Can Canada Out-think the Underpants Gnome?
I recently read a fascinating book called “How to Make an Entrepreneurial State: Why Innovation Needs Bureaucracy “ by Rainer Kattel, Wolfgang Drechsler and Erkki Karo, all of whom are influenced by Marianna Mazucatto, whose work I have discussed here and here. It’s fascinating for two reasons: first, that the book says next to nothing about how making the state more entrepreneurial or why innovation needs bureaucracy, but it is a very inclusive history of the types innovation policy structures of nation-states around the world. Possibly one of the greatest good book/terrible title combos of all times.
I will spare you a general critique of the book – Caleb Watney has already done that in Foreign Affairs, and it’s a must-read – and focus on how the book relates to the Canadian context. And, reader, you will probably not be shocked to learn that the Canadian context for discussion of innovation agencies is deeply impoverished. I count three areas where we can learn some tricks from abroad.
First, the book helpfully discusses the difference between making policy choices and implementing policy and why it is important not to assume success simply because a choice has been made and money spent. In Canada, this is very poorly understood, and deliberately so. Most of Canadian governance is about evading responsibility in the event of non-achievement of results. This is why we are so bad at collecting data, and averse to doing things like actually announcing policy targets as opposed to simply announcing the initiatives themselves. This is not specific to the Canadian discourse on innovation, but it is a specific way in which Canadian discourse on innovation is impoverished.
Second, the book helpfully distinguishes between the functions of innovation agencies/policies, of which there are broadly three. There are agencies which fund innovation, there are agencies which shape markets (for instance, by focusing on procurement and competition policies), and there are agencies which try to work on big-picture system transformation. In Canada we typically only talk about the first of these three. The idea of market-shaping is largely anathema to us because a) we’re pretty conservative when it comes to this kind of intervention and b) even when it comes to using governmental power to shape markets through things like procurement, our governments are so deathly afraid of risk that they never prioritize innovation. And when it comes to system transformation, forget it. So, what happens in Canada are big fights about the way that a single policy level is used to effect change without ever really questioning which other levers we could be pulling.
Third, we are not even particularly interested in the purpose of the agencies we create, whether it be diffusion-oriented (how do we get large numbers of firms to take up new technologies) or mission-oriented (how do we solve problem X). In fact, most of our innovation agencies are really neither because they are framed in terms of how to distribute money to various stakeholders. Probably the ones that come closest to having a mission, per se, are Brain Canada and Genome Canada. The famous CARPA proposal could have had a mission, in the way that DARPA’s mission is about US military supremacy, ARPA-E is about energy, etc. But the backers of this idea pointedly declined to give it a mission, because their assumption was that throwing money around in a diffuse manner would somehow generate product innovation equally well no matter which area was chosen. This comes from a general confusion in Canada between something being “mission-oriented” (solving a coherent set of problems) and it being a “moonshot” (i.e. something huge and out-of-this-world). The Apollo project was mission-oriented, but challenges don’t have to be gargantuan in order to qualify as a mission.
Of late, we do seem to care a little bit about the style of agencies we create, whether they are agile, like DARPA, or stable like NRC or the granting councils, or some combination of the too. But historically, we’ve been pretty consistently Weberian on this: set up stable bureaucracies to hand out money to large, stable agents like universities. The idea of small, agile organizations sheltered from political interference doing something other than hand out flipping great wodges of cash (what Dan Breznitz calls “peripheral agencies”) — while staying under the radar – is not really in our national DNA.
Put all this together and the following results: compared to many countries around the world, the Canadian policy discourse around innovation looks two- or even one-dimensional. We don’t talk about outcomes at all, and very little about agency. Missions? Not really. System transformations? Fuhgeddaboudit, especially if we are talking about social or governmental transformation related to provision of education or social services or even something as obvious as road safety. Links to other related areas, like competition policy or procurement, do occasionally show up in rhetoric. But in practice? Not so much.
I could argue a lot of things in the national psyche are responsible for this (mainly: Canadians aren’t that interested in innovation because they prefer comfort to ambition) but to drop the psycho-analyzing for a second: at a policy-structural level, Canadian governments have assumed that innovation policy = growth policy. Not only that – they also have come to believe that industrial policy, competition policy and (under strong pressure from universities, who know a golden goose when they see one) science policy are also equivalent to innovation policy. It’s all one giant mess, largely because Canadian government innovation initiatives, regardless of who is in power, never really have a logic structure more complicated than that of the Underpants Gnomes on South Park.
Step 1: Give government cheques to people doing gee-whizzy things.
Step 2: ?
Step 3: Growth!
Anyways, this book is really good if you want to understand how other governments make sense of Step 2, how they actually work to think through the how government initiatives affect real-world outcomes beyond the simple act of handing out cheques. I hope a lot of people read it, because until Canadian policy makers can prove themselves more intellectually capable than the underpants gnomes, things aren’t going to get any better.
Fortunately, we can test the Underpants Gnomes v. Government of Canada policymaking by examining the recently-created Canada Innovation Corporation. More on that tomorrow.
The State is not Entrepreneurial
If you’re interested in innovation policy, and haven’t spent time under a rock for the last couple of years, you’ve probably heard of Mariana Mazzucato. She’s the professor economics at the University of Sussex who wrote The Entrepreneurial State, which is rapidly becoming the source of an enormous number of errors as far as science and economic policy are concerned.
Mazzucato’s work got a fair bit of publicity when it was released for pointing out that a lot of private sector tech is an outgrowth of public sector-sponsored research. She has a nice chapter, for instance, outlining how various components of the iPhone – the touchscreen, the GPS, the clickwheels, the batteries… hell, the internet itself – are based on research done by the US government. This is absolutely bleeding obvious if you’re in science policy, but apparently people out there need to be reminded once in awhile, so Mazzucato found an audience.
Where Mazzucato goes wrong, however, is when she begins to draw inferences; for instance, she suggests that because the state funds “risky” research (i.e. research that no one else wold fund), it’s role in R&D is that of a “risk-taking” entity. She also argues that since the state takes a leading position in the scientific development of some industries (e.g. biotech), it is therefore an “entrepreneurial” entity. From this, Mazzucato concludes that the state deserves a share of whatever profits private companies make when they use technology developed with public science.
There are two problems here. The first is that Mazzucato is rather foolishly conflating risk and uncertainty (risk is tangible and calculable, uncertainty is not). Governments are not a risk-takers in any meaningful sense: they are not in any danger of folding if investments come to naught, because they can use taxing power (or in extremis, the ability to print money) to stay afloat. What they do via funding of basic research is to reduce uncertainty: to shed light on areas that were previously unknowable. Individual companies do very little of this, not just because it’s difficult and expensive (if a company is big enough, that’s not a problem – see Bell Labs or indeed some of the quite amazing stuff Google is doing these days), but because the spillover from such research might allow competitors to reap much of its value (a point Kenneth Arrow made over fifty years ago).
The second issue is that nearly all of the examples Mazzucato offers of public research leading to technological innovation and profit are American, and a fairly high percentage of these examples were funded by the Defense Advanced Research Projects Agency (DARPA). To put it mildly, these examples are sui generis. It’s not at all clear that what works in terms of government investment in the US, with its massive defense infrastructure, huge pools of venture capital, and deep wells of entrepreneurial talent, hold very many lessons for countries like Canada, which are not similarly endowed. Yet Mazzucato more or less acts as if her recommendations are universal.
The book’s recommendations amount to: government should own a share of young innovative companies by gaining shares in return for use of publicly-funded knowledge. But this is pretty tricky: first, there are very few cases where you can draw a straight line from a specific piece of publicly-funded IP to a specific product, and even where you can, there’s no guarantee that the piece of IP was publicly-funded by your local government (Canadian start-ups benefit from knowledge that has been created through public subsidies in many different countries, not just Canada). And while there’s a case for greater government investment in emerging companies (economist Dani Rodrik makes it here for instance), the case is not in any way predicated on government investments in R&D. In Canada, the CPP could adopt such a policy right now if it wanted – there’s no reason why it needs to be linked to anything Industry Canada is doing in science funding. To the contrary, as Stian Westlake points out, countries that have been most successful in converting public science investments into private hi-tech businesses eschew the idea of equity in return for scientific subsidies.
Worst of all – though this is not entirely Mazzucato’s fault – her argument is being picked up and distorted by the usual suspects on the left. These distortions are usually variations on: “Someone said the state is entrepreneurial? That means the state must know how to run businesses! Let’s get the state more involved in the direction of the economy/shaping how technology is used!” This way disaster lies.
So, Mazzucato did everyone a service by forcefully reminding people about the importance of publicly-funded R&D to any innovation system. But her policy prescriptions are much less impressive. Treat with care.
Missions and Moonshots
There is a crowd of policy entrepreneurs in Canada – mostly but not entirely Liberal, mostly but not entirely based in Ottawa – who have really cottoned on to the whole notion of innovation. Like many of us who have despaired over successive governments’ lack of cluefulness on this issue, they are dissatisfied with the status quo. Unfortunately, these people are currently marching with wholly unjustified confidence towards policies that are largely buzzword-driven.
It’s not just this ludicrous notion of a “Canadian DARPA”: now the policy-nerd airwaves are full of talk about “Moonshots” and “mission-based innovation”. For convenience, Ottawa’s Public Policy Forum and Toronto’s Brookfield Institute have managed to combine these terms into a single argot-tastic project known as Canada’s Moonshot: Charting A Mission-Oriented Innovation Strategy.
Let me explain why this is nonsense on stilts.
First of all, what is a “moonshot” in innovation? In contemporary Canadian innovation buzzword bingo, it seems to mean “do something big”. What, exactly, we are meant to go big about is unclear, but whatever we are doing now isn’t cutting it, so why not go big (“we must do something, this is something, etc.”). And after all, isn’t there a big literature on moonshots which suggest that they often work out well, have lots of spin off effects, etc?
The problem here is that this argument relies on people failing to understand that what the literature describes as a moonshot and what the buzzword bingo types claim as moonshots are two different things. A moonshot, by definition, is something that a country does in response to a truly existential threat. The Manhattan Project was a Moonshot to end the bloodiest war the world has ever known. The Apollo project was a literal moonshot pushed by the fact that the Americans thought they were about to lose global technical pre-eminence, particularly in rocketry. Give or take some multilateralism and the participation of the private sector, COVID vaccines might be considered a moonshot too.
But Moonshots are a by-product of existential threats (a point made in detail and extremely ably by Mark Zachary Taylor in his book The Politics of Innovation, which I reviewed back here). Countries don’t do moonshots because they wake up one morning and say “hey, let’s do big thing”, they do it because they are deeply terrified of what will happen if they don’t invest heavily in this one complex task. Canada, though, does not face any existential threats. The best the moonshot types can come up with is weak sauce like “saving health care” (a case for raising taxes, maybe, but not for an innovation program) or worse, something like “wouldn’t it be great if we were #1 in AgTech” (DARPA for Tractors). So, a moonshot isn’t a realistic option. In fact, if we could shoot the word moonshot into the Sun, that would probably improve innovation discourse in Canada enormously.
But what about “mission-oriented innovation strategies”? Well this is a phrase coined by policy entrepreneur Mariana Mazzucato, who is most famous for her book The Entrepreneurial State in which she re-discovered Kenneth Arrow’s ideas about the public role in subsidizing research and then proceeded to mangle the concept by claiming that the subsidization of basic research meant that the state was actually an entrepreneur (big hint here: if you have the power of taxation, you are not meaningfully at financial risk of…well, almost anything, and so cannot possibly be an entrepreneur). “Mission-oriented” innovation is her new shtick.
But the thing is that if you read what Mazzucato means by this term (and I think the best summary is probably this article), it’s pretty clear that what she is articulating is less an innovation process than a theory of how governments should be organized internally (fewer silos, more freedom to innovate) and how it should interact with the private sector (co-structuring markets and then getting out of the way. Read this document, which Mazzucato helped produce for the Inter-American Development Bank, where she enumerates a number of “mission-driven innovation policies” in Latin America, and where she singles out policies like reducing diabetes in Mexico, or improving education in the slums of Medellin.
There is nothing wrong in principle with what she is advocating, but it absolutely not innovation policy in the way we define this term in Canada. And more to the point it is the antithesis of a “moonshot” approach. One is about pouring money monomaniacally towards a particular technological objective while the other is largely about harnessing a wide variety of public and private actors to contribute to solutions on broader societal challenges. More broadly, one is about technology and the other is not.
It takes a deeply impoverished national discourse on innovation to think these two ideas belong in the same room, let alone the same sentence. And yet here we are. Welcome to Canada.
Look, innovation policy has some basic tennets. We have here in Canada the world’s greatest theoretician of innovation, namely the University of Toronto’s Dan Breznitz. He has written several books – most recently Innovation in Real Places: Strategies for Prosperity in an Unforgiving World– which lay out a few very simple principles for inclusive growth. First, expand flows of knowledge, demand and inputs between the local and global level. Second, increase the supply of public and semi-public goods which fuel innovation (mainly: high-quality education/training facilities and community facilities), and third, build local ecosystems that reinforce firm-level benefits of the first two fundamentals.
That’s it, that’s all. It’s not hard to understand. The question is why we don’t simply do it and stick to it instead of screwing around with ideas like DARPA for Tractors.
Or to put it another way: policy entrepreneurs are going to policy entrepreneur, but why is anyone in Ottawa bothering to listen to them?