June 18, 2018

One of the major legacies of WW2 – on both sides of the Atlantic – was a profound awareness that science and technology matter. They played a huge role in winning the war, and they must have an equal place in winning the peace.

In the United States, President Roosevelt’s wartime science guru, Vannevar Bush, laid out a plan for a post-war government-science partnership that resulted in the National Science Foundation.1 The core idea: governments need to fund basic research.2

A decade later, the Soviet Union’s launch of the world’s first satellite, Sputnik, in 1957, sent shock-waves through America and refocused government minds on innovation. How on earth could this despised centrally-managed regime have beaten the US into space?

In February of 1958 President Eisenhower set up the Advanced Research Projects Agency (DARPA – the D for Defense was added later) with a simple mission: to make sure the US would never again be surprised by developments in new technology.

DARPA’s biggest claim to fame is the internet. Or, as it was originally called, the ARPANET.

From the start, DARPA has operated as a unit of the Department of Defense, though it’s been free to think long-term and not feel bound by the need to develop weapons. It’s a nimble organisation, with only a couple of hundred employees – and a budget of $3 billion.

Its approach is to look way into the future and find and fund the smart people who will get us there. Some of its work is seriously secret.3 But many of its innovations are well known.

Its most recent claim to fame is self-driving cars. Big car companies are now sparring with Silicon Valley start-ups to turn their experimental vehicles into production models, but it was as recently as 2004 that DARPA kick-started this rapid technological development. Instead of splashing out large sums to university researchers, DARPA decided to run a competition. Anyone could compete, their cars had to trundle across 142 miles of desert to get to the finish line, and the winner would take home $1 million. Incredibly given where we are now, every single entry in that first round failed to make it. But DARPA persisted, and in 2005 five vehicles crossed the finishing line, with Stanford University taking home the prize. DARPA moved on to an urban course, and then decided its support was no longer needed.

Of course, DARPA’s biggest claim to fame is the internet. Or, as it was originally called, the ARPANET. Back in the 1960s there was a simple need: a system that could not be destroyed by a Soviet nuclear attack. Hence the idea of joining up computers – initially between universities – in a network that did not depend on any single hub. And that burning platform led to DARPA funding 70% of all US computer research in the early 60s.

It was two key thinkers working on the project, Vint Cerf and Bob Kahn, who came up with what we now recognise as the internet. The story goes like this: Cerf was hanging around in a hotel lobby and came up with the idea of a new ‘communications protocol’ that would enable networks to connect with each other. It came to be known as TCP/IP. First tested on ARPANET in 1977, it’s the standard we’re still using today. 4

The outlines of this story are well-known, but its implications less so. Because DARPA and other US government funding agencies did not just invent the internet – they devised the tools that made the digital economy possible; and enormously profitable.

As economist Mariana Mazzucato demonstrates in her celebrated book The Entrepreneurial State, for example, every key component in the iPhone was developed by the United States government. 5 And it’s quite a list. Aside from the internet itself, which is the basis of every single 21st century high-tech company (including the 60+ ‘unicorns’, Silicon Valley start-ups valued at more than $1 billion), the whole succession of technologies that power our devices has emerged from publicly-funded science. From touch-screens to voice-activated Siri (and her equivalents) we have government investment to thank.

In fact, the entire eco-system of the internet was funded by DARPA. This article on “10 amazing DARPA inventions” starts off with the internet, and moves onto hyper-text linking (which led to the World-Wide Web); an early version of Google street-view; the Unix operating system; and GPS.

And while it’s true that Sergei Brin and Larry Page wrote the algorithm that created Google, what’s less well-known is that they made their key discoveries with funding from the National Science Foundation.

If American taxpayers funded the research that led to the development of the iPhone, shouldn’t those taxpayers share in the rewards?

Mazzucato’s argument is simple. As a matter of history, the core technologies that enabled Silicon Valley were not developed by Silicon Valley geniuses funded by Silicon Valley venture capital. It was the federal government, chiefly through DARPA, who made the development of these now ubiquitous technologies possible, and who gifted them to anyone who was interested.

American (and, of course, increasingly, Chinese) companies have been very interested, and their vast profits are the fruit of investment by the US taxpayer. Public risk, private profit.

Let’s take Apple, who have sold over a billion iPhones: current net profits are around 20%, its market capitalisation is edging towards one trillion dollars, and at the moment it has nearly $300 billion just sitting in the bank. If anyone has done well out of US public science investment, it’s the company Steve Jobs built.

This story of how a government agency, DARPA, jump-started the digital economy needs to be better known, and one signal service of The Entrepreneurial State has been to help puncture the illusion that the Silicon Valley geniuses came up with it all themselves. It also raises some big questions about who drives innovation, and how it is paid for.

DARPA and other US government funding agencies did not just invent the internet – they devised the tools that made the digital economy possible

If American taxpayers funded the research that led to the development of the iPhone, shouldn’t those taxpayers share in the rewards? Mazzucato makes a strong case for a clawback, which could be recycled into further public science investments. She may be right, if it can be done in a way that does not inhibit the next round of entrepreneurs. But it’s important to note that these companies already benefit the economy – and therefore the public purse. They pay tax, have created a lot of high-paying (as well as low-paying) jobs, and had the usual multiplier economic impact on their localities and supply chains that vibrant innovative companies produce.

Nonetheless, a windfall tax on digital companies making outsize returns – ring-fenced to fund new research – could be one solution to the Mazzucato challenge. A form of DARPA dividend. It’s worth noting that Apple’s current cash mountain alone could fund DARPA one hundred times over.

A more fundamental question is whether governments should see themselves as “entrepreneurial”, and set out to do just what DARPA did, as a matter of practice. The idea that governments should set industrial policy is hotly contested, with critics tending to characterise such an approach as ‘picking winners and losers’.

It’s ironic that while European governments (and the Japanese) are happier talking like this, it was in the United States, where the rhetoric of free-market capitalism tends to shut down conversations about industrial strategy, that the DARPA-digital lift-off actually took place. There’s a good argument that governments can’t avoid industrial policy choices – either they have one they talk about, or they have one by default. On the other hand, there’s also an argument that the DARPA-digital story is unusual, perhaps unique, and certainly not something that a government could set out to replicate.

Maybe it was unique, but that’s no reason to sit on our hands. Governments have nothing to lose and everything to gain by funding far-sighted research. One of the beauties of the DARPA model has been to keep this separate from the National Science Foundation/National Institutes of Health funding, which suffers from traditional peer review (fellow-experts who tend to strangle really fresh thinking at birth), and also from pressures to make it translate readily into business ideas and profits. The internet is worth trillions, but it took a generation before that became clear.

A windfall tax on digital companies making outsize returns – ring-fenced to fund new research – could be one solution to the Mazzucato challenge

While there have been huge commercial and economic benefits from these technical advances, it is, however, important to remember that neither Vannevar Bush nor President Eisenhower had profit on their minds as they planned science and technology’s central role in America’s future. Their focus was security – keeping ahead of the Soviet Union. Their efforts succeeded beyond their wildest dreams, in part precisely because the focus was on long-term innovation rather than economic returns. There will always be private capital for the latter.

The clearest lesson from The Entrepreneurial State is that the blue-sky thinking of the mid-20th century paid off in the 21st – but that the original objectives of that research investment could not have foreseen the digital revolution that it spurred.

  1. The report that led to the National Science Foundation (and its equivalent in every developed nation) – Science, the Endless Frontier. His brief and extraordinary essay published in 1945 in Atlantic Magazine, As We May Think actually envisages an “internet” as a hyperlink system based on microfiche! “Consider a future device…  in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory.”
  2. The parallel National Institutes of Health, first funded for cancer research, is now the big brother as bioscience has flourished. Current funding: NSF $8 billion. NIH $35 billion.
  3. I was once asked to help a company get funding from DARPA, and I hired a former DARPA programme manager to make it happen. Then he explained that if DARPA really liked the company’s hush-hush product they might just decide to classify it – make it a state secret. And the company decided to look elsewhere.
  4. For more details, see the article on ARPANET in Encyclopaedia Britannica
  5. Mariana Mazzucato, The Entrepreneurial State. Debunking Public vs. Private Sector Myths. Revised edition, Public Affairs, 2015