Estimate that spending target has been hit will shake up innovation policy, says Kieron Flanagan
On 29 September, the Office for National Statistics delivered the biggest surprise to hit UK research and innovation policy in years. Announcing a change to how it estimates business expenditure on R&D, the ONS revised its measure sharply upward—by about £16 billion a year.
On Twitter, the economist Josh Martin posted a quick estimate of what this means for the ratio of R&D to GDP, the widely used indicator of national research intensity (we don’t yet know the full effect because the extra R&D spending will itself have a small impact on the GDP estimation). It had seemed the UK was stuck on 1.6-1.7 per cent, but Martin estimated that the country had in fact hit the government’s goal of reaching 2.4 per cent by 2027 way back in 2018.
Scholars and analysts like me have long argued that the quantity of R&D—which is just an input to innovation processes—is not the whole story, and that we ought also to think about the quality and composition of spending. We have also fretted that the OECD definition underpinning national statistics excludes a lot of activity analogous to R&D in services and the creative industries.
But while we may have been lukewarm or critical about targeting R&D intensity as a policy, we’ve not really questioned the data. The ONS’s move is a salutary reminder not to rely on any one indicator, even for innovation inputs.
Unhelpful success?
On the political front, former science minister George Freeman has warned that Liz Truss’s administration might claim job done on the 2.4 per cent target and abandon or reassign the funding promised in the 2021 spending review, along with the broader commitment to a modest regional rebalancing of R&D spending.
Perhaps focusing on unfavourable international comparisons of R&D intensity wasn’t such a great lobbying tactic after all. That said, 2.4 per cent is just the OECD average: many nations the UK sees as its peers and competitors spend more like 3 per cent.
The revision also has implications for long-standing questions about the UK’s performance in research and innovation.
Since the late 1980s, for example, government investment has looked quite different to that seen in most R&D-intensive nations. Under Margaret Thatcher, the country essentially abandoned large-scale investment in civil technology programmes in areas like computing to concentrate on basic and ‘strategic’ (near-basic) science.
The assumption then and since was that ‘technology transfer’ would magically turn all this science into things that industry could take up and commercialise through further R&D. So why haven’t the enormous efforts at technology transfer and commercialisation since then produced enormous economic and social benefits? A better view of business R&D might provide some clues.
Hidden figures
We also know we need a better handle on other kinds of innovation inputs. These include investment in ‘intangibles’ such as brands or designs, and R&D-like activities in the creative sector and knowledge-intensive services. This could help inform policies to support innovation in these sectors, which account for most of the UK’s jobs and GDP, and include some huge exporters.
The revisions also pose new questions. First, if, as the revision suggests, ONS statistics have significantly underestimated R&D spending in small firms, what does the profile of activity in different-sized companies now look like, and how does that compare with other countries?
On the one hand, R&D in small firms might be more ambitious and radical compared with that in a large incumbent. On the other, it might be more speculative and less well managed. Larger firms may also be better at capturing the learning spillovers from R&D that seem to be so important.
Second, if recent decades have seen small firms do a larger proportion of private R&D, does this reflect a kind of outsourcing? Pharmaceutical companies, for instance, may rely on biotech startups to make risky early investments in new technologies, only to buy them up if the bet pays off. And firms that sell R&D services to other companies also seem to be very significant in the UK.
While we’d expect to see this in other countries like the US, it might be that this kind of outsourcing is more common in the UK, or that the way we do it here delivers fewer social and economic benefits—perhaps because the average UK startup is lower quality or less well-managed, or less able to capture the spillovers from R&D.
The new picture of the UK’s private R&D spending raises many questions. Now it’s time to start looking for answers.
Kieron Flanagan is professor of science and technology policy at the Manchester Institute of Innovation Research, University of Manchester
This article originally stated that Josh Martin worked at the Office for National Statistics. This is no longer the case and the affiliation has been removed
A version of this article also appeared in Research Fortnight