One swallow does not a summer make

A recent blog by Geoff Mulgan argued that “public intellectuals” had a duty to engage in public debate: responding to and learning from criticisms about their work. He jumps to the conclusion that I am not engaging in such debate because I didn’t respond to a blog written by his Nesta colleague, Stian Westlake.

Let’s start with our common ground. I agree “public intellectuals” should engage in debate with both those who agree with them and those who do not, and gain from this process. Indeed, as he points out, this is a key way in which academics develop their thinking - through peer-reviewed journals, presenting and discussing their research at conferences and, where possible, the opportunity to discuss and refine their work through engagement with policy-makers, business leaders and other practitioners relevant to their field.

But I was surprised to read Geoff’s piece because it is exactly this kind of discussion and debate with collaborators, practitioners and commentators that I have been engaged with in the past few years since my book (the subject of the blog), The Entrepreneurial State: debunking public vs public sector myths, was first published. This has included engaging with specific academic criticisms in the economics and innovation community, regarding whether innovation policy is simply about fixing market failures and system failures, or making and shaping markets and systems. It is indeed those precise criticisms that have motivated me, in writings over the last two years, to be much more specific on exactly why the traditional market failure framework for economic policy is limited (more on this below). And I’ve also responded to criticisms of my work, both ‘live’ in presentations to audiences resisting my key points on the need to fundamentally rethink (not cut back) the role of the public sector, but also in responses to blogs, such as that by Tim Worstall, from the conservative Adam Smith Institute. Here is my long response to his criticisms which he posted near his Forbes blog.  

I was also surprised to learn that what prompted Stian to write his blog was a supposed declaration on my part of not being interested in looking at the nitty gritty aspects of innovation policy (what-works-what-doesn’t), and preferring “diagnosis to prescription”. I have no recollection of saying that and, indeed, for me to have said that I was not interested in the granular world of policy making (specific tools, organizations, etc.) would have been very strange indeed, given that one of my key interests is precisely the organizational dimension (what kinds of public organizations can welcome the explorative nature of innovation) and the evaluation dimension (of both failures and successes, and how to replace cost-benefit analysis with more dynamic measures). But let me use this blog space for what is more important, which is to address the substantive concerns with my published work.

Geoff is right that criticism should be welcomed, and so it’s a shame that in this particular instance I did not write a specific reply to Stian, instead using a number of articles (both academic and more popular) to address questions which had come from a range of sources. But I’m happy to take this opportunity –as good as any– to respond to Stian’s critiques more directly. These are essentially three:

1. That the role of government in innovation is nothing new to the innovation studies community;

2. That I don’t consider the difference between research and development, and in particular to the private sector’s role (which makes me underestimate Apple’s contribution to the iPhone);

3. That my proposals about making sure the government gets returns on its R&D investments are flawed and that I disdain taxes, and consumer surplus, as the way to get such return.

I will discuss each one in turn.

What is new in my argument about the role of the State in innovation?

Although the innovation community does see governments as doing R&D and funding innovators, few acknowledge the degree of risk-taking of the public sector as investor of first resort, and the way in which this role is better analysed as market creating and shaping, not (just) fixing market or system failures.  

The usual point, repeated again in Geoff’s blog, that the government does a lot of basic research while the private sector does development, is not the point at all. It’s the breadth and depth of public investments across the entire innovation chain and the types of risks (and inevitable failures) which this involves, that requires better theorising. These include investments not only in basic research (the classic ‘public good’ from market failure theory), but also in applied research, early stage high risk seed financing for firms, and demand side policies via procurement policies and policies providing new directions for innovation. It’s not enough to recognise that these investments exist, the real question is what they imply for the way we have been considering public sector involvement in the economy—and hence what new questions they raise.

As the usual criticism of market failure theory from the innovation community is to allude to system failures, I’ve spent quite a bit of time focusing on why the failure-fixing approach (whether markets or systems) is limited and why it should be complemented by an understanding of policy as market shaping and creating. This has involved bringing together the path-breaking work of literatures that do not speak to each other: (1) the work of Karl Polanyi and the developmental state; (2) the work of Keynes on the role of government investing where the private sector won’t, and (3) the work on mission-oriented innovation by Schumpeterian scholars like Dick Nelson and David Mowery. Indeed, in various articles since 2013 I’ve reflected on the benefits that might be gained by bringing these different conceptualizations together, such as in this SPRU working paper, a shorter version of which is forthcoming with the title "From Market Fixing to Market-Creating: A new framework for innovation policy" in the peer reviewed academic journal Industry and Innovation.

Precisely because the alternative framework has broader implications than just on classic innovation agencies, more recently, with Caetano Penna, I’ve been working on various papers on how a market shaping view can help us better understand the emerging role of public development banks in innovation, beyond their usual remit. With Gregor Semieniuk I have been using the Bloomberg New Energy Finance (BNEF) database to investigate international public and private investments in green innovation, to study the way one type of investor (public) might crowd in the other. With Carlota Perez I have looked at how governments have historically also established demand side policies that tilt the playing field for converging and synergistic innovations. Indeed, funding research or even funding innovative entrepreneurs seriously understates the multiple roles of the State in innovation. These points about ‘mission-oriented’ approaches and market shaping/creating strategies are missing almost completely in the innovation policy debate and require new thinking.

Serious academic thinkers in the mainstream economics community often tell me they believe the market failure framework works well – and these discussions have been valuable in developing my thinking on why the market failure framework is limited, and what a new framework might look like. Regardless of whether I’m right or wrong, working on the market creating and shaping role of public policy is not ‘mainstream’ in either the innovation community or in the economics community.

How much of the iPhone came from government?

Both Stian and Geoff repeat that I don’t consider the difference between research and development, meaning that the latter is what really makes the difference and that’s what the private sector does mainly on its own. Consequently, they claim I’m wrong not to give the private sector more credit for innovation. In a recent testimony to a forum in the US Senate, I focussed on precisely this distinction, and the feedback effects between research and development (the latter is done by both public and private actors) which are often missed in policies that assume ‘linear’ innovation models.

In any case, the main point Stian and Geoff object to here is the idea that “the State created the iPhone”. Yet the title of the Chapter (5) they are referring to in my book is not “The state created the iPhone” but rather “The State behind the iPhone”. The point of that chapter is to reveal, in detailed form, the public funds that invested in those technologies which enabled the iPhone to be a smart phone (Internet, GPS, touch screen display, SIRI), and how the fact that we ignore that side of the story, is what then creates a problematic eco-system. It allows some private companies, like Apple, to capture too much (relative to what they put in) of the cumulative rewards from innovation.

Indeed, I acknowledge in various places in my book (and indeed in all my presentations on the topic) that Steve Jobs was no doubt an inspiring genius, as is his evolving team in Apple which do ‘development’ brilliantly. The point is that Jobs’ ingenuity and that of his team is often presented as the whole story, ignoring the role of the public sector institutions that funded the development of technologies critical to all the “i” products—beyond the basic science (and even provided Apple with one of its first early stage investments). What Apple did was put those technologies together in innovative new products, but ignoring who funded those technologies is a major problem. Indeed, in the 600 page biography on Jobs by Isaacson—a great book by the way—there is not one sentence about this side of his success.  

So, yes Apple produces great products (who said they don’t?) and no, the government isn’t solely responsible for the invention of the iPod, iPhone or the iPad. Neither is Apple. But only one of those actors is routinely described as innovative and risk-taking while the other one, if it’s acknowledged at all, is seen at best as facilitating and de-risking. The point of my work has been to show how the bias, in both popular narratives and economic theory, has had negative implications for misinformed innovation policy, and for the problematic distribution of the rewards from innovation (more on this below).    

Similarly, Elon Musk may be a genius, but ignoring the high risk funding that his company Tesla has received, is not helpful. Indeed, a recent LA Times article showed that the support to Tesla, SolarCity and Space X (the three companies forming his empires) amounts close to $5 billion, much of this in public subsidies. In my book I discussed one of the investment sides of this support, in the Tesla S car: a $465 million guaranteed loan by the Department of Energy. It went well. But a similar amount went to Solyndra, and it went very badly—causing public outcry of government as incapable of picking winners. Yet any venture capitalist would tell you that this is normal: for every success you will have many failures. And that is precisely why governments might benefit from considering investments in a portfolio setting, not only covering the downside but also benefitting more directly from the upside. Which leads to the third point Stian raises.

Why should the government get a specific return from innovation beyond taxes?

Stian writes that “several of the recommendations in The Entrepreneurial State’s Appendix make good sense” but goes on to raise some questions about whether we really need to look at alternative ways of funding innovation and the ways in which this is done in other countries.

So what is my risk-reward point exactly? By not having a proper framework through which to understand the government as a lead risk taker and investor (not just a spender, de-risker, and market fixer) we have ended up socialising the risks and not the rewards. And we end up building what I call—a dysfunctional, parasitic ‘ecosystem’. I then try to set out some new ways in which we can think about adequate public sector rewards from innovation.

As I explained in detail in a Foreign Affairs article on The Innovative State, which appeared shortly after Stian’s blog, economists tend to assume that rewards to innovation will happen through taxation and/or through spillovers. Indeed, Stian repeats these assumptions. But while upstream basic research (a public good) may produce economy wide spillovers, it is less true for downstream investments where particular technologies and firms receive the benefits of the investments (a grant, a loan, or publicly funded applied research that gets patented). Furthermore, as patenting has increasingly moved upstream (see the excellent work of Nelson and Mazzoleni on this), even the spillovers from basic research are not guaranteed. And while consumer surplus, which Stian mentions, may capture the benefits in terms of prices (the difference between what consumers are willing to pay and what they actually pay), it’s widely recognised that this static approach is not great for capturing the dynamic benefits of innovation.

Secondly, even in the case of basic research, there is a real question whether tax today is working to bring back a public return. As William Lazonick and I argue in an article on risks and rewards (in the journal Industrial and Corporate Change), it is precisely the wrong narrative about who the innovators are that has allowed tax systems to be lobbied in problematic ways. It’s not a coincidence that it was the National Venture Capital Association in the US that lobbied for capital gains tax to be cut in half at the end of the 70s. And it is that same problematic story that is behind the constant pressure to lower upper marginal tax rates (in the US the top rate was more than 90% when NASA was founded) and corporate income tax. Stian claims tax is working to bring back a fair reward, we (and many others) believe it is not— even when it is paid.

Furthermore, it is precisely the false narrative of the state only being important to fix market failures, rather than to be a lead investor, that has made government too shy to push for a better “deal” with business. In the book I mention various ways this could happen: through equity, royalties, shares, and retention of a golden share of the intellectual property rights (IPR) etc. Another could be through the price system itself. Indeed, it’s remarkable that while the Bayh-Dole Act (1980), which allowed publicly funded research to be patented, allows government to affect prices of those products (mainly medicines) which are publicly funded, the government has never exercised this right.

Geoff also recalls Stian’s questions on the different options that could be considered for potential “clawback” (a term I don’t like or use—much preferring the simple concept of public sector sharing in the upside). So what are Stian’s criticisms? In passing, I mention the Finnish innovation agency Sitra and the equity it retained in Nokia to exemplify ways in which equity might be retained by the public sector. Stian reminds us that Nokia is much older than Sitra. This is, of course, true, but Nokia’s recent revival is commonly viewed as a new stage in its history, and indeed Sitra did benefit from the shares in this company, and did use its profits to invest in similarly innovative Finnish companies (both Geoff and I are on Sitra’s Advisory Board). It cannot be denied that Sitra's role was critical in pioneering the Finnish VC market in the early 1990s when it trained some of the first VC managers and made early stage VC investments before the private VC market was adequately developed (an ‘entrepreneurial state’). Another public fund in Finland, Finnish Industry Investment, is an active investor in the country, investing in both mature industrial companies and in high growth companies, with some great success stories. Although Sitra is barely mentioned in my book, the point of the chapter where it is found is to consider different ways in which public funds have acted as lead investors, and the importance of studying the cases to gather insights (pros and cons) on ways the state can socialize not just downside risks but also some of the upside rewards…so as to depend less on public budgets precisely when these are being cut, and also being affected negatively by falling tax rates.

Stian also raises a question on my comment on the actual and potential role of public banks, world wide, in fostering innovation and providing the type of revolving fund implied above. In referring specifically to the Brazilian public bank BNDES Stian makes two points: that BNDES’ high return on equity is due to its access to cheap funding; and that return on equity is not an indicator of its effectiveness as an investor in innovation. 

The first critique on the issue of public banks refers to where the funds for public banks come from. This differs greatly between public banks. In the case of BNDES it has evolved from coming from workers’ social security contributions (FAT) to funds coming directly from the treasury. In the case of the European Investment Bank it has been capitalised from national funds. But in all the cases the real question, as described below, is whether a public bank then is able to invest strategically and earn a return from the funds, regardless of where it comes from. An excellent PhD thesis at John Hopkins University also looks into the degree to which BNDES profits have been related to investments in innovation, showing that BNDES’ high profits are not derived from lending margins (which Stian’s blog implies), instead are explained by a low cost structure, access to more credit-worthy clients, low default rates and from its equity arm, BNDESpar.

Second, is whether the rate of return of banks like BNDES, is derived only from innovation investments, and of course it’s not. It’s coming from the entire portfolio of investments. The real question should be whether the overall return (the part that remains in the bank) is then reinvested across a broad portfolio that includes high-risk innovation, and not just to give support to national champions. Indeed, 80% of the return actually goes back to the Treasury that can choose to invest it in areas like health and education. I have addressed this issue of returns in various papers on state investment banks with Caetano Penna, both generally in 2014 across different banks, and in one particular one in 2015 where we look at specific innovation funds within public banks (for BNDES we study the Criatec fund and the BNDESpar fund mentioned above).

Stating, as Stian does, that the returns from public banks are simply from favourable treatments they receive, is not only wrong but also misses the key questions being asked today by those looking into the new economics of development banks—including their problems.

Lastly, Stian mentions Yozma, an Israeli venture capital fund. He claims it’s also not a good example of clawback since the royalties it earns are only up to the point of the initial investment. While it is true that it does not make a profit above the original funds (and I have discussed this with the Israeli Chief Scientist), the key point here is that the investment is gaining a direct return (with no loss to the tax payer) so it can enable future investments and the benefit to the State moves beyond an indirect taxation route, regardless of whether the investment is successful. This is very different from the kind of support US companies get. Take the case of Tesla, where the US government only stood to gain an equity stake in Tesla if it failed to repay its loan. That’s right – the government would only get shares in Tesla if Tesla was failing, not if it was succeeding! My Foreign Affairs article describes this further.

Besides these direct mechanisms, Lazonick and I argue that it would be fair to ask companies that receive public support in innovation to reinvest a high proportion of their profits back into innovation. Instead what we have is record level hoarding and also share-buybacks, with the Fortune 500 companies having spent $4 trillion on share buybacks between 2004-2013. Lazonick has shown that between 2003 and 2012, publicly listed companies included in the S&P 500 index used 54 per cent of their earnings (around $2.4 trillion in total) to buy back their own shares. One of our most recent journal articles on this topic (in Accounting Forum) applies this analysis to Apple, which under Tim Cook is engaging in major buyback schemes not seen under Steve Jobs. In various sectors, there is increasing evidence that this lack of reinvestment has been at the expense of innovation and investment in the real economy (and has been discussed by others thinking about the problems of financialization, such as Adair Turner, John Kay and Andy Haldane).   

So what I argue in the book, and in my more recent writings, is that to create a more symbiotic—less parasitic—innovation ecosystem we need new thinking to realign risks and rewards.  My intention is not to limit the specific ways in which part of the upside returns can come back directly to the public, but instead to open up the discussion, and to provide a framework that allows the question about risks and rewards to be seriously considered and various ideas for its resolution to be explored. I’m currently supervising a second year PhD student who is looking at the concrete ways in which this has been done across the world, comparing and contrasting the pros and cons of different experiences in varied settings.

Well known economists worldwide have recently reflected on these points on risks and rewards, from Dani Rodrik who advocates a similar type of ‘revolving fund’ (in a superb Project Syndicate article called From Welfare State to the Innovation State) and Kemal Dervis from the Brookings Institute who repeats many of my same points on this issue.  

In my recent work with Randy Wray we have also responded to those who – inspired by Minsky – have criticized the emphasis on government needing a return, given that in theory it can simply ‘create money’ (this is the ‘endogenous money’ angle, of which Wray is a key proponent). But this is a whole other story which I will not go into here except to mention that it’s one of the many criticisms I have responded to.  

Are the diagnoses and the policies wrong or right?

Of course the proof of the pudding is in the eating. Whether policy prescriptions are ‘right’ or not cannot be proven with theoretical arguments but by testing them in practice, once they are conceptually justified. The theory that says taxes are the way to make sure public returns is received for public investment falls flat in a world of massive tax avoidance, and dysfunctional tax dynamics. And Stian’s concern that handling returns to the state in the form of dividends, royalties or golden share of the IPR poses administrative challenges, indeed a “nightmare” he says, seems a strangely old fashioned worry in the midst of a big data revolution.

I am fortunate to have had a lot of interest in my work and to have had many fruitful discussions and interactions with academic colleagues, policy-makers and others from around the world. They have engaged with the ideas I have put forward and improved and developed them through their observations, critiques, research and new ideas. My interest continues to be in how to encourage growth that is smart and inclusive. To make progress towards this goal will be the work of many people. Sometimes in vibrant disagreement to improve ideas. But always with a larger sense of a common purpose.   


Mariana Mazzucato

Professor Mariana Mazzucato (PhD) holds the RM Phillips Chair in the Economics of Innovation at SPRU in the University of Sussex. She has held academic positions at the University of D…