R-2022-104

The Silicon Valley model of entrepreneurship has become closely associated with innovation, but this dominant view overlooks the role played by collaboration among diverse stakeholders, including the public sector. To promote social well-being, notes economist Sébastien Lechevalier, governments need to play a more active role in directing innovative efforts.

*     *     *

The argument I would like to develop in this review is that the private sector plays a key role in innovation but not an exclusive one. This role should also be fulfilled by other actors, including the government and civil society.

I will first describe what can be called a dominant view of innovation and then offer a criticism of this dominant view.

What I call the dominant view has been deeply influenced by what happened in the United States, especially on the West Coast, during the 1980s and 1990s. This is related to the success of the so-called Silicon Valley model, especially in information and communication technologies and also in biotech. This Silicon Valley model has been simplified, and there are several theories to explain the success of US companies in the field of innovation during these two decades. These models emphasize the importance of entrepreneurship, venture capital, and also collaboration between firms and universities.

This model gives great emphasis to the activities of private companies, particularly a specific type called startups. These companies have been at the center of attention of policymakers all over the world during the last three decades, with moderate success in some countries outside the United States, such as Israel.

According to this dominant view regarding the process of innovation, France and Japan are considered laggards, being unable to build successful innovation systems because of backward institutions. The main reasons given for this are the lack of entrepreneurship and the low rate of startup creation. The two countries—Japan and France—are not innovative, according to this definition of the Silicon Valley model.

I would add that this view of innovation is very much inspired by what one might call a Neo-Schumpeterian view of innovation, which, while closely associated with the idea of creative destruction, also incorporates several other hypotheses. Examples include the idea that all social problems have technological solutions, that private actors are the drivers of innovation, and that the role of the state should be limited to the definition of intellectual property rights and, subsequently, providing infrastructure, including education.

This view, which has been influenced by the Silicon Valley model, may not be completely misleading, but I believe it is not as general as it pretends to be. I would venture to say that a closer look at social innovations will reveal that we may need to change our definition of what innovation is.

Four Limitations of the Dominant View

I will now discuss what I believe are four limitations of this dominant view.

The first limitation is that this dominant view of innovation underestimates the importance of collaboration between diverse players. Research on innovation during the last two or three decades has emphasized the reality that collaboration is really at the center of the very process of innovation. For one, I can cite a famous paper by John Hagedoorn describing the rise of R&D partnerships. I can also cite a second famous paper, by Fred Block and Mathew Keller of UC Davis, looking at the 100 top innovations every year from the early 1970s and showing that the share of collaborative innovation is dominant and is increasing over time, especially from the late 1980s. Of such collaboration, public support is dominant. We should thus consider the importance of collaboration and the role of the public sector in this process.

The second criticism is about increasing science linkages. The Silicon Valley model, as I mentioned, does emphasize the importance of collaboration between universities and the private sector. But I would argue that it should be really central. One indicator of collaboration is the global increase in the number of citations of scientific papers by patents, although the trend is not equal in all countries and, of course, all sectors. But it is very important in the United States, and we can conclude that increasing science linkages is crucial. Even if it appears less important in Japan, figures nonetheless show that the number of patent application by universities in Japan has grown dramatically from the early 2000s.

The third point is about the role of government in industrial and innovation policy. Here, I would like to emphasize a paradox. The European Union defined more than one decade ago a new Innovation Union strategy with a key role for not only the governments of member states but also the European Union—the European Commission—itself. This strategy emphasized the role of all types of innovation, including social innovation and innovation in the public sector. The EU has also created a European Innovation Council whose goal goes well beyond the promotion of startups. The paradox here is that this EU strategy is very much influenced by Japan’s past experience.

It seems that in Japan, there is a tendency to forget what was originally the strength of the Japanese innovation system. But it has been analyzed and imitated in Europe, while the Japanese government appears to have forgotten and given up on the importance of government coordination. The famous example is the importance of research and development consortia that were, in my view, a key element of innovation process in Japan. It has been imitated in Europe, and I can bet that it will become more and more important in the future.

I would make two points about the Japanese model of innovation. First, what is clear is that Japanese companies, especially since the 1990s, have engaged in a technological development race, and this is why they continue to dominate, especially the large companies in Japan. But this effort to catch up with the United States and with the so-called new economy has accounted for so much of R&D expenditures that spending for human resource management has been a race to the bottom, as economic constraints have forced them to massively cut costs. The tendency has been to cut all kinds of expenditures for human resource management, including those related to training and improving human resources. I think it is quite striking that when there is an increase in R&D and technology expenditures, there is a simultaneous decrease in spending on human resource management and training.

I would posit that the so-called Silicon Valley model might indeed be effective in some industries and for some technologies but that this is not true for all technologies. The role of startups and entrepreneurship might be very important in some sectors. But as I showed in papers that were published almost 10 years ago, focusing on the topic of next-generation robot technologies—which are the basis of personal robots—major players in these industries are still very large companies. Large companies have achieved success through collaboration with various parties, since robotics requires a lot of collaboration. I call this model not the model of entrepreneurship but the model of intrapreneurship. There is still space in some industries for innovation within large firms. I do not deny the importance of startups, but large companies can also play a role.

Innovation does not emerge out of thin air, and it is deeply conditioned by a set of institutions. That is why when we look for policies to promote innovation, we have to look very cautiously at the relationship between science and society. This will offer a more balanced view of the capabilities of the Japanese and French innovation systems.

Finally, as the fourth point, I would like to emphasize the importance of society. This part of my argument is taken from a book that was published a few years ago. I think that at one point we all became Schumpeterians, and the problem with this, as I see it, may be the narrow, Neo-Schumpeterian thinking regarding innovation. This view does not emphasize enough the role of controversies and contestation in the innovation process.

Benefits for Society

A very interesting article was published in Nature in 2017 emphasizing the fact that there is an increasing gap between expenditures for research and development and benefits for society. My opinion is that governments, in the name of the public good, should have a greater say in how the budgets for research and development are spent and that this decision should not be entirely left to the private sector. I would like also to point out that technology is part of society. It is very important to integrate technology into society and to reconcile people with the innovation process. And for this, it is necessary to rethink what is innovation.

My conviction is that the crisis of innovation is a crisis of the relationship between technological innovation and society and that it is very important for them to be reconnected. In other words, society needs to play a much more active role in the process of innovation, and I would point out that this requires that we change our vision of innovation. Currently, this vision is dominated by the needs of private companies, which see innovation as a critical source of differentiation and comparative advantage.

This may be valid from a corporate perspective, but from a social viewpoint, innovation should be mainly a source of well-being. What we therefore need to do is to define the conditions that lead to social well-being and to set institutions to identify what contribute to achieving those conditions and what do not. I would emphasize that the social sciences and humanities can be of great help in identifying those criteria.

As a matter of conclusion, I would advocate an active role for all stakeholders in advancing innovation: private companies, universities, public research institutes, government, and, last but not least, society in all its diversity.

Subscribe to the Tokyo Foundation for Policy Research eNewsletter.