Regional research and development subsidies are working.
But not in the way most people think.
, .
In its 2017 Industrial Strategy the UK government set itself a goal of increasing research and development (R&D) intensity in the British economy from 1.7% of GDP (only slightly higher than Spain and Italy) to 2.4% of GDP (similar to the USA back then) by 2025. The stretch goal was to reach 3% of GDP (like Austria and Germany).
In the same year the UK’s Digital Economy Act was updated to give the Office for National Statistics (ONS) more powers to demand and use administrative data from other parts of government to improve its statistics.
In early 2024, releasing data for 2022, and having forewarned that big changes were coming, the ONS announced that the UK government had already smashed its target and is close to the stretch goal. Britain now spends 2.8% of GDP on R&D.
Like other big achievements of the last UK government such as English kids becoming comparatively good at reading, few people know, even fewer people care, and almost no-one is celebrating. Now’s your chance. But do it quickly because some of the rest of this blog post is less positive.
Some of this increase in R&D intensity in our economy is due to more being spent by government, charities, and companies. Mostly it is that the ONS learned from abroad, focused on the issue, and used their new powers to massively improve their methods for estimating how much is spent on R&D by companies. They used data from HMRC on which companies assert that they perform R&D by claiming R&D tax credits, made better use of information from the interdepartmental business register (IDBR), and designed better surveys to reach about ten times as many companies as before. Their new methods much more accurately capture the R&D of smaller businesses, especially in sectors that have traditionally performed less R&D. It’s great work.
It turns out that British companies have been spending about 60% more on R&D than we thought for a long time. We may well have already been meeting the 2.4% target at the time it was ambitiously set.
These updates pose a lot of challenges to a lot of economists. Low R&D spending by UK businesses has long been proposed as an explanation of the UK’s persistently low productivity and recent productivity stagnation. That story is now much less convincing.
Data is always best treated sceptically. In this case we should consider that the UK subsidises private sector R&D more than any other country in the OECD, largely through the generous tax credits I have already mentioned.
Big subsidies can cause big distortions and this government generosity incentivises companies to claim that they are performing as much R&D as possible. It is likely that companies come to believe these exaggerations and repeat them when they complete their ONS R&D survey. These distortions are likely to be larger in the small enterprises that the ONS have found doing a lot more R&D than they thought. A big company like Rolls-Royce is less likely to claim or believe that running a training course for staff to use the new office printer is genuinely R&D.
That British businesses seem to perform so much R&D while investing comparatively little in capital backs up these concerns. But few researchers in this area think this effect is enough to make British businesses look as stingy on R&D as they seemed to be until recently.
As an economist at risk of being forced to change my mind based on this new data I have been stewing my hat for a couple of years now, just in case I needed to eat it at short notice when regional breakdowns of the data were released.
The story has changed.
My concern stems from a paper that Richard Jones and I published in 2020 titled The Missing £4bn focusing on the regional distribution of R&D spending within the UK. In it we argued that the UK public sector misallocates its R&D spending by funding lots of research in places where the private sector funds little (like London and Scotland) and funding very little research in places where the private sector does lots (like Cheshire and the West Midlands).
Given the importance of proximity to knowledge transfer and talent acquisition, and other factors such as the sectoral mismatch that locational mismatches are associated with (Loughborough University does research more relevant to industry in the East Midlands than Kings College London does), this regional misallocation would reduce return on investment from UK public sector research funding.
But even though I’ve always felt that London’s businesses weren’t particularly focused on innovation, I wondered if the data on business spending on R&D was missing quite a lot of what happens there.
The new data shows that to be the case. London’s R&D uplift with the new methodology seems to be (the UK leaving Eurostat makes this data much harder to access so I’m still struggling with it a bit) the largest of all the UK’s regions and nations. North West England’s uplift is second highest. South East England’s uplift is among the lowest.
The story changes slightly because of the new data. I suspect it also changes slightly because London’s business R&D has genuinely increased since 2016 as high investment in public sector R&D (eg. the Crick Institute) starts to generate private sector spillovers (eg. Automata) and as companies in areas of comparative strength (eg. DeepMind and Wayve) massively increase R&D spending, especially where they can secure foreign investment.
Even those things together still leave London in a privileged position in UK research and development. It enjoys more public sector funding than its business spending merits while regions like the West Midlands and North West England continue to receive less. But it seems likely that London and the Greater South East of England have also genuinely turned a corner and have, like fellow preferenced national capital Berlin which we covered in our original paper, started to convert sustained high public investment into private sector spillovers.
It seems that a pattern of regional misallocation of public R&D funding that has held for decades (at least) in the UK has now started to erode. The good news is that we are now probably allocating our public sector R&D less badly and we have even more evidence to suggest that high public sector spending on R&D can crowd in higher private sector investment.
The bad news is that we may have missed a big chance to back existing strength outside of South East England and instead let that strength decay.
UK industrial strategy is a great success.
Decades of focused industrial strategy by the UK government has moved the UK’s centre of innovation South and East. We know that the government decision in 2000 to move the UK’s synchrotron radiation source from Cheshire in North West England to Oxfordshire in South East England was primarily a political one and led to the substantial shift of the UK’s related scientific excellence to the South East. We know that AstraZeneca’s move out of Cheshire to Cambridge in 2013 was motivated in large part by the high density of relevantly educated and talented people ready to work on their problems there, a by-product of centuries of government focus on Cambridge University. We know that proximity to UK central government and the huge public investment in transport, institutions, and R&D in the Kings Cross redevelopment that hosts AstraZeneca’s commercial head office helps to ensure that policy wins such as being selected to commercialise the UK’s national champion Covid vaccine are more likely. We see in the latest data that Cambridge’s region of East Anglia has now overtaken Cheshire — a region receiving almost no public sector investment in R&D since the relocation of the UK’s synchrotron — as the small region of Britain with the highest private sector R&D intensity. There are hundreds more examples and they all add up.
What’s done is done. As the UK’s national synchrotron begins its £500m upgrade (expenditure that makes very welcome partially-devolved R&D funding settlements to Birmingham, Manchester, and Glasgow look small) nearly two decades since it moved to the South East it would be a bad idea to move it back to the North West. This is not the kind of equipment it makes sense to have two of. Like so much more in this debate, and others on regional growth in Britain, it really is either/or more than why not have both?
AstraZeneca is, I’m told, working more closely with people in the North West as the high price of office space, labs, and housing in Cambridge and London starts to bite. The excellent Cambridge Manchester innovation partnership aims to boost this and broaden it to many more companies. AstraZeneca is probably mindful that much of their supply chain and talent base did not move with them to South East England and that being close to both manufacturing and supply chain can improve both R&D and commercial functions. But they will not be leaving Kings Cross and Cambridge and moving back to Macclesfield.
In the time since they moved, the UK government has built and sold a Vaccine Manufacturing and Innovation Centre in Oxford, opened a new railway line linking Oxford and London, upgraded existing railway lines connecting both Oxford and Cambridge to London, and is now building a new railway line between Oxford and Cambridge. This is on top of another decade where, despite promises to do the opposite, the UK government’s civil service and power has centralised in London, public investment in infrastructure has heavily tilted towards London with most investments into and around London such as Crossrail, Thameslink, HS1, and HS2 in the South built or under construction while similar investments in the North have largely been cancelled. For another decade public R&D has remained steadfastly focused in London and South East England at a rate well above what the market suggests is merited. A new research body ARIA has recently opened there and is likely to further deepen preference in funding towards that city and its region in the coming years.
We still aren’t trying to level up.
There is nothing we have done in the last fifteen years in the three areas of government most likely to lead to levelling up that we could have expected to work. We are still not trying. The outcomes are as expected. North England’s economy has fallen behind East Germany’s and while fiscal transfers within Germany have decreased, the UK is ever more reliant on money from London to fund public services its weak and stagnating economy outside of London cannot afford themselves. We will not return to being the world's most prosperous large country while our metaphorical national econonomic jet aircraft flies on only one engine.
Northern transport is barely improved in decades and the gap with the South East, and increasingly with Scotland, is deeper than ever. The 20 million people of North England and the Midlands who live at a similar density to the 20 million people of London and the South East enjoy vastly worse connectivity and have little chance of forming an effective and prosperous agglomeration. Sheffield’s airport has closed, promises to electrify its railway have again and again been cancelled, plans for a motorway to Manchester were abandoned. Every North English city has a similar story to tell while The Elizabeth Line has rapidly become just about the world’s most successful railway. The UK government has centralised and what devolution has been achieved has been smaller than the continued erosion of local government budgets and the still astonishingly limited local power our cities enjoy compared to those in unsurprisingly more productive countries like France, the USA, and Germany.
What matters now for Britain outside of South East England is what’s next.
In data, tech, and now AI the UK government has consistently tilted the playing field in favour of London. The Open Data Institute, Tech City, Tech Nation, The Digital Catapult, The Centre for Data Ethics and Innovation, The British Library, The Alan Turing Institute, The AI Safety Institute, ARIA, and much more have been placed in the capital. The indirect government subsidies on offer in London make it ever more rational for British talent and ambitious companies elsewhere in the UK to relocate there.
And now, despite levelling up having been spoken about for another decade and still not tried, there is a growing pushback against what are considered regional subsidies to economic development. Details of that, and a broader push to reindustrialise Britain, are better left to another blog post than tacked on here.
An end to regional subsidies?
I would welcome an end to regional subsidies within economic development in Britain. It should be unacceptable for explicitly distortionary regionalist arguments such as Sir Paul Nurse’s on deciding that London would host the Crick Institute.
But this never seems to be what advocates of reducing regional strategies mean. That means that the rest of the UK cannot unilaterally disarm on this.
I still hear no-one sensible who proposes ending regional economic development subsidies pushing against the decision to place ARIA, the AISI, and a huge swathe of other new UK government organisations in London. It turns out that if you live in London, and your friends live in London, and you fund and get paid by organisations in London that just feels like good national investment and getting around the constraints of government. How convenient.