What the government's new industrial strategy can learn from the past - and four specific things it should do.
The government has pledged that it will develop an industrial strategy. Now it is trying to decide exactly what that means and how best to do it. Crafting an industrial strategy is tricky at the best of times, but in the UK the process is complicated by history. Whatever strategy is chosen needs not only to avoid the mistakes of the 1960s and 70s - which is widely associated with economic stagnation - but also to differentiate itself from the policies of 2010-2015 - which is seen to have done too little to combat the regional and sectoral imbalances that exist in Britain today.
This article is an attempt to set out principles for a successful industrial strategy, together with some specific policies to achieve them. It begins by asking what went wrong with previous UK industrial strategies, both explicit and implicit – with a particular focus on the past six years. It then sets out general principles for what an industrial strategy should seek to achieve.
Finally it sets out four specific things government needs to do differently, together with policy ideas to help achieve them.
He who has been bitten by a snake will be afraid of a piece of rope, says the Chinese proverb. When it comes to industrial strategy, a similar anxiety exists. British policymakers have traditionally been nervous about industrial strategy in part because of the perceived failure of the version the UK tried in the 1960s and 70s. The problems with this incarnation of industrial strategy are well known: it is associated in the public mind with British industrial decline, failed attempts to pick winners and shape new technologies, and a stifling corporatism. But it is not just the industrial strategy of the 1970s that policymakers feel they need to move on from. A new industrial strategy also has to distinguish itself from the industrial strategy of the years leading up to the EU referendum.
The problems with industrial strategy in the 60s and 70s are well known. The shortcomings of the industrial strategy of the last few years, however, have been less comprehensively analysed.
While it may not have made it into the name of a government department, the previous government pursued what it called an industrial strategy since 2010 (indeed, since 2009 – perhaps it is more appropriate to date the change from the financial crisis rather than from the formation of the coalition government; there was a brief break in using the language of industrial strategy between May and July 2016).
It included quite sophisticated industrial partnerships in selected areas (such as car-making, aerospace, and pharmaceuticals), a commitment to protect scientific research funding from the bulk of public sector cuts, increased investment in a few types of innovation policy (in particular, R&D tax credits and Catapult centres), and the promotion of new business formation and growth capital.
After six or seven years of these policies, the UK still faces problems of the sort the Prime Minister invoked in her earliest calls for a new industrial strategy: overdependence on particular sectors of the economy, low productivity, insufficient opportunity and good jobs. Logically, a new industrial strategy needs to be as clear on how it is different from the policies of 2009-2015 as it is on how it will avoid the failures of the 1960s and 1970s.
The lessons of industrial strategy in the 60s and 70s are arguably pretty uncontroversial: don’t fall prey to vested interests, make room for entrepreneurship and competition, don’t over-regulate or assume the government can pick the technologies of the future. The shortcomings of the industrial strategy of the last few years, however, have been less comprehensively analysed.
It is one thing to point out outcomes that it failed to deliver (higher business R&D, higher growth outside the South East, higher productivity growth, more jobs) – but this is not the same as identifying how it could have done better. After all, it is not as if policymakers before 2015 were unconcerned with growth outside London, with good jobs, or with productivity.
What follows is a brief attempt to diagnose the issues that recent attempts at industrial strategy failed to solve...
i) A misguided focus on a particular type of innovation. Innovation is essential for business competitiveness and productivity, and any effective industrial strategy should promote it. But 2009-15 industrial policy arguably focused too much on science, and a science-push model of innovation, to the exclusion of other worthwhile projects. Thus while academic science funding was protected from the full brunt of austerity, and R&D tax credits increased, spending on things like further education and business support (such as the Growth Accelerator and Manufacturing Advisory Service programmes) were sharply cut. This would be no problem if innovation proceeded from university research out into businesses – but for the most part it does not – management and workforce skills matter just as much, as does the influence of customers (including government procurement).
ii) Geographical agnosticism. The UK is biased towards making R&D and transport investments in the South East (and in Scotland - basically places with a government - see Tom Forth’s blogs for more on this). The government compounded this by scrapping RDAs and replacing them with far less-well-resourced LEPs, only some of which are effective organisations, even now; RDAs had many faults, but they did provide a source of local economic intelligence. Lack of local knowledge makes it harder to make local investments (how do you know which proposals are credible?) – with the result that economic opportunity accumulated in a few corners of the country. (The Northern Powerhouse began to address this issue in the North West – but of course the policy was only announced in 2014.)
iii) Flying blind. Industrial strategy involves making at least some specific interventions – which implies some sort of knowledge. Promoting growth in Manchester assumes knowledge about how Manchester and its opportunities compare to other cities; setting up a Catapult Centre for Satellite Applications implies you think investing in space makes more sense than investing in, say, meat processing. But government did relatively little to develop the capability for knowing what was going on in specific sectors, technologies or places.
This constrained government's ability to do a range of industrial policies, to know what was working, or to challenge local claims (for example from LEPs, from particular industries, or from researchers). It also limited the debate about businesses that may have been strategic national assets. When ARM or Vocalink were in the process of being acquired or when the Port Talbot steelworks was threatened with closure, government had little information to evaluate whether it should intervene in these businesses from an economic or a supply-chain point of view. Innovate UK has some capability to analyse new technologies, and Smart Specialisation Hub and the Science and Innovation Audits were intended in part to help do the same, but there is still some considerable way to go, in part because these capabilities are spread across many small teams in different organisations.
iv) Lack of coordination. Industrial strategy works if business policy is coordinated with other government policies: procurement, regulation, infrastructure spending, and education policy, for example. This kind of coordination is always difficult, but it was perhaps made harder during the coalition when BIS was run by a Secretary of State from the junior coalition party.
v) Not enough money. Austerity, and in particular the sharp cuts made by BIS as an unprotected department, made all the above problems worse. These savings meant that science spending could be spared deep cuts only by cutting skills spending (indeed, some other OECD countries increased science and skills spending in real terms after the financial crisis). They spelled the end of the RDAs. They meant that the successful policies used in the automotive and aerospace sectors could not be more widely rolled out. More recently, they necessitated the unhelpful shift of Innovate UK grants to loans, which seems unpopular with most businesses. Industrial policy need not be expensive, but it is difficult to make it a priority while making very large savings from the BIS/BEIS budget.
Some reasonable goals for a better industrial strategy would be:
These goals are, I hope, uncontroversial. The first is more or less a recap of statements already made by the prime minister about the purpose of the new industrial strategy. The other two seem straightforward: the importance of innovation and enterprise to productivity growth has long been known to economists, and evidence and capability are generally important to making good public policy. A less obvious question is what specifically a new industrial strategy should do differently to increase its chances of meeting these goals.
Industrial strategy must be grounded in innovation and in enterprise, the two complementary sources of competitiveness – but it needs a wider view of innovation than just the commercialisation of elite science, and a wider view of entrepreneurship than just tech start-ups.
A new, better industrial strategy requires two types of change: a modest reorientation of its aims, and a significant change in how it is carried out. We have four specific ideas for what to do differently, which we set out below. Each one starts with a general principle, and ends with a policy idea to help make it a reality.
i) Creating the opportunity to innovate
Innovation is the source of sustainable competitive advantage. But innovation does not just come from elite scientists in the Golden Triangle or dot-com start-ups in London. It happens throughout the economy, around the country and across the population.
A major study by John Van Reenen and Raj Shetty showed that children who were exposed to technology and innovation at a young age were far more likely to become innovators and inventors themselves – but that these opportunities tended to be available mainly to a privileged few. A Nesta study by Eric Von Hippel showed that as many as 6% of the UK population were “user innovators” – people who, through tinkering and ingenuity, make new products. New technologies, like the Raspberry Pi, micro:bit and Arduino, and new settings like makerspaces (places where people can tinker with tech – see the Nesta makerspace map here) give these new innovators more power than they have ever had before.
Unleashing the creativity of individuals could provide a significant boost to the UK’s innovative potential, especially if it is linked to UK businesses.
ii) Intelligence for policymaking
All strategies depend on some sort of intelligence. You would not make a business strategy without management accounts, market data, and forecasts, and yet in the past our industrial strategy has been based on only the most cursory data about the UK’s industrial strategy. Policymakers have had little reliable information about the performance of sectors, different technologies, or the economies of local areas. This makes it hard to decide where and how to invest, or for decision-makers to have confidence in investment decisions; it also makes it hard to keep an eye on how policies are working. In the absence of timely, relevant data, inertia wins the day: investment happens in the south east, we fail to focus our resources on promising industries or technologies, potentially damaging acquisitions are waved through with no challenge.
Nesta has for the last three years been building tools to analyse industrial and innovation ecosystems, understand the relative performance of different technologies, and map sectors and clusters; we are currently working with the Welsh government to build a tool to help shape industrial strategy in Wales. In this, we have had the pleasure of working with leading edge start-ups and researchers to build data sources on the economy that would have been impossible or prohibitively expensive five years ago.
We have also developed the international Innovation Growth Lab to trial which innovation and enterprise policies actually work around the world. To make a success of industrial strategy, BEIS needs to invest in tools like these to inform its investment decisions and see what is working.
Build an analytic tool to help inform industrial strategy along the lines of Arloesiadur, the Welsh government’s analytical dashboard, to map and measure clusters, the performance of industries (especially emerging ones), and networks between businesses, researchers and investors. Develop the habits of making investment decisions and policies in a data-rich environment, and regularly checking the data to see what is working.
iii) Build the capability to do industrial strategy
Effective industrial strategy depends on capable, experienced people both within and outside government. Germany’s industrial strategy depends on sophisticated and well-established industry bodies and local government to help shape strategy. France, which has also got some of these things right, has built a cadre of officials with experience of business, technology and investment.
There are a few examples of where this has worked in Britain. The Aerospace and Automotive Councils are good examples of well-functioning trade bodies that have helped shape government policy. And perhaps unsurprisingly, Britain has developed strong policy expertise in dealing with the financial services sector, including very high quality officials who have spent time both in government and in the sector itself.
Industry councils are not appropriate for all industries: they make more sense when a sector has well established supply chains and is not undergoing disruptive innovation. But there are many industries that fit this bill where there is more scope to build capability. This is doubly true at a local level, where the abolition of the RDAs and the slow emergence of some LEPs has left a vacuum in local economic policymaking.
Set out now to build a cohort of policymakers who can help shape industrial strategy, for example in the establishment of the new UKRI. Training programmes like the Global Innovation Policy Accelerator, run by Nesta with the Universities of Manchester, Oxford and Cambridge can help.
Invest in LEPs and local areas, for example by funding local business mentoring partnerships such as the Wenta mentoring partnership in Bedfordshire and Buckinghamshire or Nesta’s Creative Business Mentor Network.
Support the development of industry councils, along the lines of existing structures in the aerospace and automotive sectors, offering them a role in allocating funds and shaping policy for their sector. (This is especially approproate for sectors where incremental, rather than disruptive innovation is the norm.)
iv) Join up the dots
Industrial strategy must be strategic. One aspect of this is that it must be aligned with other government policy that affects business, especially regulation, procurement, infrastructure and education.
Economic growth in cities outside London depends on them exploiting economies of agglomeration, which in turn requires better transport and planning.
Done right, procurement is potentially the biggest source of government innovation funding businesses can hope to receive (and it lay behind many of America’s most successful tech businesses, not to mention that of erstwhile British stars like Autonomy and ARM). Regulation can make or break a new business, especially one with a potentially disruptive business model.
To some extent better coordination of industrial strategy across government will be facilitated by the new Industrial Strategy Board, that the prime minister will chair (a development we have long supported). But institutional nudges will also help. The suggestions below will help coordination procurement and regulation in particular with the industrial strategy.
If a department is serious about using the money it spends to promote innovation in the national interest, it should consider setting up an Systems Innovation Accelerator. An SIA is a departmental body whose role is to promote innovation to meet the department’s needs, and to grow the ability of businesses (and civil society) to help meet them, in particular by connecting its procurement budget with its long-term strategic needs. Energy, health and transport are promising areas for this kind of body.
In a few areas where new technologies are having a big effect on particular industries, set up a regulatory task force, to ensure that regulation is dealing effectively with changing business models. The UK has a record of doing this well when we try: it enabled the UK’s world-leading alternative finance sector, and has helped government stay on top of developments in the sharing economy. Here’s how it could work.
* A careful critique of the failings of British industrial strategy in the 1960s and 1970s can be found in Geoffrey Owen’s From Empire to Europe: the decline and revival of British industry since the Second World War. As Paul Nightingale recently pointed out in his evidence to the Business, Innovation and Skills Select Committee inquiry into Industry Strategy, some of the failures of this version of industrial strategy were due to external forces (such as currency and oil crises, or attempts to meet military as well as industrial requirements), and some aspects of the policy worked well (the nationalisation of Rolls-Royce in 1971, for example). But on the whole the critique that the strategy overreached, neglected entrepreneurship and failed to promote innovation still stands.