3-1 of report), in all dimensions - its research, universities, entrepreneurial outcomes, size of cities and economic activity. To give one example:

University R&D Expenditure (£000s, total of 2016-2018, source: SQW report)

University R&D Expenditure


To dig into more detail on scale - as well as R&D differences, the make up of university sectors across the two countries is very different, with London having more than double the student numbers of Boston and Los Angeles, and similar student numbers to New York whilst having less than half the total population. This reflects the higher overall levels of expenditure on research in the USA. Leading edge American universities that we see as models for the UK are actually more akin to research institutes than UK institutions whose business model is founded more on teaching. Although London produced a creditable 106 spinout companies over these three years, this performance is inevitably dwarfed by 260 in Boston, 364 in LA and 178 in New York. A focus on IP income in the same timeframe shows the same picture, with London generating £166 million to Boston’s £778 million, LA’s £1.51 billion and New York’s £1.15 billion. Although LA and New York are supercity sized compared to London in terms of population and economic output, interestingly Boston is half the population size of London, though with comparable economic output.

It is important though to distinguish between university performance and scale. London is actually more efficient in turning R&D expenditure into spinout companies, producing a spinout for every £35.35 million R&D input compared with £60.42 million, £42.60 million and £54.5 million for Boston, LA and New York[3].This is also highlighted in the report published today by Research England on commercialisation data, which compares US and UK university sector wide commercialisation performance and shows distribution of US and UK activity once normalised for scale is very similar:

IP income per institution normalised by its individual research resource, for the 50 institutions with the greatest normalised IP incomes, the UK and the US.

Why does this scale matter?  There is a wealth of evidence (though not all straightforward[4]) that suggests scale and density matter in clusters. Knowledge intensive workers congregate together. These dense connections spur innovations. There is more likelihood of the high risk-taking needed in entrepreneurship, particularly in very speculative deep tech and knowledge intensive forms, occurring in dense agglomerations. These clusters reduce the consequences - and hence fear - of failure. Entrepreneurs have many fall back options if one venture fails.  It is also easier to justify and to get value from the specialised ecosystem components that accelerate innovation and entrepreneurship if it is provided at scale.

We should not though jump to conclusions. We will consider the SQW proposal of scaling US ecosystems to match with UK. But we also want to look at Europe and Middle East comparisons - Paris, Tel Aviv, Nordic models - to understand ecosystems in smaller nations.

What might this mean for policy and practice? First, the Government’s roadmap commitment to increase public spending on R&D to £22 billion is vitally important as hitting directly on the scale issue, alongside the developments in the roadmap focused on innovation, commercialisation and regional clustering. At the practice level, we need to consider whether we have the imagination, will and connectivity to think and do big - embracing not just London, but possibly a greater south east cluster (with systems of equivalent scale needed across UK?).

Our initial worry was that London was too big to think about as an ecosystem - but possibly it is too small?

[1]  Bruce Katz and Julie Wagner, The Rise Of Innovation Districts: A New Geography of Innovation in America, (Washington, DC: Brookings Institute, 2014),

[3] Source: 2018/19 HE-BCI data (London) and 2018 AUTM – STATT Survey (LA,Boston and New York)

[4] For example: Cooke, Philip & Asheim, Björn & Martin, Ron. (2006). Clusters & Regional Development.

Thomas, Llewellyn & Autio, Erkko. (2020) Innovation ecosystems in management: an organizing typology.