The UK’s Missing Two Trillion
The Incredible Economic Promise of UK Innovation
At a recent private event hosted by a venture capital (VC) fund in London, three partners at a top-tier Silicon Valley VC fund were visiting to meet UK startups. I was struck when one of the partners disclosed that their multibillion Silicon Valley fund is allocating a hundred “partner days” to the UK this year, up from fifteen the year before. A recent article on one of the fund’s recent investments came to mind. It revealed that the Silicon Valley fund had been looking to make a substantial investment in the AI semiconductor space. Their global search for the space’s most innovative firm did not end in the US or China; it led them to Bristol.
The UK is a formidable intellectual power, and its innovation venture scene is advancing rapidly. It is still far, however, from commercialising its extraordinary intellectual property (IP) at peak. If it does, the economic gains will be dramatic. In the order of trillions.
UK IP: Intellectual Powerhouse
Much has been made of China’s meteoric rise in science and technology research –especially in strategic sectors such as Artificial Intelligence (AI)– and for good reason. China’s science and technology research publications have skyrocketed from near UK levels in 2003 to surpass US levels in 2016.
However, if we look at what truly matters –the quality and influence of the IP, and not just its quantity– a strikingly different picture emerges. The UK is a global powerhouse, punching far above its weight.
Measuring research impact, the UK consistently ranks first in the world, ahead of both the US and China. Per an established measure of research impact, the field-weighted citation impact¹ (FWCI), the UK overtook the US in 2007 to become the highest ranked, and has maintained its lead since.
In 2018, the UK’s FWCI was 11% higher than the US, a third higher than the EU 27, and over 50% higher than China and the world average. In fact, the UK has increased its lead over the US and OECD in the last few years. While China is closing its gap (its FWCI has increased by 24% from 2014 to 2018), it remains well below the UK.²
The UK’s intellectual leadership is confirmed when considering the world’s most influential publications (i.e. most highly-cited). The UK is a leader in the proportion of its research that is the world’s most influential. In 2018, the UK had 2% of its publications among the most highly-cited in the world. This was double China’s figure, over two thirds higher than both the EU and OECD, and about a fifth higher than the US. The UK has maintained its leadership in this measure since 2010³.
If we account for research productivity, together with impact, the UK maintains its lead over China, despite the vast difference in research output volume. The metric that is widely used to measure both the productivity (number of publications) and impact (number of citations) of research is the h-index⁴. According to that measure, the UK ranks second only to the United States across all science and technology fields (indeed across all fields). The UK ranks consistently higher than China, which produces around five times as many publications, on average. (In some hot sectors, such as AI, this difference is even higher.)
To gain deeper insight, let’s look at country performance by h-index in the key innovation fields of AI, computer science, engineering and medicine.
In the nationally strategic field of AI, where China and the US have deployed colossal resources, the UK is markedly more productive than the two giants in generating influential IP.
The US and China boast the largest numbers of the world’s AI researchers by far –at 28,536 and 18,232 respectively at the end of 2017, whereas the UK only had 7,998⁵. Nonetheless, and quite remarkably, the UK’s h-index in AI is higher than China’s –who had more than twice the AI researchers. The UK also recorded more than half of the US’ h-index –despite the US having roughly four times as many AI researchers (in fact, counting top researchers, the US had 4.4 times that of the UK)⁶.
For computer science generally, the UK maintains its place, second only to the US, while China drops to sixth place.
Across engineering, the picture is similar to AI, with China coming third after the US and UK despite China’s large lead in volume of publications (close to seven times the UK’s).
The UK’s influence is stark in medicine. While it comes a strong second to the US, the UK has notably higher productivity in producing impactful research. The UK’s h-index is two-thirds that of the US, but with less than a third in publication volume. China drops to the 19th slot, a notch above Brazil, with an h-index (404) that is 60% lower than the UK’s.
Even looking at sheer number of publications alone, the UK has consistently played with the heavyweights: ranking in the top 3 and lagging only the US and China. In 2018, the US produced 22%, China 19%, and the UK 7% of the world’s publications. Since 2004, the top 3 producers of publications have been the US, China and the UK (when China overtook Japan). Over the past five years the UK has maintained its share despite the significant increase in output from China. (China has increased its share from 3% in 1998 to 19% in 2018).
For the most influential publications, the UK has an even stronger showing. The top three nations remain the same, but the gap narrows. In 2018, the US produced 37%, China had 20%, and the UK had 14% of the world’s most highly-cited publications⁷.
In fact, for highly-cited publications in general, UK researchers are four times as productive as Chinese researchers and almost twice as productive as US researchers⁸.
The UK’s exceptionally prodigious intellectual impact begs the question: what is the economic potential of this extraordinary IP? And is the UK fully realising it?
The Curious Case of the Missing Two Trillion
In 2014, a Stanford study⁹ looked at the economic impact of VC-backed companies in the US. VC-backed firms are a proxy for IP commercialisation, as VC is the primary vehicle for funding such ventures. The study covered the period from 1979 (after a regulatory change allowed VC to take off in the US¹⁰) through 2013. The findings were striking.
VC-backed firms were nearly half (43%) of all public companies founded in that 34-year period, comprising 57% of the total market capitalisation and creating 38% of total jobs by these public companies. Remarkably, they were responsible for the overwhelming majority (82%) of all research and development (R&D) spending by these US public firms. To get an idea of their impact, VC-backed US public companies spent $115 billion on R&D in 2013, up from effectively null in 1979. The economic impact of VC-backed firms has only mushroomed since 2013, given the dramatic increases in the number of US unicorns –that skyrocketed by more than 5x from 2013 to 2019¹¹– and the swelling wave of VC-backed IPO’s in the last 5 years, which set a new annual record for exit value in 2019 and saw average exit valuations smash the record set in 2012 (due to Facebook)¹². In short, commercialisation of research and innovation plays an increasingly dominant role in the US economy, acting as a prime driver of economic growth, value creation and employment.
Is the UK commercialising its IP as effectively as the other two nations in its intellectual league? Unfortunately, no. Not yet. The UK considerably lags the US and China in that regard currently.
Given that the economic impact of innovation takes years to materialise at scale, it would be fairer if we didn’t judge on outcome just yet. Let’s instead take a snapshot of the respective activity taking place and assess how the UK compares.
A key metric for IP commercialisation is the amount of venture “scaleup” investment taking place. Scaleups are startups that are effectively gaining commercial traction¹³, which is what matters to our analysis, and they constitute the vast majority of startup economic activity (e.g. 80% in the UK). Globally, scaleups raised over £245 billion over the 4-year period of 2015 to 2018. 5.2% of that global total, or £12.6B, went to UK scaleups. That sounds impressive. But when compared to the US and China, the UK does not compete as it does in intellectual output.
US scaleups raised ten times as much as the UK, at £120.9 billion (49.3% of the global total) and Chinese scaleups raised four times as much, or £49.9 billion (20.4% of the global total)¹⁴. If we compare these differentials with the research outputs discussed above, there is a large gap for the UK to close. For example, for influential (highly-cited) research, the US published only 2.6 times the UK rate. While IP is not the only input into venture activity, it is a key (arguably primary) ingredient as has been evidenced by the extreme concentration of venture funding in US research hotbeds –70% of US venture funding goes to Silicon Valley, Boston-Cambridge and New York, with Silicon Valley alone attracting over 40%¹⁵. If the UK attracted venture funding in proportion to intellectual output at the same ratio that the US does, the UK could see two to four times the amounts it sees today.
The good news is that the UK is catching up. Growth in VC investment exceeded 40% for the third year in a row, well ahead of China and US rates¹⁶. Despite these rapid advances, the UK can still go much farther. (We’ll explore this below.)
I have met some who have been skeptical that the UK can ultimately perform as exceptionally in the commercial realm as in the intellectual one. They often cite structural, geographic or even cultural impediments. To those, my answer is simple. The UK has already proven it can. In fintech.
The UK has attracted almost a third of the funds that the US has in fintech, a much narrower gap than the 10x differential overall (and closer to the proportional performance in intellectual output). In 2019, the UK ranked second globally (and first in Europe) with $4.9B of investment –up 38% on 2018– across 359 deals. The US saw $16.3B in investment across 1,095 deals¹⁷. The UK has also commensurately a third as many fintech unicorns as the US, at 10 and 30 respectively¹⁸. (Other sources have cited different fintech unicorn numbers, suggesting an even narrower difference with the US¹⁹.)
In fact, London ranks first globally in fintech deal volume overtaking top-ranked New York in 2019, with San Francisco coming third. In deal value, however, San Francisco maintains its lead, with London coming second, ahead of New York²⁰.
So what is the UK’s true economic potential if it commercialises its IP at peak effectiveness?
One could argue that comparing the UK to the US and China is like comparing gooseberries to oranges and persimmons. Let’s therefore compare the UK to an economy of its size, stage of economic development (i.e. not emerging) and, critically, with equally prolific high-calibre IP output. Let’s look west –to California.
The economies of California and the UK are roughly equal in size. California’s, at around $3T, just inched past the UK’s $2.8T last year. Both economies have similar compositions (with services dominating²¹), and both have staggering income disparities between regions, reflecting disproportionately concentrated economic engines (with personal incomes in the richest regions reaching four to five times those of the poorest)²². Most importantly, their leadership in intellectual output is strikingly similar²³ –each boasting four universities of the world’s top 20²⁴ (California with Caltech, Stanford, Berkeley and UCLA; the UK with Oxford, Cambridge, Imperial and UCL).
Where the two economies diverge significantly, however, is in their output per capita and economic growth. In 2018, California’s gross domestic product (GDP) per capita, at $74K, was over 75% higher than the UK’s, at $42K²⁵. The difference is even starker when comparing labour forces: the UK’s is 75% larger than California’s²⁶ (whereas the population is only 68% higher), suggesting that California’s labour force productivity outperformed the UK’s by 83%.
This gap has persisted for decades, resulting in considerable cumulative outperformance by California. The figure below²⁷ illustrates how much California has outpaced both the UK and US, in the ten-year period from the end of the great recession in 2009 to 2018 (by indexing 2009 at 100). In real terms (adjusted for inflation)²⁸, California’s cumulative growth for the period was nearly double that of the UK, at 34.3% versus 18.6% respectively.
Zooming out, the same picture reasserts itself. The figure below shows the cumulative real GDP growth since 2001 (after the dot-com mania was purged) to 2018 (with 2001 indexed at 100). California’s cumulative real growth was again double that of the UK, at 59.8% versus 33.1% respectively.
This divergence in economic fortunes has been driven by a sustained differential in annual growth rates. California has outpaced both the UK and US significantly. The figure below illustrates California’s superlative real GDP year-on-year growth, versus that of the UK and the US.
A closer look at the annual real growth rates of the UK and California, for the ten-year period from 2009 to 2018, puts the variance in stark relief. The annual real growth of California’s economy averaged twice that of the UK, over the period.
California’s sustained growth differential, that has surpassed both the UK and US for decades, shows the promise of relentlessly commercialising innovation at high effectiveness and scale. The UK can achieve these levels, if not exceed them. It would require concerted perseverance by the private, public and academic spheres to further develop and drive a confluence of critical enablers. The good news is that the UK is off the charts in the most elusive and critical of the enablers: prolific output of deep IP.
At play are a variety of other enablers –including expanding and deepening multi-disciplinary talent pools, increasing research funding, scaling venture funding, strengthening capital market liquidity, developing conducive regulatory environments, setting immigration policies that attract the best and the brightest, maintaining tax incentives for risk-taking, furnishing advanced affordable infrastructure, ensuring open and free access to large market blocs, such as the US, EU, Japan and China, among others²⁹.
A caveat worth noting, however, is that commercialising innovation at peak is not in itself a panacea , as California clearly demonstrates with its socio-economic and fiscal difficulties. It is imperative that the economic momentum is harnessed to build virtuous cycles that amplify economic and social progress, widen participation, and catalyse leadership in tackling the universal challenges of our age, such as the climate.
The UK is capable of commercialising innovation at peak effectiveness. The case for setting that as a national priority is overwhelmingly compelling. It promises a doubling of economic growth and commensurately dramatic boosts to per capita income. The UK, as a nation, has a responsibility to convert its intellectual output to maximal economic gains for its population and future generations. It did so once before with spectacular success. It is arguably, again, time.
¹ The field-weighted citation impact (FWCI) is a measure of how much impact a set of publications have had. It compares the total citations received by a researcher’s publications to the average number of citations received by all other similar publications from the same research field in the same year. The global mean of the FWCI is 1.0, so an FWCI of 1.50 means 50% more cited than the world average; whereas, an FWCI of .75 means 25% less cited than the world average. [Excerpted from Subject Guides: Measure Research Quality and Impact — Research Guide: Researcher Impact, Murdoch University]
² Scopus, an abstract and citation database provided by Elsevier; UK Department for Business, Energy & Industrial Strategy [BEIS], 2019.
The following figures on FWCI, share of own research among the world’s most highly-cited publications, and share of the world’s most highly-cited publications, were excerpted from the UK BEIS report, International Comparison of the UK Research Base, 2019.
³ Scopus; UK Department for Business, Energy & Industrial Strategy, 2019.
⁴ The h-index is an author metric that attempts to measure both the productivity and citation impact of the publications of an author. The h-index is a measure of an individual’s impact on the research community based upon the number of papers published and the number of citations these papers have received.
The method of calculating an h-index is the number of articles in a database that have received the same number or more citations over time. As an example, a researcher with an h-index of 20 has (of their total number of publications) 20 papers which have been cited at least 20 times each. [Excerpted from Subject Guides: Measure Research Quality and Impact — Research Guide: Researcher Impact, Murdoch University]
⁵ China AI Development Report, China Institute for Science and Technology Policy at Tsinghua University, 2018.
AI researchers are defined as researchers in the field of AI with issued patents and/or published English papers in the last ten years.
⁶ China AI Development Report, China Institute for Science and Technology Policy at Tsinghua University, 2018.
Top AI researchers are those with an h-index score that ranks among the top 10% of international AI researchers. Counting only the top AI researchers (top decile per h-index), the UK surpasses China, coming second after the US. China drops to 6th place with 977, whereas the US leads with 5,158 top researchers, or 4.4 times those in the UK.
⁷ Scopus; UK Department for Business, Energy & Industrial Strategy, 2019.
⁸ The World Bank, UNESCO Institute for Statistics.
Based on number of researchers per country, calculated as researchers per million, multiplied by population in millions. 2017 numbers.
China = 5.9x UK. US= 4.79x UK.
China: 1235 researchers/million * 1386million =1,711,710 researchers
UK: 4377 researchers/million * 66million = 288,882 researchers
US: 4256 researchers/million * 325million = 1,383,200 researchers
⁹ How Much Does Venture Capital Drive the U.S. Economy?, Insights by Stanford Business, October, 2015.
¹⁰ “The US VC industry came into its own only after a regulatory change in 1979 that allowed pension funds to invest in VC. That rule change, known as the Prudent Man Rule, led to a greater than tenfold increase in the money entrusted to VC funds: VC funds raised $4.5 billion annually from 1982 to 1987, up from just $0.1 billion 10 years earlier.” How Much Does Venture Capital Drive the U.S. Economy?, Insights by Stanford Business, October, 2015.
¹¹ Pitchbook, Crunchbase, Statista.
¹² PitchBook NVCA Venture Monitor Q4 2019. 2019 Will Be A Record-Breaking Year For US Tech IPO Exit Value, CB Insights.
¹³ As defined by the OECD: high-growth scaleup firms have an “average annualised growth greater than 20% per annum, over a three-year period, and with ten or more employees at the beginning of the observation period. Growth is thus measured by the number of employees and by turnover.”
¹⁵ Pitchbook, 2019.
¹⁶ UK tech sector beats both US and China to lead global growth in 2019, Tech Nation, dealroom.co.
¹⁷ Pitchbook; 2019 fintech investment landscape, Innovate Finance.
¹⁸ CB Insights, January 2020.
¹⁹ A Tech Nation, Dealroom.co June 2019 report listed 21 fintech unicorns in the UK (18 in London), and higher numbers for the US (31 in SF Bay Area; 12 in NY). The report suggests that the UK has more than a third as many fintech unicorns as the US does.
²⁰ Global Trends from a UK Perspective, London & Partners, 2019.
As of August 2019, fintech deal value: San Francisco: $3.02B; London: $2.11B; New York: $1.93B. Fintech deal volume: London: 114; New York: 101; San Francisco: 80.
²¹ The services sectors constitute over 70% of both the UK and California economies, while manufacturing and agriculture together make up less than 20% in either.
²² In California, per capita income in Marin (San Francisco Bay Area) were four times that in Kings (in the Central Valley) [Source: US Bureau of Economic Analysis]. In the UK, personal incomes in Kensington & Chelsea and Hammersmith & Fulham were around five times that in Nottingham [Source: UK Office for National Statistics].
²³ UK and California are relatively quite close in research impact. I was not able to get consistent quantitative scores for California. One source placed the UK slightly ahead of California, where another placed California slightly ahead of the UK.
²⁴ The Times Higher Education World University Rankings 2020.
²⁵ California: GDP: $2.94T; population: 39,557,045. UK: GDP: 2.81T; population: 66,573,504. [Sources: IMF, Bureau of Economic Analysis, World Population Review, US Census.]
²⁶ 2019 labour forces: UK: 34.2M; CA: 19.5M. [Sources: UK Office for National Statistics, State of California Employment Development Department.]
²⁷ The following three figures were generated on fred.stlouisfed.org, by the Federal Reserve Bank of St. Louis.
²⁸ Real GDP measures a country’s total economic output, adjusted for inflation. Unlike nominal GDP, real GDP accounts for price changes and provides a more accurate figure of economic growth –as it shows comparisons for both quantity and value of goods and services, across years. [Excerpted from Investopedia.]
²⁹ Going deeper into these enablers, and studying the data on their effectiveness and interplay, is a rich and evolving topic. Perhaps more on this in another post.