资源环境科技发展态势分析平台

The U.K. government has released its plans for emulating the storied high-risk, high-reward U.S. funding agency, the Advanced Research Projects Agency (ARPA). During the 2019 election campaign, the Conservative government promised to set up such an agency, and Prime Minister Boris Johnson’s former adviser Dominic Cummings pushed for it. Now, it has a name: the Advanced Research & Invention Agency (ARIA), and a confirmed budget. Business secretary Kwasi Kwarteng announced today it would be funded by an initial £800 million over 4 years, and would begin to disburse grants by 2022.

Unlike ARPA, now known as DARPA because it specializes in defense-related research, ARIA so far lacks a focus, which concerns some experts. “The fabric of an ARPA organization is its mission,” says Anna Goldstein, a science policy researcher at the University of Massachusetts, Amherst, who provided evidence on ARPA to U.K. politicians. Without that, she says, “ARIA is a solution in search of a problem.”

The government will soon introduce legislation in Parliament to enshrine the agency in law, aiming to give it more permanence than pet political projects sometimes enjoy. It will sit within the Department for Business, Energy and Industrial Strategy, but remain independent from the main research funding agency, UK Research and Innovation. The new law may also allow ARIA to sidestep normal spending rules, audits, and even freedom of information requests. But those checks and balances don’t hinder risk taking, says James Wilsdon, a science policy researcher at the University of Sheffield. “You don’t have to be a scathing political critic of the government to raise a quizzical eyebrow about attempts to put chunks of the research budget in a place where they’re not subject to any of those checks and balances.”

DARPA, created as ARPA in 1958 after the Soviet Union launched the Sputnik satellite and the United States panicked about falling behind on military technology, is credited with innovations in the development of the internet, GPS, and robotics. Such is the power of the brand name that it spawned spinoffs for intelligence, homeland security, and energy technology (ARPA-E). Unlike traditional funding agencies, which hand out grants through peer review, DARPA and its ilk give program managers the freedom and budget to seek out promising work, fund it with no bureaucracy, and cut it off early at signs of failure.

Because of limited public data, DARPA’s triumphs can only be tracked through a series of anecdotal success stories, Goldstein says. But in 2020, Goldstein and her colleagues analyzed the more comprehensive data from ARPA-E and found that projects it funded led to more patents than those it rejected or comparable projects that found funding elsewhere—a hint that the agency has a knack for hitting on innovative research.

In announcing ARIA, Kwarteng nodded to two possible focuses: disease outbreaks and climate change. “Those would both be excellent choices,” Goldstein says. “Why not create two ARIAs and do both?” Recruitment for the agency’s leaders will begin in the next few weeks. But that is risky without having determined the purpose of the agency, Wilsdon says, because different goals demand different leadership.

The new agency will likely only take up about 1% of the government’s total research funding, which Johnson’s government has pledged to increase drastically over the coming years. But the funds are a “reasonable starting point,” says Nancy Rothwell, president of the University of Manchester and co-chair of the government’s advisory Council for Science and Technology. It’s enough to do something useful, she says, but not a reckless gamble. “Time will tell whether or not [its budget] should increase.”