The director of the new “high-risk, high-reward” research funding agency, ARIA, should not be a “bog-standard vice-chancellor” but a “brilliant” person with “good taste in scientific ideas and scientific researchers”, according to the prime minister’s former advisor, Dominic Cummings.
Mr Cummings, who took a job in Number 10 partly to drive forward the “ARPA-like entity”, gave an account of his contribution to the new Advanced Research and Invention Agency (ARIA), to the Science and Technology Committee yesterday (Wed 17 March).
Among his suggestions for the infant agency, Mr Cummings was adamant it should have “extreme freedom”, a broad mission statement, little ministerial oversight and operate with a “very flat organisational structure” with just a few key decision-makers.
The agency is dependent on finding the right people to run it, Mr Cummings said. “[The ARIA model is] you have a director in charge of it, and then they have good taste in finding people. There’s no alternative to this fundamental problem. You have to have someone in charge who has good taste in scientific ideas and in scientific researchers.”
Referring to great scientists, like those responsible for the Apollo moon landings, Mr Cummings said: “It certainly does seem to me that [these experts]… are very, very odd people. And the most obvious way which [it] will fail, [is] if you just get some bog-standard vice-chancellor and put them in, then obviously it won’t work.”
There are too many restrictions. Certainly, in my model of it, you wouldn’t have ministers anywhere near making decisions about how it spends money, I think that would be a disaster
– Dominic Cummings
The former special advisor (SpAd) to prime minister Boris Johnson said he asked his former boss to agree to four conditions before joining Downing Street, of which planning an organisation based on the Defense Advanced Research Projects Agency (DARPA) in the United States of America was one. He left government last year.
Mr Cummings said: “The prime minister came to talk to me [in my living room] the Sunday before he became prime minister and said would I come into Downing Street to help try to sort out the huge Brexit nightmare.
“I said: ‘Yes if first of all you are deadly serious about getting Brexit done and avoiding a second referendum. Secondly, double the science budget; third, create some Arpa-like entity; and fourth, support me in trying to change how Whitehall works, and the Cabinet Office works because it’s a disaster zone.’ And he said: ‘Deal.’”
Mr Cummings said he would not be involved in ARIA in the future. “I wouldn’t be involved. I’m not seeking to be involved. I wouldn’t want to be involved. I shouldn’t be involved.”
The agency is not expected to be fully operational until 2022.
He favours a small agency free from ministerial and departmental oversight, headed by a director and four trustees. He continued: “You need to strip out all the horrific Whitehall bureaucracy around procurement, state aid, human resources, civil service pay scales, all of these sorts of things. And huge tender processes from the Treasury as well.”
His vision for ARIA differs from the one laid out in the government bill proposed earlier this month, which ministers say “will also ensure [agility] is balanced with necessary accountability and oversight”.
According to the government, ARIA will “proactively” share information about its activities with the secretary of state, who will have “robust” powers to intervene in its operations “in the interests of national security if required”. It will publish annual reports to be scrutinised by Parliament and the National Audit Office.
Asked to comment on the government bill, Mr Cummings said: “There are too many restrictions. Certainly, in my model of it, you wouldn’t have ministers anywhere near making decisions about how it spends money, I think that would be a disaster.
“In my model, it would be extremely simple. You would find a director, you’d have probably [a] maximum of four trustees, so that they are actually real trustees, and they have real control, not one of these normal government things with like 20 people on a board so that no one’s actually exercising serious responsibility for it.
“And you would cut it loose of the rest of the system. So no, I’m not confident about how it will work out.”
Mr Cummings said the plans for ARIA had been watered down by civil servants opposed to projects that are “completely hostile to normal bureaucracies”. “[A]lmost all science funders globally operate in the same way, they use pretty much the same metrics, like papers, they have pretty much the same kind of horrific bureaucracy. They waste huge amounts of time for the researchers in filling out all these forms, and they get pretty much the same results.”
A different type of agency would give the UK competitive advantage, Mr Cummings claimed. Asked if his model would instead lead to nepotism and cronyism, Mr Cummings disagreed, adding: “You’ve got to pick the five people very carefully.”
Committee chair Greg Clark asked Mr Cummings if the goal of ARIA – of empowering scientists to take more risks – could be achieved by increasing general quality-related (QR) funding to universities.
“[U]niversities themselves are also an absolutely massive source of bureaucracy,” Mr Cummings claimed. “And one of the things that in my opinion should happen is that the government should be very aggressive with universities and say, we’ve got a lot of money to hand out, we’re going to double the science budget, you’ve got to do your part; you’ve got to remove all the following nonsense from your maths department, your physics department, your engineering departments, if you want our cash, and create some really strong incentives on the universities.”
Mr Cummings said the £800 million agency budget was “[fine] to begin with”, but he thought it likely more investment would be needed. Extreme freedom, however, is more important to the project than the budget, he claimed, elucidating: “I stress how the money is spent is even more important than just increasing the amount of money.”
The committee asked the former SpAd to explain what he thought the agency should aim to achieve.
By its very nature it should not have a mission statement, Mr Cummings said. “I wouldn’t give it a specific defence focus; I certainly wouldn’t try and import a bunch of buzzwords like AI and quantum and net-zero and things like that. I just think it doesn’t add anything. I would stick to stick to a broad mission statement.”
Mr Cummings said he had sought the opinions of people with experience in DARPA, who agreed “essentially, if [ARIA] isn’t failing, then it’s failing.” Pushed to provide a likely numerical success rate, Mr Cummings hypothesised: “Maybe a third of them are working out brilliantly, and two-thirds of them aren’t.”