USING MESSENGER RNA to make vaccines was an unproven idea. But if it worked, the technique would revolutionise medicine, not least by providing protection against infectious diseases and biological weapons. So in 2013 America’s Defence Advanced Research Projects Agency (DARPA) gambled. It awarded a small, new firm called Moderna $25m to develop the idea. Eight years, and more than 175m doses later, Moderna’s covid-19 vaccine sits alongside weather satellites, GPS, drones, stealth technology, voice interfaces, the personal computer and the internet on the list of innovations for which DARPA can claim at least partial credit.
It is the agency that shaped the modern world, and this success has spurred imitators. In America there are ARPAs for homeland security, intelligence and energy, as well as the original defence one. President Joe Biden has asked Congress for $6.5bn to set up a health version, which will, the president vows, “end cancer as we know it”. His administration also has plans for another, to tackle climate change. Germany has recently established two such agencies: one civilian (the Federal Agency for Disruptive Innovation, or SPRIN–D) and another military (the Cybersecurity Innovation Agency). Japan’s interpretation is called Moonshot R&D. In Britain a bill for an Advanced Research and Invention Agency—often referred to as UK ARPA—is making its way through Parliament.
An agency needs agency
As governments across the rich world begin, after a four-decade lull, to spend more on research and development, the idea of an agency to invent the future (and, in so doing, generate vast industries) is alluring and, the success of DARPA suggests, no mere fantasy. In many countries there is displeasure with the web of bureaucracy that entangles funding systems, and hope that the DARPA model can provide a way of getting around it. But as some have discovered, and others soon will, copying DARPA requires more than just copying the name. It also needs commitment to the principles which made the original agency so successful—principles that are often uncomfortable for politicians.
On paper, the approach is straightforward. Take enormous, reckless gambles on things so beneficial that only a handful need work to make the whole venture a success. As Arun Majumdar, founding director of ARPA–E, America’s energy agency, puts it: “If every project is succeeding, you’re not trying hard enough.” Current (unclassified) DARPA projects include mimicking insects’ nervous systems in order to reduce the computation required for artificial intelligence and working out how to protect soldiers from the enemy’s use of genome-editing technologies.
The result is a mirror image of normal R&D agencies. Whereas most focus on basic research, DARPA builds things. Whereas most use peer review and carefully selected measurements of progress, DARPA strips bureaucracy to the bones (the conversation in 1965 which led the agency to give out $1m for the first cross-country computer network, a forerunner to the internet, took just 15 minutes). All work is contracted out. DARPA has a boss, a small number of office directors and fewer than 100 programme managers, hired on fixed short-term contracts, who act in a manner akin to venture capitalists, albeit with the aim of generating specific outcomes rather than private returns.
The first challenge for the new ARPAs is to secure the breathing space required for such experimentation. SPRIN–D illustrates how difficult this can be. The concept was approved by Germany’s cabinet—“and then the Federal Court of Auditors came along,” sighs Barbara Diehl, SPRIN–D’s chief partnership officer. After the auditors issued their recommendations, the agency lost its exemption from standard public-sector procurement rules and pay scales, restricting who it could hire and the sorts of risks it could take. Existing government ministries exert influence through the agency’s board, stymying radicalism, says Ms Diehl. Dominic Cummings, a former aide to Boris Johnson, Britain’s prime minister, who demanded a British agency as a condition of his employment, has said he is concerned by the provisions for ministerial oversight in the legislation creating it.
Without freedom from political interference, the risk-taking instincts of those at the cutting edge are curbed. The administrative and research directors of Germany’s Cybersecurity Innovation Agency recently quit, frustrated by political interference. In America the homeland-security ARPA was established in 2002, but has been hamstrung by power struggles in the department that gave it its name. “It has never been allowed to make independent decisions, it has never been allowed an independent budget,” says an observer. There is a debate about whether the Biden administration’s health ARPA (ARPA–H) ought to stand alone, or be part of the National Institutes of Health (NIH). The latter would be less of a legislative challenge, but may infringe its independence.
DARPA’s budget in 2020 was $3.6bn, equivalent to just 8% of the NIH’s. If all goes to plan, ARPA–H will be on a similar scale, but none of the others receives such funding (ARPA-E got $425m last year, roughly as much as one of DARPA’s six offices). Since the model works by making lots of bets in the hope that a few will come off, stingier funding means fewer wagers, which reduces the chance of success and thus of continued political support. This is especially true given the difficulty of measuring progress. As a paper by Pierre Azoulay of the Massachusetts Institute of Technology (MIT) and his colleagues notes: “It is impossible to accurately measure the incidence of one-in-a-thousand ideas, much less one-in-a-million ideas, on a timescale relevant to political decision-making.”
The new agencies must also work out how to get their innovations out of the lab. There is a close relationship between DARPA and the Department of Defence, which is a customer for its work. But other agencies lack such a pipeline. Research by Anna Goldstein at the University of Massachusetts, Amherst, and her colleagues finds that, although new “cleantech” companies sponsored by ARPA-E produce more patents than others, they are not more likely to raise venture capital, be acquired by larger firms or list on public markets. So far, at least, the agency’s innovations have struggled to leap into the real world.
When ARPA-E began in 2009 the hope was that venture capitalists would pick up innovations emerging from it. They have proved reluctant. Energy technologies take far longer to reach the market than venture capital’s favourite investment, software. ARPA-E has thus tweaked the DARPA model to add a “tech-to-market” team, to guide projects through the industrial jungle. Last year it began handing out grants to promising previous award-winners seeking to grow. William Bonvillian, a science-policy expert at MIT, suspects one missing ingredient is simply time: “We created the internet in ’69. It didn’t scale up until ’91 or ’92. So we’ve just got to get used to it taking a while.”
ARPA-H may face similar difficulties. It is based on the idea that the NIH is too conservative, focusing on biology at a time when many life-science breakthroughs happen where biology, chemistry and computer science meet. Mikko Packalen and Jay Bhattacharya, of Waterloo and Stanford universities respectively, provide supportive evidence, finding that the NIH’s funding of work building on new advances has declined. ARPA-H’s other goal, though, is to pump money into treatments for rare diseases, eschewed by the private sector because of limited moneymaking opportunities. As with ARPA-E, this lack of commercial interest may make the transition from innovation to the real world tricky.
Michael Stebbins, who was an official in Barack Obama’s administration, and is an advocate for ARPA-H, hopes that someone from DARPA can be recruited to lead the new agency. Replicating DARPA’s freewheeling culture is such a challenging task that there have been times when DARPA itself has failed. It went through a fallow period in the late 1960s and early 1970s, and many feel its ambitions have been trimmed back in recent decades—minimising failures, but also successes.
The defence focus also has an inbuilt advantage. By failing to build a terrible weapon, American leaders can reassure themselves that their adversaries won’t either. There is no such reassurance in failing to cure cancer. But that has not been enough to dissuade politicians in America, Britain, Germany and Japan. The lesson many have learnt from DARPA is that mere difficulty is no reason to avoid something. It may even be a reason to do it. ■
Correction (June 5th): A previous version of this article said ARPA-E awarded grants of up to $150,000. In fact there is no limit.
A version of this article was published online on May 30th 2021
This article appeared in the Science & technology section of the print edition under the headline “Cloning DARPA”