Interview with Dan Breznitz

Dan Breznitz (co-director of the Innovation Policy Lab at the University of Toronto, where he is also Munk Chair of Innovation Studies) took some time out of his busy schedule to talk with SASE about his work, especially his recent SER article “The politics of partial success: fostering innovation in innovation policy in an era of heightened public scrutiny“, co-authored by Darius Ornston (Munk School of Global Affairs and Public Policy at the University of Toronto).


What motivated you to become interested in innovation policies as a field of inquiry?

The path toward innovation policies has been quite ironic. Before the advent of the Israeli High-Tech boom, I opened a software company with two of my childhood friends in the early 1990s. There was no VC in Israel at the time, so we started with a tailor-made software business to fund our R&D. We found it very difficult in the beginning, since the reaction in most first meetings with clients was: “why should we trust three young people in their early 20s? You are too young to know how to develop software, you need more worldly experience!” After a few years the going got better and we had a cash positive business; however, I decided after that this was not my calling and that I wanted to become a political theorist. I went to MIT and very quickly discovered that modern analytical political philosophy was also not my calling. Depressed, I went to Israel for the summer vacation thinking that this whole PhD thing was a terrible mistake. As it turns out, Israel had completely transformed – not only was starting a software company right after your military services no longer a rarity, it was common! – even high school students could knowledgably debate with me about the merits of term sheets offers from VCs, who suddenly seemed to be everywhere. It was an exciting change, and it seemed as if the whole Israeli economy and society was rapidly changing for the better. Suddenly, comparative socio-political economy became my calling. I started to look at the historical socio-political origins of these changes, only to find that all paths lead to the state. I then went to Ireland and felt firsthand that countrywide transformation that was known as the Celtic Tiger, with their indigenous software industry being the very first sector in which Irish companies reached global success. By then, the depression about my decision to go to MIT was gone, and I turned from a participant in the “innovation revolution” into a scholar of the same. I still wish I could pose as a philosopher, from time to time, though.


You argue that innovation has gone from a low-profile, marginal policy domain to a high-profile political priority.  Moreover, you claim there are many pitfalls due to these developments.  Why is this? And what does it mean to talk of “partial success” in such circumstances?

Our argument – since this research stream is a collaborative effort with Darius Ornston to such a degree that neither of us can any longer recognize who wrote what sentence – is that, inherently, innovation policy necessitates the three things that politicians and policymakers find the most difficult to do. To successfully promote a rapid-innovation-based industry, first, requires the development of radically new policies to target non-traditional actors and activities. Second, there is a need to experiment and demonstrate the feasibility of these policies, testing and validating models that then can be scaled-up. Finally, and following directly from the importance of experimentation, there is a need to constantly scale-up what works, stop what does not, and come up with new sets of policies as the industry and technology change both domestically and globally.  Now, if you think about the three most difficult things for politicians and policymakers to do, they are: i) developing and implementing radical new policies; ii) closing down programs; and iii) experimentation. For extra difficulty add on top of this the fact that it needs to be a continuous process! It is an almost mission impossible to be successful over time in transformative innovation policy.

Think about an elected official – a Minister, for example – first declaring (which she must do in order to garner support) that the radical new policy she just devised is of such importance and urgency that it necessitates a new budget and new programs since it will deliver economic miracles – but then having to shut down that program a few months later? Maybe, if she plays it right she can survive one such cycle (perhaps by spinning it to show that she is honest), but no politician can survive doing this on a regular basis.

This means that truly transformative innovation policy can only be done successfully at the periphery of the state, where you have enough obscurity to be truly experimental, and where limits on both resources and connections with mainstream industrial partners force you to be experimental with your resources and to look for new partners and ideas (since mainstream policies are already taken by other more “important” agencies). This is exactly what happened until the mid-1990s in Israel and Finland, where two peripheral agencies acting as what we call ideal-type Schumpeterian Development Agencies (SDA) with both the mission and just enough resources to experiment (but not enough to build white elephants or support zombies) developed successive waves of policies that ended up completely transforming their respective economies. However, with success came significant limits to their ability to innovate in innovation policy. With their success, innovation became “the” economy, and thus drew intense political attention. As such, both the Office of the Chief Scientist in Israel and SITRA in Finland quickly found their ability to experiment curtailed, and the ability of the whole economy to continuously develop successful innovation policies – that is, to innovate in innovation policy – sharply declined.  

This leads us to the current predicament. Innovation is not the esoteric and marginal politics/policy domain it was before the mid-1990s; it is now one of the most political and media-worthy domains. Consequently, the question we ask is: Can you really succeed in innovation policy under those conditions? Our answer is “partly yes”. By partly, we mean that the chances of having an agency that can truly act as an SDA (like SITRA and OCS) is minuscule, and so the ability to have truly transformative innovation polices is significantly reduced. Instead, we argue that contemporary reform-oriented politicians and policymakers must grapple with a trade-off between implementation and experimentation, which we term the “politics of partial success.” In a politicized environment, innovation agencies with the mandate and resources to administer new policies, however modest, are more likely to act as “effective upgraders,” implementing non-radical policies or applying new ones to existing industries, than as SDAs.


What do you find to be the biggest changes when considering innovation policies? 

The three most significant changes are, first, the way we now globally produce goods and services – that is, the current system of globally fragmented production. This system has drastically changed how societies/economies can successfully engage in rapid-innovation-based growth. Indeed, my first two books came out of trying to figure out this system, and arguing that, unlike what too many scholars maintain, globalization significantly increases the choices available to emerging economies. However, these choices are real – they have consequences in terms of what these societies look like, specifically in terms of economic distribution and inequality. The second big change, which we do not yet fully understand, is how changes in the rules and uses of intangibles – especially the creation and creative legal utilization of property rights – has changed who profits on what activities by employing whom, and who reaps sustainable profits from innovation. This has, in turn, vastly amplified various aspects of the financialization of the economy and made international trade treaties into arenas for IPE fights over what societies should look like. This goes all the way from IPR to fights about data and privacy, and we have a serious lack of understanding concerning these changes. We have practically no good theories or even conceptualizations of what these fights really are about nor how specific political decisions would translate to what outcomes in particular economies in the mid- and long-term. As a matter of fact, it is not clear we even have a good concept of work and labor anymore.  And let me note here that all our theories of property stem from an argument about labor. Last, but certainly not least, the whole domain of innovation has become a heated political arena, from domestic politics and issues about growth and inequality to new global power competition. As Darius and I have stated, these are the worst conditions in which to develop effective innovation policy or to think about where your innovation policies should lead you (in terms of what you want your society to look like).


What findings in this five-country comparison surprised you the most? 

For me, personally, the surprise was how easy it is to completely disregard the best policy ideas, even when they come from an agency that is specifically built to be strategic and is stuffed with the “smartest persons in the room”. Our five-country comparison made it very clear that if an innovation agency does not have the resources to at least seed its ideas, it quickly becomes politically irrelevant. This stands even for officially higher-ranked agencies. It was especially surprising that this turned out to be the case in Ireland (Forfás), where the realization that innovation was already a proven promise to change the Irish economy for the better led to a very thoughtful reorganization of the developmental agencies in the early 1990s, which was specifically aimed to allow for radical and transformative innovation policies. The story of Ireland’s innovation policy leading to the bust and stagnation of the indigenous high-tech industry is the sad surprise of this comparison.


What have you found to have the highest impact on innovation and industrial renewal?

Institutionalization and patience were clearly important in all stories of success (partial or full) –not the amounts or even the inherent brilliancy of the programs, but the fact that those agencies and their program were there year after year, constantly pushing for more innovation. This is an important reminder to everyone that innovation is a long-term game, a lesson we tend to forget. Just think for a moment what is it that you are trying to achieve with innovation policy: behavioral change of the highest order. Accordingly, innovation is a generational, not a three-year, game. Even if we talk about specific programs, it takes 2-3 years before the clients (firms and individuals) learn about the program and co-learn (together with the agency) how to make efficient use of it. If you talk about a completely new area of knowledge (either global or local) and you need people to start acquiring academic research degrees in order to just start playing with it, then you are talking about 5-7 years minimum. Last but not least, innovation is a high-risk and high-uncertainty game – if private market partners see that agencies are changed every five years, or that all programs are as permanent and reliable as the next election cycle, they view them as increasing risk and uncertainty, and would avoid them at all costs. In short, you need an agency filled with capable people and equipped with a modest budget that would devote itself to innovation over a long period of time, not the showy gushing of a massive budget with the hope of creating an innovation tsunami overnight. In more technical terms, initialization with modest but secure resources is always superior to temporary massive resource outpouring.