European Darpas won’t create new industries overnight

Barely a week goes by in Europe without a call to set up an equivalent of the US Defense Advanced Research Projects Agency (Darpa), the body credited with some of the 20th century’s most transformative inventions, to end the continent’s technological vassalage. 

The EU’s high priest of economic decline, Mario Draghi, and economics Nobel Prize-winner Philip Aghion, have made the case for more Darpa-like agencies in Europe. Germany and the UK have already set up new civilian innovation agencies modelled, to some extent, on Darpa. 

The Netherlands is also planning its own agency, and the European Innovation Council (EIC) is piloting technology challenges it claims are Arpa-style, referring to the Darpa model but without a defence focus.

But what can Europe actually expect from such agencies? Science|Business spoke to Benjamin Reinhardt, a New York-based entrepreneur who wrote one of the most comprehensive guides to how Darpa actually works. He’s been consulted by Dutch officials on setting up their own agency, and had just returned from a visit inside Darpa. 

His first piece of advice for Europe? Do not slavishly replicate Darpa, but build something new. “Don't try to copy the American thing,” he said. 

To be fair, none of the European Darpa-flavoured innovation agencies, from Germany’s Sprind to the UK’s Aria, are attempting a one-to-one simulacrum.

Instead, they’ve taken certain elements of Darpa, such as its speed, low paperwork and programme managers, people given the freedom and money to explore technological ideas without layers of decision by committee. 

The EIC, meanwhile, is largely a grant and equity-making body, with very weak programme managers who sometimes have to oversee hundreds of projects. But despite these organisational differences, at least rhetorically, Europe is still very much enamoured with Darpa. 

While Darpa is a worthwhile model, “Europe should do its own institutional experiments,” said Reinhardt, who is also the founder of Speculative Technologies, a not-for-profit research organisation with a focus on materials and manufacturing. 

Reinhardt also stressed the sheer amount of risk Darpa takes: it has to fund some very whacky-sounding failures on the way. For every successful project, “there are several dozen at least, that were complete flops.” 

Don’t centralise

One question is whether Europe should centralise its Darpa-style innovation agencies. In Brussels, it’s sometimes argued that combining budgets will allow a leading European agency the critical mass of funding it needs to make risky bets on unusual projects. Having many national agencies risk fragmentation. 

But for Reinhardt, the more innovation agencies Europe tries the better. A spread of bets is more likely to hit on a model that works.

“I'm a deep institutional pluralist,” he said. He worries that EU states won’t try their own agencies because of a complacency that “the EU will just take care of it.” 

What’s more, Europe can arguably try out more types of innovation agencies than the US, because European countries have the power to set up their own bodies in a way US states don’t. 

Indeed, Reinhardt is critical of how the US federal government has largely monopolised responsibility for research since the Second World War. “You've seen much more centralisation across the board,” he said. 

Don’t expect immediate returns

European politicians also shouldn’t expect Darpa-style bodies to quickly create new industries or economic growth.

“It is very good at inventing new technologies,” he said of the Darpa model. “It is not necessarily the best mechanism for creating new industries, at least not on the timescales that governments operate.” 

Despite its economy-transforming inventions, Darpa’s goal is not to boost the US GDP. Instead, it’s to prevent what the US military calls “strategic surprise.” Tt was set up to prevent a repeat of the Sputnik shock of the 1950s, when the US found itself losing the early space race. 

Value capture?

One “awkward” problem for Europe is that a technology invented in a European Darpa might be commercialised elsewhere, Reinhardt went on.

“One of the reasons that the US is so able to capture the value that is created by Darpa is because it's so big,” he said. This is a particular problem for smaller state like the Netherlands, he said: how do you make sure enough of the benefits stay within national borders?

Related articles

• EIC to pilot ARPA-style challenges in 2026 
• The Netherlands take a step towards establishing a Dutch Darpa 
• Inside Germany’s Sprind innovation agency 

Another question is start-up funding. European innovation agencies, such as Sprind or the EIC, often fund early-stage technology companies as a part of their mission.

But Darpa-style projects are more about “unlocking a research bottleneck” than backing an idea already on the path to commercialisation, said Reinhardt. 

It’s not that supporting start-ups is a bad idea. It’s just that they may succeed anyway, without innovation agency intervention. By contrast, Darpa in the 1950s and 1960s seeded pre-commercial technologies that would have come along at least a decade later, or never, said Reinhardt. “Maybe that will happen with Sprind, but I'm more sceptical.”. 

Darpa loses its shine?

Reinhardt’s guide to Darpa isn’t without criticism. There’s a “pretty stark” difference between the world-changing leaps it seeded before 1972, such as weather satellites, personal computing, and the internet, and the impressive but more modest breakthroughs afterwards, such as the Big Dog autonomous robot or the challenge that kicked off self-driving cars.

This apparent post-1972 slowdown may have been due to political pressure that clipped Darpa’s wings. Following increased scrutiny of the military after the Vietnam war, Darpa was placed under tighter Congressional oversight, and limited to more strictly defence-related projects.

What’s more, Darpa’s system of contracting out research to outside labs and companies no longer works so well in the US. “There are many fewer independent research organisations,” Reinhardt said. “There's been massive consolidation, at least in the US.” 

Largely gone are quirky independent public labs or R&D-focused small businesses. What’s left are universities, which chase academic papers, and start-ups, which, since the birth of the venture capital system in the 1980s, are under “extreme pressure” to make money and don’t want to take on a Darpa research “side quest.” 

The Darpa model “does not work as well in the 21st century as it did in the 20th,” he said.

For this reason, Reinhardt’s Speculative Technologies is trying to raise money to build its own physical lab. “We actually need to be a place that can hire the people who are actually doing the research,” he said.