Innovation Leadership Forum

Meeting the challenges: Developing an Innovation Action Agenda

Alan Hughes
Centre for Business Research, Judge Business School, University of Cambridge, United Kingdom

PP: 359

Abstract

Innovation is a major source of competitiveness for firms and nations around the world. The challenge is not just to be innovative, but to lead innovation. To develop strategic advantage from innovation, Australian firms need to be ahead of the game when it comes to developing new products and services, using new organisational arrangements and incentives, and building new business relationships.

The summit provided an opportunity to bring together multiple perspectives on these critical questions. Its objectives were to build a community of interested parties to share knowledge on what we know and don't know about innovation in Australia. Its intention was to produce a clear statement on the significance of, and responses needed to, the challenges of developing an innovative Australia.

The Melbourne summit was preceded by invitation-only working forums in Brisbane and Sydney. The Brisbane forum in October looked at policy frameworks for innovation, while the November Sydney roundtable examined a variety of corporate perspectives on approaches to innovation.

Overall, about 250 key people from governments, industry associations, individual firms, research agencies and universities contributed to the dialogue.

From the active dialogue at this and preceding summits in Brisbane and Melbourne a working statement has been drawn together summarising:

1.       Areas of broad consensus

2.       The challenges Australia faces

3.       Issues we need to address

4.       Leadership priorities

5.       Open issues and matters for further attention.

The proceedings of the Melbourne Forum were structured as follows:

• Morning keynote presentations and discussion as thought starters and to set the scene
• Working lunch table discussions about possible leadership priorities
• Early afternoon panel sessions to crystallise views on the leadership challenges; and
• Closing sessions on setting a forward looking national agenda.

This record of the proceedings mirrors the programme for the day.

Going into 2007 the Innovation Leadership Forum will continue with a further series of roundtable meetings to address the open issues and the matters we have identified for further attention. We hope that this process will further broaden the emerging coalition of people committed to promoting an innovative Australia.

We thank the organisations and individuals who have generously sponsored and supported this series of forums, Brendan Lewis who organised the Melbourne event, and the staff from the University of Queensland Business School. It has been a truly collaborative exercise.

The Melbourne Manifesto - 06 December 2006

The summary of proceedings:

Article Text

Optimal Innovation Systems: Lessons from the UK and the USA

Professor Alan Hughes, Margaret Thatcher Professor of Enterprise Studies, Cambridge University

What I want to do is give a very high level overview of some areas of innovation policy and, in particular, the experience of the US. First of all I need to emphasise that while I'm speaking here as an individual the work I'm going to report is actually based on collaborative work with a number of colleagues and individuals who have been looking at US and UK innovation policy and performance over the last decade. In particular I'll make use of a survey which was carried out in the course of this collaborative research in the UK and the US covering about 3,500 firms and I'll present some evidence on that later on.

In thinking about innovation policy I want to begin with a little bit of anthropology. This illustration is not the English test cricket team's aeroplane, although it might be appropriate if it were so, it's an aeroplane constructed as part of one of the Melanesian cargo cults. Cargo cults have a long history in the Melanesian islands but they became particularly prevalent in the period after the Second World War. The essence of the cargo cult was that, having observed large quantities of cargo arriving in the Islands on American planes and British planes, and having observed the huge differences in living standards between the indigenous islanders and that implied by the arrival of cargo on these planes, a set of cults developed around the notion that if you simply mimicked or reproduced the objects that were associated with the arrival of the cargo then you could get the cargo delivered. The illustration is an example, a very finely constructed raffia aeroplane, but the cult went to much greater lengths than that. There were raffia air runways, carved head sets, control towers and so on. Of course the cargo never arrived.

Now why is this of relevance to innovation policy? I think that there is an element of cargo cultism in the way we are exhorted today to think about the key elements of science and innovation policy. A rather imperfect view of what has been happening in the United States is used to shape the main elements of science and innovation policy in many other countries. So what I want to do is look at what the elements of the so-called American model are, try and unpick those elements, and then put them back together in terms of a broader view of innovation policy based on the way in which innovation actually occurs in modern societies. So that's basically what I'm going to do. I'll rely on the particular pieces of work which I've carried out with colleagues at MIT in the United States but also draw on a rather wider set of analyses carried out by the OECD and others to illustrate my argument.

If you look around the world today every country is developing science and innovation policies, and you will find a number of key elements highlighted in their policy statements. The first is the tremendous emphasis on R&D. In fact, Australia is very unusual. Australia sits alongside the United States in a select band of two countries that have not announced a R&D target, much to Australia's credit I believe. But everybody else, the EU as a whole, every other OECD country, has set a target for raising R&D.

The second thing is the emphasis, to a very high degree, on the significance of exploiting the science base through the particular mechanisms of spin off companies and licensing. In particular, there are moves to constantly try and link what universities are doing to funding for commercialisation related activities, and you'll find that highlighted again and again in nearly every country's policy.

The third plank to all of this is that a key part of the way in which this science will be transferred into commercial activity is deemed to be through the spin off of new entrepreneurial firms. There is a tremendous emphasis on spin outs from the science base, encouraging entrepreneurialism, and solving what's often seen as the stigma of failure and, compared to the United States, less willingness to take a chance. This focus produces a wide range of policy, some of which is very expensive indeed in terms of tax breaks, which is designed to improve this kind of spin off activity in a narrow range of industries, namely high tech producing industries. If you look in the UK, Australia, or Canada you find a very similar range of industries constantly being emphasised: nano-technology, bio-technology, and ICT. In the UK, we spend about several billion pounds a year in tax offsets such as lower capital gains taxes to promote risk taking in these sectors through the promotion and sponsorship of venture capital activities, while the investors who fund the venture capital funds are also given tax breaks if they put money in this direction. So again a key part of innovation policy throughout the world is designed to produce a venture capital market thought to be like the private sector venture capital market of the United States.

So these are the key elements of the common model of science and innovation policy, and you will find them replicated everywhere you look.

The popularity of this model is strongly linked to the view that this is the basis for the United States' superior economic performance. There is no doubt the US has recently been an extremely successful economy, both in terms of its performance relative to other economies and in terms of its absolute performance in the recent past. However in looking at what the United States has done I would argue there is a kind of cargo cult in which a certain number of elements are picked out and emphasised. This thinking lies behind the prevailing policy model I've described. The cargo cult arises from the perceived success of a 'US' model of superior productivity and GDP growth based on:

• High-tech producing sectors as key drivers (ICT, biotechnology etc)
• New firm entry, especially spin-offs and new firm commercialisation from the science base
• Universities as key drivers of growth and productivity
• Private sector venture capital funding for new firm start ups and spin-offs based on greater private sector willingness to take risks

Now I want to argue that this is actually a partial and misleading picture. The first thing is to actually understand what has happened to economic performance in the United States.

Figure 1: US Productivity Growth

The most dramatic thing to note about the United States' performance is how its recent improvement is heavily concentrated at the end of the last century and at the beginning of this century. Secondly, we need to note that this is a return to their long run trend performance. There is not a tremendously dramatic change in their overall rate of growth. What was odd was that period after the oil price shock. Nevertheless, we're talking about quite a substantial recent change - almost a doubling in their rate of productivity growth over the period from 1995 - so understanding what happened is very important in trying to think through what the implications might be. One way you can analyse this change in performance is to ask which sectors actually led the turnaround in productivity growth. If you break down the turnaround in productivity growth in the United States, break it up into the bits which each industry accounted for, it depends on two things. An industry's contribution depends on its own change in productivity growth, and on its size, because the economy is a weighted average of all these different sectors.

If you look at the first period from 1995 to 2000 the key sectors in the United States economy on this basis were:

• Wholesaling
• Retailing
• Security and commodity broking

These accounted for all of the acceleration in productivity growth. None of these are technology intensive sectors in any conventional sense. If you look at the second period, the most recent years for which data is available, seven sectors accounted for 85% of all the productivity growth:

• Retailing
• Finance and insurance
• Computer and electronic products (one of these high tech producing sectors at last emerges in this list)
• Wholesaling
• Administrative and support services
• Real estate
• Miscellaneous professional and scientific services

None of these, with the exception of computers and electronics, are in any sense conventionally R&D intensive which leads to my punch line: it's a Wal-Mart not a Microsoft led turnaround . And it's not the traditionally identified R&D intensive sectors which have carried most weight.

I do want to make clear what I'm not saying. It's clear that Sam Walton, the founder of Wal-Mart from Arkansas, matters more than Bill Gates in the sense that retailing 'accounts' for a bigger share of productivity turnaround than 'hi-tech' producing sectors such as IT. It's also clearly true, however, that both of these individuals and their organizations are highly innovative. But one is in a sector which no one thinks of as being in the vanguard of a science and innovation led policy and yet Wal-Mart has transformed - many people would argue, much for the worse - a whole variety of social structures in the United States and delivered enormous productivity in the retailing sector. The other is a high tech producer which supplies the capacity for many of these changes to occur in the 'low-tech 'sector. So in that sense Sam Walton isn't more important than Bill Gates and they're certainly both innovators and are involved in complementary activities. Sam Walton and Wal-Mart are more important to productivity turnaround than Bill Gates and Microsoft because of the scale of the activity which is transformed by the activities of a company such as Wal-Mart when it implements IT linked business transformations. Difference in services productivity growth accounts for most of the difference in national productivity performance between the USA the UK and Europe in the past decade.

So high tech activity is a small part of the economy if you think of it as a producing activity but enormous gains are to be had by the use of that technology elsewhere. Therefore you have to think extremely carefully about the mechanisms by which high technology activity is diffused through the rest of the economy. As Geoffrey Blainey explained so clearly, you wouldn't identify mining and agriculture on any standard list of high technology industries and yet the transformation of those sectors over long periods of time by the use of all kinds of innovations and scientific discoveries elsewhere has been fundamental to their performance, so it's use that matters and not just production of technology.

Now I want to turn to the issue of new spin offs and their role in this system; I call this the golden oldies versus the new kids on the block debate. The new kids on the block are all these new high tech spin off firms that are attributed such an important role in the science and innovation process. Now I want to contrast some facts about spin offs, especially in the United States, with what we know about the way in which the golden oldies dominate many changes to industry structures and productivity growth.

The first thing is to get a sense of proportion. The United States economy has some 500,000 firms starting up each year. That's firms of all kinds from small restaurants to boutique high tech businesses, not just IP businesses. If you look at the US as a whole, in 2004 there were 462 IP based start-ups where the IP is from a US university. That's actually an extremely impressive performance so I'm not saying that's not a significant activity, but look at the scale of it when compared to claims of what might be gained in other economies from such spin offs. When I was reading the Australian Productivity Commission's recent report on the contribution of public science, I came across a reference to a recent study of five extremely successful Australian IP based technology start-ups. After describing how successful these were the cited study argued that if Australia could grow 200 to 250 more research based companies like these over a five year period the result would be around $20 billion in annual export earnings and Australia would be well on the way to reducing the national debt. Well since the whole of the United States system produces 462 spin offs and the vast majority will not be outstanding successes it's important to get a sense of perspective and recognise the extreme optimism and massive expansion in university spin off activity which that statement actually implies.

Secondly, although IP produced by US universities produces significant numbers of start-ups and it produces a lot of patenting and licensing, it's insignificant numerically compared to the total amount of licensing of research activity in the United States. IBM in the year 2005 alone registered 2,941 patents with the US Patent office, Canon 1829, and HP 1790. The whole of the University of California state system, which is one of the most dynamic, productive and innovative university systems in the world, produced 388, MIT 136 and Stanford 90. It's a tremendous university performance, but again it's important to keep it in perspective and think of universities as a part of a wider system in which they are perhaps a quantitatively small but very important part.

Finally, and it bears both on the issue of start-ups and the numbers of them and licensing, the returns from this activity are enormously skewed. There are a couple of statistics here which illustrate just how skewed. Only 167 out of 27,322 patents held by 193 US university institutions in 2004 made over $1m. The vast majority of US university technology licensing offices barely break even or don't make a profit, so the chances of hitting a home run are small. That doesn't mean to say you shouldn't do it; in fact you can't win the race unless you place a bet, but you have to be realistic about what the odds are.

We can now look at this in a slightly broader way. Instead of just looking at the spin off activity by US based universities we can look at the whole structure of start-ups in economies in the OECD. There's some very interesting work been done which breaks down the change in productivity in particular industries across the whole of the OECD . This work includes Australia, United States and the UK and enables you to break down the productivity growth between the gains in productivity that are made by the firms who are there throughout the period you study. They are what I call the golden oldies, they're there at the beginning and they're there at the end. Then there's the impact on productivity of firms that leave. Of course, that helps. If bad firms leave, if the worst firms drop out, there's a kind of batting average affect and the average productivity rises. And there's the effect of those new entries, the new kids on the block, the spin offs and new start ups. They enter the system and grow over the period that you look at. Well what is clear from this work is that the vast majority of the productivity growth that is experienced in any economy and an industry in any time period is driven by the transformation in productivity of the golden oldies; that is, it's the improvement in the performance of the firms that are there all the time. So if you focus entirely on driving innovation and productivity by looking at what you do with new firms you miss a very important part of the story.

For instance in UK manufacturing net entry, that's the difference between the new firms coming in and the firms going out, only accounted for 16% of the manufacturing productivity growth. The rest was all to do with the transformation of the existing firms and shifts in market share between them.

Now there are some industries and there are some conditions which are relatively favourable to the success of new entry. A couple of factors affect the chances of getting somewhat better returns from new entry. The first is where the nature of the technology is constantly changing the basis on which you can successfully compete - so there's turbulence in the technological basis of the industry - and where entry is relatively low cost. In those circumstances you can get a lot of experimentation in new entry and some success. Also, if the incumbents - the golden oldies - in the industry are very heavily committed to an existing technology then there's a better chance of a new entity succeeding because the conservatism that goes with very heavy investment in a standard technology makes the incumbents relatively slow to react. Finally, you have a good chance of success if the things you require to exploit your new business - complementary assets - are not owned by others. If these complementary assets, which are necessary to extract value, are owned by somebody else you are unlikely to capture the value by going it alone, but if that's not the case then new entry can matter. What this means is, first, it's not new entry as such that matters but it's how entry plays out within the whole system and what the survivors are doing. Secondly, very few of the entrant firms will succeed. Only 30 to 50% of new entrants survive for five years. Finally you have to look very carefully at what the industry circumstances are before you can attribute a likely successful role for high tech start-ups.

Figure 2: Use of sources of knowledge for innovation (% of companies)

Now I'm going to say something about the roles of universities and I'm going to rely on the survey that I mentioned at the beginning. The first thing that we did was to find out the extent to which firms used universities as a source of knowledge for innovation. In other words, if one asks businesses - and our sample covered a wide range of sizes of firms in the US and the UK - what were their sources of knowledge for innovation what you find is that universities are not cited as frequently as many other elements of the innovation system. This is shown in Figure 2. So customers, suppliers, competitors, and the firm's own internal knowledge are the dominant sources. There is a distributed innovation knowledge system and universities are a small part of it. In both the USA and the UK universities are relatively low in frequency of use as direct sources of knowledge for innovation and the same is true in Australia. This does not mean they are not important but what it does mean is that they are a small part of a complex system.

The same is true for Australia. Figure 3 provides some data that I calculated from the Australian Innovation Survey, and the same picture emerges.

Figure 3: Key sources of ideas or information for innovation in Australian innovating business 2001-2003 (% companies)

What I want to emphasise is that the role of the university is a small part of a complex system, but in that small part what does it actually do? Is it all about spin offs and licensing or is there something else going on? Well what do universities do? The first thing they do, and I always find this important to keep reminding myself and audiences, their first role is actually educating people and they produce skilled graduates. This role tends to get lost in a lot of the innovation and science policy debate. The second thing, and this is deeply connected with research, universities increase what might be called codified knowledge. University staff publish things, they write things, they patent, and in many engineering faculties they develop proto-type machines. Then there are a very wide range of problem solving activities that are often carried out - often in a very detailed regional and local way, but sometimes on an international basis - actually directly addressing problems which are brought to the attention of the universities through contract research, co-operative research, and all kinds of faculty consulting. Often university laboratories will have equipment which can be used for testing out various kinds of commercial outputs.

What tends to be not emphasised as much as it should is what, following Richard Lester and Michael Piore, I call here the public space function of universities. Universities have a very distinctive role to play in society and in the innovation system as public spaces in which other interested parties can play, if that public space is correctly structured. This is a whole range of soft things, but none-the-less extremely important, to do with network forming, stimulating social interaction, influencing the direction of research processes by identifying commonly experienced problems, setting standards of a technical kind, setting up entrepreneurial centres and so on. So there are a whole range of public space functions. The role of the university, therefore, is very diverse.

Figure 4: The university role is multi-faceted

We asked our firms how they interacted with the universities and what kind of emphasis they placed on interactions. Figure 5 shows that they interact across the full spectrum of those elements which I set out in the previous exhibit. The most frequently cited form of interaction is informal contacts, and it's not only the most frequently cited - it's also given great significance as an important factor. All the conventional modes of university output are also highly ranked: business values the undergraduates and graduates, the publications and the conferences. In that sense there is no real conflict between what the business community says it wants from universities and what academics themselves want to do.

Figure 5: Types of university industry interaction contributing to innovation (% companies)

It is worth noting that US firms appear to use and value internships more than their UK counterparts. Licensing and patenting - all those heavily emphasised activities in innovation policy - are amongst the least frequently cited, especially in the US. In some industries they're important, but in the generality they're low down.

Finally I want to turn to the issue of private sector venture capital for the risk taking community in the United States, and the view that what's required is huge subsidies to promote the private venture capital industries in countries that don't have the vibrant system that United States has. My point here is that if you look at the United States, one of the the most powerful, most proactive venture capital organisations is a public sector organisation. It's the Small Business Innovation Research (SBIR) program .

SBIR was established in the middle of that period of very low United States economic growth. The United States in the period 1972 to 1990 experimented with an impressive range of industrial policy mechanisms to counteract what it perceived, correctly, as its failure to deal with the threat commercially of Germany and Japan. One of the things they did was introduce this program. They also introduced a wide range of very big, collaborative programs which I won't have time to talk about today, but this particular one was established to support businesses with less than 500 employees, and provides 100% funded contracts to carry out particular contractual obligations to US Federal Agencies. The US Federal Agencies advertise technical or research related problems they need solving, there's an open competition and then a contract is awarded to do them. So it's not a grant it's actually a contract. The US government currently mandates 2 ½% of Federal Agency total R&D spend to SBIR, and that is absolutely an enormous sum, it's 2 billion dollars annually, and they make 4,000 awards and some extremely big and successful companies have been party to this program. Amgen, Qualcomm, and Genzyme, for instance, all have SBIR connections in their origins. These sums are roughly equal to the total angel finance and two to three times the early stage venture capital funding, because the private venture capital in the US is pretty much like it is in the rest of the world - it's not particularly risk loving and mainly goes in for all kinds of large scale company buyouts and so on. It's only a small proportion of funding which goes into early stage finance. So actually what the SBIR does is produce a situation in which a lot of the very early stage investments have this huge public sector mandated activity already providing certification and proof of performance capacity in these small firms that win these contracts.

So, what are the overall lessons? Well, the first is that US productivity and growth performance is not based on high tech production per se; it's based on the diffusion throughout the system and the transformation of what people would regard as low tech sectors by general purpose high technology.

Secondly, productivity gains are in general driven by what you do with the firms that are in existence, so thinking about the existing firms and their innovation performance is critically important in the innovation process, not just looking at start ups.

Thirdly, public sector procurement has potentially a very powerful part to play.

Fourth, universities have to be seen as part of a complex system - there's no one single or simple role they play and it's a very complex system in which their direct contribution is relatively small compared to other components in the system. The mechanisms for that interaction with business are very diverse and you have to foster the most productive in particular circumstances. Licensing and spin offs are one part of the story and they produce very uneven returns, and a 'one size fits all' economic development or innovation strategy for any country or any university is not appropriate. In particular all universities are not the same. The introduction into countries of systems of evaluation, and Australia is about to embark on an RAE like activity for its universities, is potentially very dangerous unless it's done very carefully because it can force every university into the same mould. All universities do not have the same talents, capacities or objectives, and all sectors and fields of science are not the same and they don't require the same kinds of interaction.

Finally, the returns to university based spin offs and licensing are highly skewed and that means there is no silver bullet. A strategy based on the notion that universities are going to fund themselves in a substantial way from spin off gains and licensing streams is a very risky one. Lots of things that academics normally do are valued by businesses. If you make universities just like businesses they lose their interest and attraction to academics and businesses alike. Far better to develop informal and people based public space interactions and, coming back to the point about evaluation, encourage a diversity of mission across universities; don't encourage mission overload on every university.

... continues ...

View references

Navigate: Archives | Contents | Top of page

Search

Translate

Related Publications from eContent Management