What’s the problem?
One of the reasons capitalism and markets work well is because of innovation: competition and enterprise come up with new goods and services that improve people’s lives and make them happier. Indeed, if you had to boil down the last 10,000 years of human history to a single significant story, the rise of liberal innovation and the massive improvement in human existence in the countries that gave it free rein would be it.
But while innovation has its roots in ingenuity and enterprise, good governance has an important role to play in making it flourish. First of all, governments set the rules for how markets operate — including everything from competition policy to health and safety rules to technological standards — and they can do this in ways that encourage innovation and novelty, or in ways that favour incumbents and quash innovation.
Secondly, governments have a role to play in investment in innovation. Developing an idea and turning it into a product requires lots of investment in intangibles like research and development and design. Economists have long known that these investments tend to have big spillovers: the business that makes them can’t be sure that it will reap the rewards rather than a competitor¹. (Think of the CT scanner, invented by the British business EMI, but sold worldwide by Siemens and General Electric, who got most of the commercial benefit.) Investors know this too — for all we think of Venture Capital as funding “tech”, for the most part, successful VC funds take little true tech risk.
This is why governments should fund innovation, whether in the form of basic research, applied research, targeted funding in fields like defence and energy, through tax credits for R&D, or through artificial property rights in innovation in the form of patents. The twentieth century provides many examples of significant publicly-funded research projects that led, in due course, to big economic benefits, from GPS and the Internet to hydraulic fracturing and the graphical user interface. (Although it’s wrong to suggest, as some do, that the government is the main or only risk-taker when it comes to innovation: all these inventions also involved large amounts of risky investment by the private sector. For every pound the Government invests in R&D, the private sector invests two, and that for every pound of R&D investment, there are ten pounds of investment in things like design and marketing.)
We believe that both when it comes to public funding of R&D and designing regulation that does not quash innovation, the UK government can do a better job.
UK public funding of R&D lags that of most other rich countries (0.5% of GDP compared to levels of 0.7 to 0.8% of GDP in countries like the United States, Germany, France, or the Netherlands). The ratio of public to private R&D in countries like the USA and Germany is about the same as that in the UK; our total levels of R&D investment are lower because both our public and private sectors invest proportionately less.
What is more, the way we fund innovation has some unusual quirks. Compared to other rich countries, more of our public research funding goes through universities than through public labs or government research agencies like Germany’s Max Planck and Fraunhofer Institutes. IN addition, much of our recent increases in public R&D funding have gone through the Industrial Strategy Challenge Fund, which has developed as a funder of big joint industry-business projects in four so-called Grand Challenge areas (chosen by Government) and signed off by ministers. There are two drawbacks to this funding model:
- The dominance of university research risks missing a trick, to the extent that some cutting-edge public funding bodies, like DARPA, seem to have a good track record of original, impactful research linked to practical application. Many academics do brilliant work; but the incentives to generate publications and citations rather than either longer-term or more practical forms of impact are strong.
- The Industrial Strategy Challenge Fund seems to favour quite broad and generic challenges (the official four grand challenges include “the Ageing Society” and “the Future of Mobility”). What’s more, ISCF funding is decided through a politicized decision-making process (in which BEIS ministers make the final investment decision), which seems likely to favour easily explicable projects rather than radical breakthroughs. To put it another way, we find it hard to imagine a breakthrough project like ARPA’s Information Processing Techniques Office being funded through the high-level Whitehall and ministerial meetings that run the ISCF.
When it comes to regulation, we face two challenges. While British regulators have a relatively good reputation globally, most of them are not tech experts: it is simply not how they are set up. But increasingly, they are facing questions with big technological dimensions: is it a problem that Google enjoys a large share of the online advertising market? If a water utility uses AI leak detection to reduce its costs, how should those savings be passed on to consumers? What remedy is there for rights-holders if search engines routinely direct users to pirated version of their content? It is not clear that our regulators are staffed up to meet these new issues.
The second issue is a global one. One of the arguments made for Brexit was that it would allow the UK to set its own rules in new technologies that would give British businesses the edge over rivals governed by putatively more onerous EU laws. This will be much more likely to be true if the UK can not just develop better rules, but can also convince other countries to accept them. All other things being equal, the EU has more international clout to get its rules accepted than the UK does alone. Overcoming this challenge will be important if the UK is to make the most of whatever “regulatory divergence” opportunities arise in tech.
Finally, the UK has an unusual geographical problem when it comes to the commercialization of our leading-edge academic research. Because academic research funding is awarded on the basis of excellence, much of it has traditionally gone to places like Oxford, Cambridge and the centre of London. It just so happens that these places have very onerous planning restrictions, making it very hard to build new homes, offices or factories. This is an economic problem: if it’s hard to build new offices, it is more expensive and difficult to set up spin-out businesses. If it is hard to build houses, it will be hard to attract skilled workers to work in spin-out businesses. In this way, planning constraints in effect turn into powerful constraints on innovation and high-tech business growth. Or to put it another way: the most effective tech transfer policy is planning liberalization in R&D-intensive places.
Our goal
We need a government that invests in research and technology in a comparable way to other advanced economies. It must do so in a way that encourages radical, high-impact innovation, which will require a diversity of funding streams and room for experiment. It must be backed up with a regulatory system that creates space for new, tech-enabled insurgents to enter established markets, and a capability to persuade other countries to adopt our rules and standards. Finally, planning policy needs to reinforce tech policy, making it easier to live, work and build businesses in places with a history of excellent research.
Specific policies
Four specific policies can help deliver this vision.
A significant increase in public R&D funding, increasing public R&D funding from 0.5% of GDP to 1.0%. This would take the place of the current target of increasing total R&D to 2.4% of GDP, which although worthy in its intention, is not clearly in the Government’s control, since most R&D is not government-funded.
New funding mechanisms. Rather than channeling new public R&D funding through vague, broad Grand Challenges, with consortia of large businesses and ministers closely involved in bid selection, we should focus a proportion of new funding to one or more radical breakthrough funding agencies, closer to DARPA or ARPA-E than to the ISCF.
To the extent we fund challenges, we should make them narrowly focused and technically demanding (the current STEP project to develop a spherical tokamak for nuclear fusion seems much more appropriate than broad challenges on themes like “healthy ageing”).
Competition in regulated sectors. As well as the broad trends we’ve described, and the policies we suggest to respond to them, there are some niggling problems with the UK economy that an economic-minded Conservative Party should consider.
The first is the growth of the utilities regulators. These were initially set up to manage monopoly infrastructure: in energy, for example, we have an at least notionally competitive retail sector and production sector, but distribution, while privately owned, is mostly done by national and regional monopolies (the National Grid and the Distributed Network Operators). The same goes for other privately-run infrastructure that is either a natural monopoly (like the water network) or effectively a monopoly (like the railways or airports).
The regulators were set up to monitor these monopolies and ensure that they did not capture all the value that could otherwise be passed on to consumers, and to regulate the newly deregulated consumer markets for a time limited period (while the market reached maturity). However, mission creep has led them to accumulate lots of other responsibilities that have turned many of them into bureaucratic juggernauts. Ofgem and Ofcom both employ more people than many government departments, with annual budgets of around £90 million and £120 million respectively.
These sector regulators have ‘concurrency’ powers that mean that they, not (in practice) the Competition and Markets Authority, have responsibility for competition enforcement in their sectors. But very few competition cases have ever been brought, compared to France or Germany, for example.
Unlike a competition enforcer that has to stick to an established, cross-sectoral rulebook, regulators make ad hoc interventions that amount to micro-management of many aspects of the market. As well as this, repeated interactions between regulators and industry mean that regulators have incentives not to punish existing players too harshly — regulators rely on established players for information and cooperation, and staff move back and forth between industry and the regulator.
Regulators tend to make static analyses and as a result they do not put much weight on innovation or barriers to entry as important factors in the market improving. The result of this is that incumbents in these sectors are focused more on the regulator than they are on consumers and competing with their rivals.
A solution to this may be to redesign the regulatory system so that it is not sector-specific, with competition powers and a requirement for market investigations given to the CMA, a new independent agency for consumer protection, the n+1 regulator discussed below, and one for access and network regulation.
This single access regulator could be given the narrow mandate to protect consumers by guaranteeing open access to monopoly infrastructure, in order to facilitate new entry and promote the interests of consumers, instead of trying to micromanage or centrally design the market structure.
A new cross-sector sandbox authority. John Fingleton, former CEO of the Office for Fair Trading, has proposal for an “n+1 regulator” that would sit across all sectors of the economy, and aim to support new companies with innovative business ideas that existing laws or regulations meant could not operate.
It would grant five-year licences to new companies whose business models conflicted with existing regulations — the next Uber, for example — and allow the companies to operate for this period whether they were in breach of existing law or not. The companies would be required to take out liability insurance, or in some cases the regulator itself may offer it (at a price) if the private sector would not insure an innovative business.
This approach already exists in healthcare, where treatments that haven’t had regulatory approval are allowed to be used under certain circumstances, and to some extent in fintech. It would have two responsibilities: one, it would allow innovative companies into the market that would be blocked by regulation otherwise, and two, it would be responsible for negotiating with the sector regulators to change the rules so that they could continue to operate in the long-run. Setting this up would put the UK at the forefront of global regulatory practice and be a statement of intent about our regime after Brexit.
We also need to project this influence globally. To that end, the government should charge UK Research and Innovation and our national laboratories and standards bodies to establish a centre for international technological standards and ethics, dedicated to undertaking groundbreaking international work on standards and norms for new technologies, and putting in the work to get these standards adopted as widely as possible. (Standards and norms aren’t sexy, but the UK is good at them, and compared to many “big science” projects, they are dirt cheap: a relatively small investment in the field will go a long way to helping the UK make the most of our global technological ambitions.)
Tech cluster planning reform. A lot of ink is spilt on optimising ”tech transfer” policy, often involving tweaks to licensing regimes or nugatory funding pots announced to much fanfare. But: the simplest and best policy for tech transfer is to make it easier to build homes and offices near places where excellent research happens. The Government should throw its weight behind a major expansion of either Oxford or Cambridge, our two most research-intensive towns, where despite decades of wrangling, it is still hard to build significant amounts of new housing (Cambridge is currently doing a better job than Oxford). The Government’s vision should be a significant, beautiful expansion of one of the two towns, inspired by historical projects like Edinburgh New Town or Bath. They should be delivered to high architectural standards and with adequate infrastructure, funded by land-value capture. Having an open tender for Nansledan/Poundbury style urban extensions, masterplanned as a single project like Fitzrovia, Bloomsbury or Marylebone, may not be the first-best way of building out but in a world of tight planning rules it may be the second- or third-best.
Finally, if planning reform near our oldest universities proves to be absolutely impossible, we need a serious rethink about the geography of innovation. If local politics makes it prohibitively expensive to build significant numbers of offices, houses, or factories near where the bulk of publicly funded R&D happens, Government should consider shifting the location of some of that R&D to places where building is possible, especially when there is already a lot of private sector R&D going on there. This would involve working closely with elected leaders in cities like Birmingham, Manchester or Bristol to greatly expand public funding to their excellent universities, together with plans for development and links with local businesses. If the mountain will not come to Mohammed, Mohammed must go to the mountain.
A couple of more specific issues:
Procurement for innovation. You can’t spend long reading about innovation policy before you come across someone saying “if only the government spent a bit of its procurement budget on innovation, it would change everything”. This is technically true, but it’s much easier said than done. The general problem is that procurement officials generally work to tight budgets, are aiming to deliver value for money, and have few connections with the world of innovative entrepreneurs. Procurement is difficult enough on its own without giving the people responsible additional goals that are often in direct opposition to their core objective.
And the risk-averse culture of politics is a big disincentive for civil servants or ministers to take a chance on a small supplier. The only place where procurement has historically led to innovation is the US defence sector, which is characterised by vast budgets, fifty years of total commitment to technological superiority, and a degree of political untouchability.
The way to address this is for the Government to pick one or two specific areas of public service spending where there is real political desire to innovate. (“Real political desire” means it must be one of the Secretary of State’s top three priorities — accept no imitations.) Then, the relevant department should set up a focused team dedicated to nurturing external innovations, funding them with R&D money (which UKRI should support), and connecting them to the department’s procurement pathways, using challenges and procurement tools like SBRI. This was nearly done for renewable energy innovation in 2016, but the change of government saw the project cancelled.
Doing research better. The past thirty years of British research policy have seen research funding become significantly more managerialist and more metrics-driven, a trend exemplified by the Research Excellence Framework that academic researchers are periodically subjected to. Like a lot of managerialist public service reform, this trend has had some major benefits: it has professionalised research funding, and focused attention on quality and, to some extent, on impact. But as many researchers have pointed out, managerialism can go too far. There is the risk that impact assessment is encountering diminishing returns (the time and effort required to comply with the system may no longer be justified by the benefits the system brings). Others have raised concerns that it discourages really radical research, or that it discourages the production of useful outputs like digital tools.
Government should commit to reduce the bureaucracy of research funding, and to pilot new methods like funding lotteries for scientific grants. A good way to start this off would be by commissioning a review into research funding bureaucracy, led by a senior researcher. This is a complicated area, where there is a risk of throwing the baby out with the bathwater, but an important one if we care about research.
At the same time, we should be ensuring the UK leads the field in what we might call meta-research: understanding what research methods work most effectively (especially in the light of modern technologies), and ensuring our research funding system encourages them.
[1] We note in passing the remarkable fact that many people involved in government innovation policy don’t seem to have a clear model of why government funds R&D.
Next: the Long Run