Notes

Establishments

2020-12-24

About

The origin story of a new Government research funding agency, broadly modelled on the US Advanced Research Projects Agency — ARPA — and announced by the UK government in 2019. Related background reading on the UK Government and its research and innovation system, complete with a brief Who's Who.

Drama, intrigue, and procedural machinations in government departments, and among MPs in the Commons. Selections from a series of science policy hearings.

The quest for a new funding agency and the challenge of making it fit in the existing landscape.

Introduction

I've got a new favourite show: the oral evidence sessions of the science committees of both houses of Parliament.

The science committees, filled with MPs in the Commons and Lords in the upper house, regularly invite officials from government departments and leadership from public bodies to present evidence on various issues. They invite professors in and hear from industry experts, from leading figures in all kinds of organisations near and far.

These inquiries amount to a body of knowledge that is assembled for the benefit of the public. The committees serve as guides to Government thinking and upcoming legislation and new initiatives. They keep the Government in check. The committee members ask often insightful questions for the benefit of their constituents and the general public.

Few things are as entertaining as hearing brilliant people discuss some timely topic of interest. And these truly are leading figures in public policy that the committees receive evidence from. Each expert alone could provide wise commentary on any subject in their field, but typically there's a panel full of them, each speaker building on the previous one. The consistency and quality of the conversation is remarkable, almost unbelievable. On the other hand those without good answers duck and weave around the questions in a highly entertaining fashion.

Sometimes one can catch fascinating reactions and sentiments, because these are real people, this is real life. The witnesses are truly concerned, truly under pressure, truly squirming under incisive questions, truly enthusiastic. And there's humour. Everyone understands the context, they understand the process and protocol and how to work within various systems. There's references and nods and contrasts and conflicts if you know where to look.

This write-up is also a chronological report of sorts. Reflections on a subject I find particularly interesting: the creation of the new UK research funding agency. This new agency — modelled after the legendary ARPA, the US research organisation with roots in the Cold War era — was granted some serious funding in 2020, but at the end of the year, it's form and integration into the existing research ecosystem remains shrouded in mystery. And, of course, there are some fantastic characters mixed up in all this.

At the centre of the story is one man, Dominic Cummings, a contrarian dreamer with great ambitions, little patience, and the opportunity of a lifetime. Cummings had a vision for a particular flavour of greatness, but he wasn't quite able to clearly articulate or deliver it. And yet, he set in motion a process and potentially a new institution that might well amount to something remarkable one day.

Background

Her Majesty's Government

The United Kingdom of Great Britain and Northern Ireland is a unitary parliamentary constitutional monarchy. The UK is unitary in the sense that the sovereign state is not a federation like Germany and the United States. Notably, Scotland, Wales, and Northern Ireland each have autonomous devolved powers granted by the central government — autonomy within a unity.

The UK parliament has two houses: the elected House of Commons and the appointed House of Lords. Both houses can introduce legislation, and both have to approve it. There's a thousand years of tradition as to how this is done.

Due to the long and varied history of the nation, the UK has an uncodified constitution. No single document outlines the fundamental laws of how the state works. Instead, it's all just tradition: statutes, conventions, treaties, and judicial practice. Her Majesty The Queen is the head of state, and as monarch appoints the prime minister to head the government. By convention, the monarch respects the prime minister's decisions on government.

The prime minister is typically the leader of the political party or coalition that holds the largest number of seats in the House of Commons and who in this capacity is most likely to command the confidence of the chamber. The Commons is made up of 650 Members of Parliament (MP) who represent their single member constituencies. The MPs get elected in general elections using the first-past-the-post system, meaning that the electorate primarily votes for the single candidate from their preferred party. While MPs represent their constituency on the national stage, in effect the general election is an election to select the next prime minister.

The prime minister selects all other ministers of state by invitation from the monarch. This Cabinet comprises elected members from the Commons and selected members from the House of Lords. Cabinet ministers usually head their respective government departments and serve as secretaries of state. Some cabinet ministers have no department, some departments have no minister, some ministers are not secretaries of state, and so on. Select departments maintain the traditional ministry designation: Ministry of Defence, Ministry of Justice, and Ministry of Sound.

The purpose of government departments is to prepare and implement policies regardless of the government's composition. They therefore adhere to political neutrality. A legion of employed officials, the civil servants, do the actual heavy lifting in public policy, forming the permanent bureaucracy of the Crown. On the civil side the head executive in charge of managing the department is typically the permanent secretary. Below the permanent secretary there are director generals in charge of policy areas, and then directors and managers and officers all the way down.

Government departments work together with various non-ministerial departments, executive agencies, tribunals, corporations, and other public bodies and units. There are currently 23 ministerial departments, 20 non-ministerial departments, ~400 agencies and public bodies, ~100 "high profile groups", and 13 public corporations. The Cabinet Office maintains a list of government organisations.

In addition to a cabinet minister, government departments may have several ministers of state who assist the secretary as department executives. The lowest of the three ministerial levels is the parliamentary under-secretary of state, responsible for some particular policy area — these are commonly junior MPs from the ruling party or coalition. Finally, rank and file MPs may serve as parliamentary private secretaries (PPS), effectively unpaid assistants to a minister. Government ministers may also appoint special advisers (SPAD) to assist with government policy work. These un-elected characters often have an influential role feeding ministerial vision to the civil service.

In addition to representing their constituency and possible ministerial duties, members of parliament can participate in committees that look at particular issues in detail, including government policy, expenditure, and new laws. Select committees in both houses check and report on a range of areas. Commons committees primarily shadow the work of government departments, while the Lords investigate specialist subjects. There are also joint and grand committees for special topics.

Finally, to improve processing speed, a general committee of fewer than 50 or so members is appointed to see through each proposed Bill of legislation. Each Bill is read several times in both houses, chewed up in committees in between, then amended and reconsidered. Ultimately all successful Bills get sent to Her Majesty for royal assent to become law as Acts of Parliament.

All of the above is a rough approximation at best. Every detail has an exception and special cases.

The UK Research and Innovation System

The state distributes funding through its departments. The Ministry of Defence has its own laboratories and initiatives, the Department for Environment, Food and Rural Affairs runs its own agencies, the Department of Health and Social Care has its own research bodies, and so on. The largest and the most extensive of government research machines can be found under the auspices of the Department for Business, Energy, and Industrial Strategy (BEIS).

In addition to the work done in the departments, the prime minister has his own science policy units closer to No 10. The Government Office for Science (GoS), led by the Chief Scientific Adviser, reports to the prime minister and the Cabinet on forward-looking science matters, often referred to as "horizon scanning". The GoS also hosts the Prime Minister’s Council for Science and Technology (CST), effectively a Government-internal policy think-tank.

Some recent publications:

UKRI

BEIS funds some organisations directly, including the Met Office and UK Atomic Energy Authority, but directs most of its research funding through a non-departmental public body, a kind of a research umbrella organisation. This funding agency, UK Research and Innovation (UKRI), launched in 2018 with a mandate to distribute billions in research and innovation funding every year. UKRI charter is found in the Higher Education and Research Act 2017.

UKRI is made up of nine funding councils: seven disciplinary research councils plus Innovate UK and Research England. The research councils, reformed to fit inside UKRI in 2018, fund research projects in their respective focus areas, and also invest in research infrastructure and related human capital. In its umbrella role, UKRI tries to help the whole research sector work better together, including across the research council boundaries.

The two other UKRI councils work more on the innovation front. Innovate UK "drives productivity and economic growth" by supporting businesses in developing new ideas, including spin-offs from the existing research base. Research England supports England-specific research and knowledge exchange between universities and the wider UK research ecosystem. The devolved nations have their own funding bodies (and agencies) for knowledge exchange.

UKRI research funding is available to individuals and teams in research organisations. In both cases certain criteria have to be met. Fellowships and other individual funding opportunities typically require a PhD and association with an eligible UK research institution. Some individual grants target specific career stages. Funding for a research team effort requires a principal investigator at a qualifying research organisation. The calls are competitive, but open to all qualifying research organisations. Submissions are carefully examined by the host councils.

In addition to the "academic route" through higher education and public research units, UKRI funding is available to UK registered businesses. Business funding proposals typically do not require a PhD from the key people, but equivalent research/innovation experience and the ability has to be demonstrated somehow. Other eligible host organisations include Catapult Centres, which are purpose-built sectoral government tech innovation hubs; research institutes; NHS bodies and public sector research establishments; and independent research organisations with existing capabilities and capacity for research.

On the policy side, UKRI's independence as an organisation "at arm's length from the government" is often highlighted. The reality may be different. In any case UKRI has a close relationships with both the parent organisation BEIS, and the Government Office for Science.

The UKRI councils follow a standard proposals and grants process, including peer review. This is in agreement with what politicians like to call the Haldane principle, roughly the idea that decisions on individual research proposals are best taken by researchers themselves. Another "term of art" for science policy is found in "the linear model", a research policy straw man one can try to torch for reform support. Historian David Edgerton has comments on both.

Notable programmes that UKRI runs directly include:

Research in Numbers

The UK Government has made a commitment to increase overall research and development spend across the whole economy to 2.4% of GDP by 2027 (and 3% in the longer term). 2019 UK GDP was £2.17 trillion, so total research and development spend would amount to some £60 billion. Total UK research spend was some £37 billion (1.7% GDP) in 2018, with public funding accounting for £9.6 billion (26%) and the business sector for £20.3 billion (55%). The rest comes mostly from overseas (£5.1B) and private non-profits (£1.9B).

In other words, the UK plan is to increase total research spending by doubling the public research funding effort (to ~£20B), with an aim to win an equal amount of further research investment from the private sector (another ~£10B). This would bring the total spend to around £55―£60 billion, which would more or less meet the target 2.4%. This figure, 2.4% of GDP, is the OECD average. Germany is at 3.1% of GDP, the US at 2.8%, France at 2.2%.

For comparison, Alphabet (parent of Google) spent £16 billion on R&D in 2019, Samsung spent £13.0 billion, Microsoft £12.9 billion. Top R&D companies in the UK in 2018 were AstraZeneca (£4.6B), GlaxoSmithKline (£4.1B), HSBC (£1.6B), Rolls-Royce (£1.3B), and Lloyds Banking (£1.2B), followed by APTIV, Unilever, Royal Dutch Shell, BT, and Micro Focus International.

For more, see, for example, this House of Commons briefing (17 June 2020).

ARPA

Following the devastation of World War II, there were two major players standing on the international stage: the United States and the Soviet Union. Both superpowers entrenched their position by building alliances with like-minded nations that needed a strong partner. The US joined forces with Western Europe and formed NATO (1949), while the socialist republics of Eastern Europe signed the Warsaw Pact (1955) to band together with the Soviet Union. These two teams would engage and rival one another in games of cultural hegemony, proxy wars, and propaganda for the next thirty years of the Cold War.

One of the theatres of Cold War was science and engineering. In the West, people were trying to figure out what do with all this nuclear power that now seemed within reach. In the East, rocketry developed further, with ever greater range and payload capacity. The 1950s marked both the beginning of the nuclear era and the beginning of the space age. By the end of the decade, the arms race was in full swing. Both sides were engaged in a desperate game of simultaneous pulling ahead and catching up.

In the aftermath of WWII, the US had set up the Department of Defence, the Defence Science Board, the National Science Foundation, and various programmes for scientific inquiry, but it was still institutionally unprepared for the fireworks that the Soviet Union put together between 1957 and 1958. Not only did the Soviet Union reach parity with the US on nuclear armament, but the Soviets also demonstrated a prototype intercontinental ballistic missile (ICBM) capable of delivering these bombs. To top it all off, the Soviet Union was able to launch the first satellite, Sputnik 1, into low Earth orbit in October 1957. Radio amateurs across the globe could easily tune in to the satellite's chatter.

The US Congress created ARPA, the Advanced Research Projects Agency, in 1958 at least in part as a response to the threat posed by Soviet advances. Ever since, the ARPA mission has been to invest in breakthrough technologies" and to "catalyse the development of new capabilities" in order to prevent and create technological surprise. ARPA added an extra D in the title for Defence in 1996, and has ever since been known as DARPA.

Throughout its existence, the agency has relied on a number of principles that distinguish it from other funding bodies and research initiatives:

While DARPA research is done for defence purposes, in many cases its impact is later felt more dramatically in the civil sector — both nationally and globally. Through the visionary leadership of its key people, and the broader ecosystem it catalysed, DARPA brought about the digital age. ARPA set in motion an intricate academic and economic engine that eventually produced innovations such as GPS and the Internet, and turned the science fiction of self-driving cars into a viable future.

In 2019, the UK government announced plans to set up a research funding agency, modelled broadly on the early days of ARPA. In January 2020, leading UK think tank Policy Exchange published a report on this opportunity. The report, "Visions of ARPA", has insightful essays from top UK science policy names, great introductory notes on ARPA, a brief history of UK science policy, and a set of (fairly general) recommendations.

Principal Cast

UK ARPA

Dom Cummings Writes a Blog

The story of UK ARPA begins with a peculiar character. Dominic Cummings (born 1971) studied history at Oxford, graduated in 1994, and then spent a few years doing odd jobs home and abroad. At the turn of the millennium, Cummings turned his attention towards British politics and ended up working in various Conservative campaigns for the next 20 years. He was Michael Gove's chief adviser, ran the successful Vote Leave campaign, and then ultimately found himself in the Cabinet Office as Boris Johnson's chief special adviser.

Never charismatic or likeable enough for political office himself, Cummings found his true calling working in the shadows of shrewd political opportunists. He was able to combine his own nonconformist disposition with an appreciation for the nationalist political right, and developed a Brexit message that proved effective. In Boris Johnson, Cummings found a fellow "get it done" results-first believer with great ambitions and little patience. As the doors to No 10 opened for the two in late 2019, the stage was set for Cummings to bring his radical reforms to life.

In his writing, Cummings has repeatedly argued against bureaucracy and bloated government. In Cummings' view the government, as it stands, is no longer able to carry out complex projects of any kind: the capture by the civil service is complete. If only things were done his way, all would be well. Education, the civil service, policy- and decision-making — it all had to be overhauled, and as quickly as possible. Unsurprisingly, in the real world these plans were met with considerable resistance.

It's not difficult to see why Cummings' message would sound appealing to upstart Conservative politicians who look back to the golden age of the Empire and the powerful political personalities of decades past. Cummings the historian wants us to go back to the era of mega-projects and strong leaders, a simpler time when policies from the central government could really move the nation. He wants to explore what is possible with the right mindset, supercharged with modern information technology.

However, as I'm sure many political strategists before Cummings have learned, the civil service is a formidable adversary. In government, it's much easier to build new things than to shape something old to your liking. And so we arrive at Cummings' vision for the ultimate quango. If there could be a new organisation, a government body for great projects and radical research, what would it look like?

It's clear Cummings has been thinking about research funding and the legacy of the ARPA era for a long time, probably ever since his Oxford years. A late 2014 blog post calls for a new agency outside Whitehall HR and EU procurement rules. It's a nice introduction to both Cummings' interest in ARPA, and his erratic style of writing.

In 2017, Cummings wrote more on the ARPA mentality and the extraordinary wealth generated by the wider community that it seeded. ARPA funding brought together a whole generation of American researchers, who eventually delivered what was to become the foundations of the information age. Xerox PARC and other commercial ventures had an important role in that unfolding, but it was the agency and its inspired leadership that set the stage.

"As we ponder the future of the UK-EU relationship shaped amid the farce of modern Whitehall, we should think hard about the ARPA/PARC example: how a small group of people can make a huge breakthrough with little money but the right structure, the right ways of thinking, and the right motives." — Dominic Cummings, in his mashup post on the Brexit referendum and the ARPA story (2017)

In his 2017 post, Cummings quotes at length from Alan Kay, one of the legends of that story. Kay speaks about the "almost invisible context" of early ARPA, that "catalysed so many researchers to be incredibly better dreamers and thinkers". ARPA "funded people, not projects" and promoted "visions rather than goals". Research today is about avoiding failures, when it should be about trying to "capture the heavens". Cummings took Kay's argument, mixed in ideas from a few other visionaries, and packaged it all up in a Government policy proposal in his distinctive style.

"I think it is possible to create something new that could scale very fast and enable us to do politics and government extremely differently." — Cummings, ibid.

A New Agency

Cummings further developed his vision for UK ARPA in 2018. He produced a short blog post that barely introduces his 50 page screed of a vision for ARPA, an attempt to bring together two of his blog series.

In September 2019, only a couple of months after Boris Johnson assumed office, No 10 convened a round table discussion on what a UK ARPA might look like. Jo Shanmugalingam of BEIS was part of the initial discussions, Alexandra Jones from BEIS sent out the invitations on behalf of then secretary Andrea Ledsom. This discussion paper starts with a Kay quote and then outlines what is missing from the UK R&D sector, and how a new agency on the US ARPA model could make a difference.

What might a UK ARPA do?
What can we learn from the original ARPA model?
Are there gaps in the current UK funding system when it comes to funding ground-breaking,
  high risk / high return research? What could success look like for ARPA?
What could its approach be to:
  1. funding
  2. selecting people/ programmes?
  3. specific themes such as net zero? — No 10, meeting agenda for 25 September 2019 (FOI request)

Her Majesty The Queen delivered a most gracious speech to both Houses of Parliament on 14 October 2019. This speech outlines the government's programme for the parliamentary year, which this time included provisions for the development of a new funding agency. The detailed briefing that accompanies the speech elaborated that the government is indeed preparing a UK ARPA:

"[We are] backing a new approach to funding emerging fields of research and technology, broadly modelled on the US Advanced Research Projects Agency. We will work with industry and academics to finalise this proposal." — The Queen's Speech briefing, lobby pack for the opening of Parliament on 14 October 2019 (FOI request)

It was starting to look like Cummings would have his agency. He wanted to get his house in order, too, so he decided to kickstart 2020 by trying to recruit help through unorthodox channels. This is the point where I and much of the Internet heard about Cummings for the first time. There was an unusual character running the show in No 10, and he wasn't happy with Whitehall. The government needed more Bret Victor. (NB: Every administration has tried to reform the civil service to their liking. See also Sir John Kingman's recent presentation on civil service reform.)

In March 2020, the government published the 2020 budget, with significant provisions of new funding for the sciences. Also included was a statement that the government will invest at least £800 million in a new "blue-skies funding agency", modelled on the US ARPA. "This agency will fund high-risk, high-reward science."

As the spring rolled on and the highly infectious virus situation in the East started to turn into a global pandemic, the Commons' Science and Technology Committee resolved to inquire into the nature and purpose of this new UK research funding agency. Much later in 2020, the Lords also opened up an inquiry on related research and innovation policy, focusing on Catapults.

The pieces were moving. Everything was set up, Cummings had the support and the funding. All he had to do was articulate what exactly he wanted to do with UK ARPA and then make it a reality. But first, somebody had to figure out how to put it together and pass it through Parliament.

Inquiries

The Commons Select Committee for Science and Technology has an open inquiry on the topic "A new UK research funding agency". The Lords Select Committee for Science and Technology have a related one on the role of innovation Catapults. The committees do a number of things, but most interestingly for the general public, they may gather written evidence and hold oral evidence sessions.

In advance of the oral evidence sessions, the Commons committee called out for written evidence on what a UK ARPA might look like. The committee posed a series of questions for the UK science community:

What gaps in the current UK research and development system might be addressed by an ARPA style approach?
What are the implications of the new funding agency for existing funding bodies and their approach?
What should be the focus be of the new research funding agency and how should it be structured?
What funding should ARPA receive, and how should it distribute this funding to maximise effectiveness?
What can be learned from ARPA equivalents in other countries?
What benefits might be gained from basing UK ARPA outside of the ‘Golden Triangle’ (London, Oxford and Cambridge)?
 — Commons Select Committee for Science and Technology, call for evidence (parliament.uk)

Written submissions to this call, published at the committee's discretion, are intriguing, though quite numerous and somewhat repetitive. In the case of UK ARPA, the committee received contributions from individual professor level senior researchers, a good many universities, all kinds of agencies inside and outside the government, businesses, think tanks, and more.

Some highlights from the written submissions, many have excellent further reading references:

Several news outlets report on policy highlights, including oral evidence sessions. See, e.g., Science Business or Research Professional News for headline coverage.

Oral Evidence Sessions

The Commons Science and Technology committee has held three public evidence sessions on the subject "A new UK research funding agency". The Lords have met twice on Catapults, with another session scheduled for January. Recordings from these events — and many more — are available at Parliamentlive.tv. A query like this should yield all of the recordings.

I find the oral evidence sessions of the select committees fascinating. Given the subject is of interest to me for a variety of reasons, I find the discussion itself interesting, but in addition to the content, there's all this procedure that I find charming. Many great characters in the Commons and the bodies, plus the external experts. Even the Lords are gamely — I've included one entertaining session from the Catapult review as well.

Commons CST (18 March 2020): 2020 Budget Spending Review (non-inquiry)

Commons CST (9 September 2020): University Research Funding (non-inquiry)

Commons CST (23 September 2020): A new UK research funding agency, first hearing

Commons CST (7 October 2020): A new UK research funding agency, second hearing

Commons CST (11 November 2020): A new UK research funding agency, third hearing

Lords CST (8 December 2020): Contribution of innovation Catapults to delivering the R&D roadmap

Cummings Exit, Stage Left

On 14 September 2020 Business Insider reported that Dominic Cummings' top candidate to lead the UK ARPA, physicist and computer scientist Michael Nielsen had declined the position. This is well before Solloway's hearing in the CST committee, during which she made it clear that the Government still didn't have a solid plan on how to go about putting UK ARPA together. One might conclude from this that Cummings had an active role in shaping the plans, and didn't want to commit to anything prematurely.

Nielsen apparently felt that he would not have enough freedom to run the agency as he wished. This can be seen as a comment on the incomplete state of the UK ARPA bootstrapping effort, but it's possible that interference from Cummings himself or the wider administration would have been the main concern. As Dr Dugan explained about her own appointment to Wellcome Leap, the focus of the mission made a huge difference to her. High calibre people see through Cummings' half-baked plans.

Failing to get Nielsen on board was the beginning of the end for Cummings' tenure at No 10. All through his stay, Cummings was an easy target for politicians and the media alike, and he got tarred and feathered for his lockdown manoeuvres. More problematically, his personality and frustration with the civil service and politicians probably got the best of him, and on 13 November 2020 Boris Johnson had to let his special adviser go. It was reported that Cummings had spoken against the prime minister.

The idea man's departure meant that UK ARPA was left in a bit of a limbo. Some speculated that the mission would be "kicked in the long grass", postponed indefinitely. Instead, the Government was quick to reaffirm its commitment to both increased funding for science and research in general, as well as continued support for the new "high-risk, high-payoff" funding agency. Indeed the push to increase research funding predates both Cummings and Johnson.

On 25 November 2020, the Chancellor Rishi Sunak published the 2020 Spending Review, which had and item about extra funding for UKRI, the initial setup funds towards a new funding agency. In other words, the new agency will be incubated inside the existing research umbrella.

"The chancellor also confirmed that Cummings’ pet project, a UK version of the US Defense Advanced Research Projects Agency, is still going ahead, with a modest start-up budget of £50 million allocated to UK Research and Innovation for next year." — Kieron Flanagan, for Research Professional

As the strange year 2020 draws to a close, we still have no details on what UK ARPA will look like. In any case, it will be interesting to read the Commons STC report on the new agency once it is published, and of course, to see what kind of an agency the Government will produce in 2021.

As for Dominic Cummings, he has certainly demonstrated a talent for picking up interesting ideas, and also some ability in seeing how things might fit together. He was able to convince a lot of people about a novel idea that might well amount to something. For all his faults, that is to be commended. And something tells me we haven't seen the last of him yet.