Science and Innovation Policy

Cristina Chaminade; Bengt-Åke Lundvall

doi:10.1093/acrefore/9780190224851.013.179

Summary

This is an advance summary of a forthcoming article in the Oxford Research Encyclopedia of Business and Management. Please check back later for the full article.

Scientific advance and innovation are major sources of economic growth and are crucial for making social and environmental development sustainable. A critical question is if private enterprises invest sufficiently in research and development and, if not, to what degree and how governments should engage in the support of science and innovation. While neoclassical economists point to market failure as the main rationale for innovation policy, evolutionary economists point to the role of government in building stronger innovation systems and creating wider opportunities for innovation.

Research shows that the transmission mechanisms between scientific advance and innovation are complex and indirect. There are other equally important sources of innovation, including experience-based learning. Innovation is increasingly seen as a systemic process where the feedback from users needs to be taken into account when designing public policy.

Science and innovation policy may aim at accelerating knowledge production along well-established trajectories or at giving new direction to the production and use of knowledge. It may be focused exclusively on economic growth, or it may give attention to the impact on social inclusion and the natural environment. An emerging topic is the extent to which national perspectives continue to be relevant in a globalizing learning economy facing multiple global complex challenges, including the issue of global warming. Scholars point to a movement toward transformative innovation policy and global knowledge sharing as a response to current challenges.

The Origins of Science, Technology, and Innovation Policies

Science and technology policy, defined as government intervention in the economy to support scientific discoveries and the development of technological solutions, can be traced back in history (Lundvall et al., 2005). Science, technology, and innovation (STI) policies have shaped the world as it is today. Already in medieval Europe, the kings and princes were engaged in competitive support of science and art. This competition was deemed to be crucial for the relative success of Europe as compared to, for instance, China, where the centralized control slowed down scientific progress. The mode of competition fostered a culture, still with us today, of open science, where the individual scientist has positive incentives to share knowledge and contribute to the common growing foundation of scientific information (David, 2005).

In 1541 King George VIII established the Royal Iron Works in Weald, Sussex, under the leadership of William Levett, with the clear mission to develop the technology to produce iron cannons. The success of the enterprise was based upon immigration of blacksmiths from continental Europe, and it was crucial for the political and military dominance of England. But, more importantly, it started the process where England moved from being a low-technology to a high-technology country. It was an early example of national technology policy that shaped the modern world and a critical episode in triggering the industrial revolution and in leading to England’s dominant economic position in the 19th century (Yakushiji, 1986).

The Manhattan project, when the U.S. government brought together world-leading physicists and scientists from other disciplines to develop the atomic bomb, served to demonstrate the potential of state support to science and technology, and it was logical that Vannevar Bush, who initiated the Manhattan project, should become author to Science: The Endless Frontier (Bush, 1945). The report outlined a future for science and technology policy in the United States for the postwar period that started with investments in research related to health and the military and culminated in the establishment of the National Science Foundation in 1950 with the mandate of supporting basic science.

These examples illustrate that more or less articulated science and technology policies have been in existence for a long time but also that the state may support science and technology for different reasons. Most governments engage in support of basic research resulting in open science with allowance for national scientists to share the knowledge they produce with state support with scientists from other countries. But substantial shares of government resources are allocated to research and development (R&D) used for nation-specific purposes and with restrictions to the open flow of knowledge.

While science and technology policy address the political missions of the state and especially the issue about military strength or scientific supremacy, the more recent emphasis on “innovation policy” reflects growing recognition that knowledge and innovation are fundamental for national economic performance (Lundvall et al., 2005). But again, this perspective has historical precedents. In the 19th century, economists in countries that were lagging behind the world leader, the United Kingdom, in terms of productivity pointed to the need to build a stronger knowledge base in order to catch up. This was certainly true for Friedrich List in Germany, who argued that intellectual capital was more important than financial or physical capital (List, 1841/1904). On this basis, he argued that economic policy, including trade policy, should focus on interventions that strengthened the national knowledge base. This was his main argument for protecting infant industries and for rejecting the free trade ideas as developed by Adam Smith (Freeman, 1991).

In the next section, the main differences between STI policy are discussed and the following questions addressed: What are the rationales for policy intervention? When, how, and where should policymakers intervene? To answer these questions there is a need to understand the dynamics of the modern economy, including the workings of the innovation process and the different theoretical framings that offer answers to the questions, from the neoclassical and evolutionary economics to the innovation system perspective.

There are several challenges ahead in the research and practice of STI policy. When what appears as rational action seen from the perspective of the single nation-state undermines the long-term viability of the world system economically, socially, and environmentally, contributing to rather than responding to global challenges, there is an urgent need for public policies that give innovation new directions and takes into account growing global interdependence. More specifically, it is important to discuss how the innovation system perspective can help policymakers to develop a transformational innovation policy that aims at global economic, social, and environmental sustainability.

What Is Science, Technology, and Innovation Policy?

Although it is common to use the term STI policy as one type of policy, one could think of it as three different “ideal” types of policies—science policy, technology policy, and innovation policy—each with distinct characteristics (Lundvall & Borrás, 2005). Science policy is about the promotion of the production of scientific knowledge and, as such, deals with the allocation of resources between different scientific activities. Science policy might serve different objectives from pure curiosity about understanding the world to specific military objectives such as the atomic bomb (Jaffe et al., 2015). Science policy is sometimes based upon a linear model where it is assumed that research efforts will translate more or less automatically into economic and societal outcomes.

Technology policy, in contrast, focuses on promoting the development and use of specific technologies seen as being of strategic importance for the country. Technology policy is thus based upon the assumption that there are specific strategic technologies that have a major impact upon the economy and on societal objectives, and it focuses on the development and improvement of those technologies or their wider adoption. These technologies can be specific to a particular economic activity or more generic, like information and communication technologies also referred to as “general purpose technologies.”

In a sense, innovation policy incorporates science and technology policy as it aims at intervening in the innovation process as a whole—from science (exploration) to the application to specific technologies, its introduction to the market, and its wide diffusion (exploitation). Innovation policy gives attention not only to the scientific and technological content of innovations but also to the institutional framework and the wider changes that are necessary for innovations to be introduced into the market and used. It also pays attention to other forms of learning beyond science and technology that might also lead to innovations like learning by doing, using, or interacting.

Policy can then be understood as the deliberate action from governments in the economy with the aim of attaining objectives through stimulating changes in the behavior of individuals and organizations. When it comes to STI policies, the aim is to affect the rate and direction of processes within STI. When designing and applying policies, governments need to have a theory of how innovation works in general and empirical evidence of how it works in a particular country, region, or sector. Evolutionary and neoclassical economics represent two alternative theoretical framings with different implications for what type of policies governments should pursue.

Figure 1 summarizes the main differences between these three types of ideal policies.

Open in new tab

Figure 1. The main differences between science policy, technology policy, and innovation policy.

Source: Author’s elaboration based on Lundvall and Borrás (2005).

What Drives STI Policy? Different Looking Glasses

The Neoclassical Perspective: Market Failure and the Focus on R&D

In the late 1950s, U.S. economists such as Kenneth Arrow (1962) and Richard R. Nelson (1959) supplied arguments for why governments should take on the responsibility for investments in scientific research. They pointed out that scientific information is a “public good,” meaning that it could be used by others without losing its value and that it is impossible to exclude others from using it. The implication is that there is little incentive for the private enterprise to invest in science and there is an important role for government to compensate for what can be referred to as “market failure.”

Neoclassical economics assumes that all transactions take place in markets where rational agents make choices to maximize their own individual utility. Under certain specified conditions, neoclassical modeling shows that the uncoordinated decisions of the multitude of agents result in “general equilibrium,” where resources are allocated optimally—under these ideal conditions any intervention by the state would make someone worse off. Most applied economic analyses operate without explicit reference to this foundational model, but they are nonetheless rooted in the model. This is why policymakers trained in this tradition require evidence of “market failure” before engaging in public policy.

The argument that private actors cannot fully appropriate the benefits of knowledge investments and thus will invest less than what is socially desirable is generally accepted as a base for public policies aiming at increasing the investment in knowledge. Societal rates of return (their contribution to national economic growth) on public investments in knowledge have been shown to be high in spite of the fact that they often have aimed at noneconomic objectives (Hall et al., 2010). Such theoretical and empirical contributions have made the investment in R&D a central target for innovation policy under the neoclassical paradigm.¹

Seen from a neoclassical perspective, innovation is a linear process rooted in R&D and resulting in technological progress and technical innovations. However, the relationship between investments in science and innovation and national economic performance is intricate and uncertain. Not all research aims at innovation, and innovations draw upon other kinds of knowledge than scientific information (Kline & Rosenberg, 1986). New ideas need to be transformed into new processes and products, and it is only when the new technologies become diffused that they have an impact on the economy as whole. Innovation policy may aim at any of the different stages invention, innovation, and diffusion.

The Evolutionary and Interactive Model of Innovation and the System of Innovation and Systemic Failures

The most basic distinction between neoclassical and evolutionary economics is that, while neoclassical economists compare equilibrium situations, evolutionary economists study processes of change. Key concepts in evolutionary economics are diversity, selection, and reproduction. Innovation, where new resources are created, is seen as more important than allocation of resources with given characteristics. Individuals and organizations are seen as diverse in terms of motivations and capabilities, and, most importantly, it is assumed that they can learn and through processes of learning they may change not only in terms of what they know and can do but in terms of their priorities, which may change on the basis of experiences. The evolution of knowledge is at the very focus of evolutionary economics. In this universe, there is no reference to general equilibrium state and no certainty about outcomes, and the market is seen as only one out of several alternative forms of transaction (Dosi et al., 1994; Nelson et al., 1982).

Another important difference is the assumption that innovation is the result of a complex, uncertain process emanating both from science-based and experience-based interactive learning (Lundvall, 2008). Two points are particularly important here: (a) the understanding that science is not the only source of innovation and (b) innovation is the result of interactive learning.

Diversity of Sources of Innovation

As early as in the writings of Adam Smith there was recognition that, while some innovations draw upon science, others emanate from processes of learning in the production system, reflecting the creativity of manual workers. Innovation scholars have on the basis of historical and empirical studies developed a taxonomy of learning:

•

Learning by scientific discovery

•

Learning by doing

•

Learning by using

•

Learning by interacting

The notion of learning and knowledge production as a result of basic research and science overlaps with the traditional understanding of innovation as the outcome of an act of discovery and has been the main focus of neoclassical STI policies based on market failures.

The insight that experience-based learning matters as much as scientific progress flagged by innovation scholars under the evolutionary and systemic model raises issues about what kind of knowledge matters most for innovation. As mentioned, there is a tendency among mainstream economists to translate knowledge into “information” and to assume that knowledge can be easily communicated in coded messages through computers and electronic media. But the different forms of learning result in skills and competences that are tacit rather than explicit and embodied rather than “embrained” (Lam et al., 2006). Both individuals and firms know much more than they can specify, and their skills/capabilities cannot be easily copied by others.

While scientific results are increasingly used as the basis for developing new products and processes not only in high-technology fields (such as pharmaceuticals, aviation, electronics) but also in “low-tech” areas (clothing, furniture, and food), the successful development, diffusion, and use of innovations depend on tacit knowledge and on learning by doing, learning by using, and learning by interacting.

The concept of learning by doing was introduced by Kenneth Arrow (1962) in the context of the production of complex systemic products (airplanes). His data showed how productivity was growing continuously over time as the organization gained experience from finishing a new version of the product. Actually, the examples given by Smith of workers developing innovations may be referred to as outcomes of learning by doing.

Later, learning by using was introduced by Nathan Rosenberg (1982) in the context of the use of complex systemic products (airplanes). When a user acquires the new product, the costs in terms of operation and maintenance will be high but with time the costs are reduced as the user organization becomes gradually more proficient in mastering the new technology. The concept has a much wider application. Frequently users of new products will require a learning period, and this is true both for consumers and business organizations,

Almost at the same time, Lundvall (1985) introduced learning by interacting in connection with product innovation. His analysis showed how the producer of a new product depends on feedback from the users and how users require assistance from producers when taking new products into use. The fact that most innovation processes reflect such feedback and network relationships is at the core of modern innovation theory and it stands in contrast to the original linear model where innovation was seen as springing directly out of scientific progress. This concept is important to understand how STI policies are justified, designed, and implemented under the evolutionary and systemic framework.

Innovation as an Interactive Process

Innovation may be defined as a process combining existing elements of knowledge in new ways and with new knowledge as outcome. Different organizations and individuals with distinct elements of knowledge interact in the innovation process. Interactive learning takes place within organizations as well as between organizations.

Within organizations barriers between different functions and divisions (e.g., marketing, R&D, production) may block the kind of interaction that is necessary to innovate. One reason for the increasing use of management techniques based on knowledge management principles, such as job rotation and interdivisional teams, is that it results in functional flexibility supporting innovations rooted in interactive learning. Organizational characteristics at the firm level will have a major impact on both the capacity to develop innovation and the capacity to successfully absorb technology developed by other organizations (Lam, 2004).

However important internal knowledge flows are for innovation, one very general result from innovation surveys is that firms “do not innovate alone” but in continuous interaction with suppliers, customers, knowledge institutions, and sometimes even competitors (De Bresson et al., 1991; Rothwell et al., 1974; Von Hippel, 1988). Most of the collaboration lasts for longer periods and is informal, but there are also many examples of contractual agreements where companies enter strategic alliances to develop technologies together (Mowery et al., 1996).

There are two fundamental reasons for engaging frequently in collaboration for innovation. One is the need to reduce uncertainty and to overcome information problems while the other reflects the need to share knowledge. Regarding the first, innovation is per definition an uncertain process—if the outcome of the innovation process was known in advance, it would not deserve to be called an innovation. On behalf of the producer it involves uncertainty regarding future sales of the new product; on behalf of the user it involves technological uncertainty about the actual performance of the new product/process. With a pure market, implying arm’s-length relationships and with access only to information on quantities and prices, the uncertainty on both sides would make innovation spurious (Lundvall, 1985).

The other reason for collaboration is that the growing complexity of the knowledge required to develop innovations makes it impossible for even the largest companies to rely exclusively on internal expertise. The management concept of “open innovation” refers to this fact. While neoclassical economists emphasize intellectual property rights and pure markets, business behavior often involves knowledge sharing either via formal contracts or through informal barter of information.

Given the importance of different sources of knowledge as well as of interactive learning in processes of innovation, policies designed with inspiration from evolutionary and systemic theory focus primarily on supporting capacity building, network formation, and developing support frameworks conducive to interactive learning and innovation. Government intervention in STI is seen not as a response to “market failures” but as response to “systemic failures.”² Among those systemic failures it is common to refer to infrastructure failures, capability failures, network failures, and hard and soft institutional failures linked to formal rules as well as more informal ones (e.g., culture) among others (Chaminade & Edquist, 2010; Smith, 2000; Woolthuis et al., 2005). Overall, STI policies addressing systemic failures differ from STI policies addressing market failures, although it may be argued that they are complementary rather than substitute and necessary at different points in time (Weber et al., 2012).

Mission-Oriented, Transformative Policies, and the Proactive State

Both the neoclassical and the system of innovation justification for STI policies have been criticized for being rather reactive, addressing market or systemic failures rather than taking a more proactive, entrepreneurial, and leading role (Leyden et al., 2015).

On the one hand, it is argued that governments should be able to take risks and invest in those areas where the private sector will not due to their short-sighted operating framework (Mazzucato, 2015). The benefits from advanced science and technological developments cannot be maximized without the state taking the steering position. In fact, Mazzucato (2011) provides many examples of breakthrough radical technologies that have been the result of publicly funded research, from the Internet to nanotechnology.

On the other hand, some authors argue that policies addressing systemic failures tend to produce only incremental changes while the grand development challenges, ranging from tackling climate change to eradicate poverty, require radical system transformations (Elzen et al., 2004; Foxon, 2015; Schot et al., 2016). In this context, some authors argue that the role of the state is not only to create the right infrastructure, to set the rules supporting the macro economy, or to provide funding for specific technologies but to articulate new visions about societal desirable goals, steering STI efforts according to those visions. Examples of these “steering visions” are the United Nation’s 2030 Agenda and the 17 sustainable development goals or the European Union’s Grand Societal Challenges framework.

Articulating mission-oriented or challenge-oriented STI policies involves addressing new types of challenges in relation to directionality, demand articulation, reflexibility, and coordination (Weber et al., 2012). Directionality refers to the need to articulate collective priorities and the direction of change. Demand articulation refers to the need to anticipate user needs and mobilize the demand in the direction of the challenge (Edler & Boon, 2018). Reflexibility refers to the ability of the systems’ agents to anticipate changes and mobilize actors. Finally, coordination refers to the need to manage policies in different realms (labor, education, industry, trade) to steer the system in the desired direction. It has also been argued that challenge-oriented STI policies require a change of perspective from the traditional top-down government intervention (also called the command and control) to new forms of governance and partnerships between public and private actors including the civil society (Borras et al., 2014).

Instruments for STI Policy

The instruments of STI policies are the techniques developed by governments to achieve the desired STI objectives, which could be, for example, increasing the percentage of the gross domestic product devoted to R&D, stimulating entrepreneurship in the economy, or strengthening the linkages between university and industry (Borrás et al., 2013; Edler et al., 2017). The final goal of any policy instrument is to steer the behavior of the different actors in the economy in a certain direction. But, for that, the policymaker needs to have a certain theory of how the different actors will react to specific interventions and how these interventions ultimately could contribute to achieving the desired goals. Policy instruments are thus strongly related to theoretical framings (Laranja et al., 2008; Martin, 2016).

Supporting R&D—The Neoclassical Perspective

Under the neoclassical paradigm, governments may intervene to expand total investments in R&D through three main instruments:

Designing and redesigning intellectual property laws.

Investing directly in research taking place within public universities and laboratories.

Subsidizing private investments in R&D directly or indirectly through tax deduction for R&D-investments.

Protecting Intellectual Property

The first known patent law, introduced 1474 in the city-state of Venice, granted inventors exclusive rights to their inventions. Since then there has been a lively debate on the legal regulation of intellectual property (Boldrin et al., 2002).

One reason for controversy is that most intellectual property rights regimes have a mixed impact on innovation and economic growth. On the one hand, protection offers an incentive to the innovating firm or inventive individual. On the other hand, the protection creates a barrier for competitors’ inventive activities and slows down the diffusion of new ideas. There are innovation scholars who actually think that the capitalist economy could be at least as innovative and dynamic without any kind of legal protection of intellectual property.

Other controversial issues relate to the extension and administration of such rights across national borders and their role in global trade agreements. It is well documented that the rich countries in their early development borrowed (or stole) ideas from each other, and imposing strict rules on poor countries’ use of technologies has been referred to as “kicking away the ladder” (Chang, 2010).

There are few serious proposals to abolish intellectual property rights in general—they have become institutions deeply rooted in the world capitalist system. But it remains controversial how far they should be extended when it comes to protect knowledge related to specific fields of knowledge. Examples are knowledge about the human body—the human genome is one specific issue (Jensen et al., 2005)—and software where, if established, they may limit the access to data as well as data processing tools crucial for the general advance of science (Boldrin et al., 2002).

Scholars in favor of wide use of intellectual property rights point to the fact that the disclosure of technical information obligatory in connection with patents might actually stimulate innovation not only directly but also indirectly. To obtain a patent, the patentee must reveal elements of new knowledge in the application. This might offer insight for competitors that would not be available if the inventor/innovator used secrecy to protect the new technology. It has also been argued that firms with well-established intellectual property rights may be less reluctant to engage in certain forms of “open innovation” (Chesbrough, 2003).

The general principle for designing patent systems is that they should not be imposed on elements of “generic knowledge” that constitute building blocks for science and for developing new technology. If there is a need to engage private investors in building such elements, an alternative that does not block the diffusion of new knowledge is for governments to offer prizes to the successful inventor.

Public and Private Investments in R&D

It is generally accepted that local, regional, and national governments operate organizations engaged in research and education. Statistics shows that in 2015 governments in Europe on average used around 0.7% of the gross national product (GNP) to finance R&D taking place in higher education institutions or public laboratories. The level of public investment in R&D was at the same level in other high-income countries such as the United States and Japan. But in these countries the investments made by private enterprises was markedly higher than in Europe (Europe, 1.3%; United States, 1.9%; Japan, 2.8%). The statistics also reflect the internationalization of R&D activities with foreign firms contributing to national research funding corresponding to 0.2% of GNP in Europe (more than 30% of total R&D in Bulgaria, Latvia, Slovakia, Lithuania; Eurostat, 2017).

Innovation studies show that public investment in R&D has positive effects on private sector research (Guellec et al., 2001). For instance the pharmaceutical industry in the United States draws upon research taking place at medical faculties at public universities and the military’s investment in research on Information and Communication Technologies (ICT) at universities has been crucial for the strong global position of U.S. firms in this field (Mowery, 1994). But in order to absorb and use public knowledge, it is crucial that there is a certain level of R&D efforts in private firms. Former developing countries that have succeeded in catching up with high-income countries have been characterized by a strong growth in business R&D.

One widely used instrument to promote business R&D is programs with tax deduction proportional to firms’ investment in R&D. The number of Organisation of Economic Co-operation and Development (OECD) members that make use of such programs doubled between 1995 and 2011. Almost half of all public support for business R&D comes through tax credits within the OECD. In France this support reaches 70% of total support, equaling 0.27% of GNP. The wide and increasing use reflects that this kind of policy is less demanding in terms of public policy competence and that it (at first sight) appears to be neutral, offering all enterprises equal access. However, there is an emerging understanding of serious problems with tax incentive programs (OECD, 2013):

They are not very efficient. A tax deduction proportional to the level of expenditure may have very limited impact on the size of R&D-investments. In this case it would be more efficient to give direct support to specific projects. If the size of the tax deduction is made dependent on the increase in the firms R&D expenditure, the efficiency would be higher but it may lead to strategies that do not support long-term performance of the firms.

The instrument is far from neutral and may work as a means to attract foreign firms through tax competition. Worldwide business R&D is extremely concentrated: 80% of all R&D takes place within 1,500 multinational firms, and those firms are in a situation where they can make use of tax credits to avoid paying taxes. In contrast, small and medium-sized firms operating mainly within national borders have very limited opportunities to benefit from tax deductions.

Supporting the Diffusion and Adoption of Technologies

Joseph A. Schumpeter’s (1939) classical distinction between invention, innovation, and diffusion has been used for decades to inspire innovation policy. “Invention” refers to the formation of a new idea while “innovation” takes place only when the idea is brought to the market.³

While all countries engage in attempts to increase the R&D effort, there are fewer examples of public policies that explicitly aim at affecting the diffusion of new technologies. Theories of diffusion have been developed with inspiration from the spread of infectious diseases, assuming that a more efficient technology will grow slowly to begin with but that as the number of adaptors grow, information about the new technology will reach more potential users and therefore the rate of growth in demand for the innovation will grow until it gets closer to a satiation level, where most potential users have adopted the new technology (Rogers, 1962/1983).

Innovation policy taking such a model as a starting point would primarily focus on adaptors’ access to information. More rapid and complete access to information would result in a more efficient rate of application, not necessarily in a speed-up of the diffusion since more information about new and more attractive versions still to come might lead potential adaptors to postpone acquisition (Stoneman et al., 1994).

Speeding up the diffusion of new technologies may not always result in an acceleration of economic growth. To promote productivity, innovation policies with a focus on competence building and organizational change may be as important as policies with a focus on the technology. To illustrate this, in the 1990s economists were puzzled by the fact that productivity did not grow in spite of the rapid diffusion of ICT technology, including automation equipment used in manufacturing. This is the so-called Solow paradox (the U.S. Noble Prize winner in economics Robert Solow stated that “we see computers everywhere but in the productivity statistics”). This paradox took a special and more dramatic form in the case of Denmark, where increased investment in information and communications technologies was combined with a fall in productivity (Lundvall, 2002; OECD, 1991). The end of the 1980s was characterized by growing investment in advanced technologies in manufacturing and a fall in productivity. A close analysis showed that the fall in productivity was concentrated in firms that invested in radically new technologies without investing in training and without changing their organization. As a consequence the Danish government developed, in collaboration with industry and as part of its innovation policy, a program with a focus on diffusion organizational good practices.

Historically the most prominent and successful examples of diffusion policies have been focused on the use of technologies in agriculture, and the first studies of diffusion processes referred to agriculture (Griliches, 1957). But in recent years a great deal of attention has been paid to network technologies (telecommunications, Internet, Facebook, etc.) where the value of the single artifact/service increases with its diffusion. This has given new importance to diffusion policies and specifically to the public investment in ICT infrastructure. Governments may engage in building information highways, resulting in a more rapid diffusion in the national economy. The development in China is outstanding in this respect. The public investments in ICT infrastructure has resulted in creating the most networked economy in the world.

But innovation policy aiming at affecting the rate of diffusion can intervene in other ways and sometimes with greater impact. For instance, public procurement plays a major role in promoting the diffusion of health-related technologies. Some technologies can only be widely spread if the public sector invests in infrastructure (roads for automobiles, fiber cables or satellites for computers and telecommunication equipment). The diffusion of process equipment will depend on the skill level of the national workforce, and the diffusion of consumer technologies will depend on the level of education of consumers.

The adoption of technologies rather than their development might be particularly strategic for small and developing countries. In Denmark, the government has stimulated the use of computers in schools and in public administration as well as among all ordinary citizens. One powerful instrument has been to change the mode of communication between citizens and public authorities. The only way for the citizen to communicate with central and local governments is through the Internet. This has been combined with training programs for older people. This kind of “forced user learning” may give Denmark an advantage when it comes to adapting new network technologies developed abroad. This is important since the well-being of small countries such as Sweden, Denmark, and the Netherlands depends more on adapting foreign technologies than on hosting the pioneer firms that develop them. This is also the case for developing countries.

Supporting Interactions and the System of Innovation

An important role of innovation policy is to support the formation and reconfiguration of interorganizational networks. Competition policy needs to take into account that long-term relationships between firms are crucial for the success of innovation. But the task of public policy is not just to promote linkages. One important insight from innovation studies is that sooner or later the historical pattern of network relationships becomes problematic, and breaking them up might require intervention from government. Classical examples are the transformation of the office machinery and the watch industry from mechanical to electronic industries.

These insights based on historical and empirical studies challenge some traditional simplistic ideas about the role of innovation policy. While it still is relevant for governments to take on responsibility for ensuring sufficient investments in science and to promote strategic technologies, there is a need to go beyond a focus on the supply side. Upgrading the capacity of users of innovations—individuals as well as firms—becomes as important as promoting the supply of scientific knowledge. Promoting learning organizations and offering workers autonomy may be as powerful as stimulating entrepreneurship and elite engineering. Furthermore, ensuring that there are appropriate institutional frameworks to support innovation becomes a fundamental task of STI policies.

One way to bring all these aspects into the analysis is to evoke the concept the innovation system (Freeman, 1982, 1987, 1988; Lundvall, 1985, 1988, 1992). This concept refers to the interaction among individuals and organizations in connection with innovation and learning processes. It is built on the assumption that the performance of the national economy reflects the nation-specific economic structure as well as nation-specific economic institutions. The concept makes it possible to develop a historical and holistic understanding of the role of STI policy. It takes into account neoclassical assumptions that markets, competition, and profit seeking behavior are important elements of the modern capitalist economy, but it also takes into account that those elements are mixed with social elements and that the national context and international competition matters for innovation and for innovation policy.

The innovation system concept has become widely used as framework for innovation policy all over the world both at the national and the regional level, and thus it is common to refer to the national innovation system or the regional innovation system when discussing STI policies. The micro foundation of the concept is an understanding of innovation as based on relationships and processes of interaction between users and producers of knowledge and innovations. Since the economic structure and the institutional set-up (institutions understood as shared norms and routines) differ across regions and countries, the preconditions for innovation are different. Therefore the innovation system concept points to the need for innovation policy strategies to be country or region specific.

The concept was originally developed to demonstrate that competitiveness based on policy strategies aiming at keeping wages low is not sustainable and that in the long term strong competitiveness requires a strong knowledge base as well as systemic interaction aiming at innovation.

STI Policy Instruments in Narrowly Defined Innovation Systems

The innovation system may be defined broadly or narrowly, and the STI policies that might derive from each approach are different. Narrowly defined, the national innovation system includes those organizations that form the core of the system, those having the production and use of knowledge as their main activities. That would include universities, technological institutes, and R&D departments within firms and governmental organizations directly involved in regulating knowledge production. If innovation policy is defined equally narrowly it would mainly encompass some of the elements already presented such as (a) support to private and public R&D, (b) diffusion policy including building technological service infrastructure, (c) building science and technology parks, (d) investing venture capital and other forms of entrepreneurship stimulation, and (e) developing the interaction between university and industry. But the system’s perspective would lead to a different approach in each of these policy areas. For example, the motivation for supporting private R&D would emphasize that in-house R&D is a prerequisite for absorbing knowledge and technologies developed by other firms. In connection with diffusion policy, more attention would be directed to the formation of linkages between producers and users of innovation. In relation to the interaction between university and industry, the focus would be on mutual benefits from the interaction and on the capacity of industrial enterprises to absorb scientific knowledge. Fostering interpreneurship (incentivizing agents who specialize in initiating new relationships between enterprises) and intrapreneurship would be seen as important as fostering entrepreneurship (Acs et al., 2017). Of special interest are policies aimed at supporting university–industry interactions.

Many initiatives are aimed at creating stronger links between universities and industry. The basic idea is that much research takes place at publicly funded universities and that the outcomes from that tax-payers’ funded research may be of indirect or direct use for innovation processes in private firms, eventually leading to economic growth (Gibbons, 1994). One problem with the sharing of knowledge between the two types of organizations is that they are driven by different types of incentives and characterized by different modes of behavior (D’este et al., 2016). While researchers are expected to create new knowledge and to publish their results as soon as possible, the private firm has an interest in keeping critical knowledge in house—at least until it has been protected as intellectual property.

In addition, many public policy initiatives strengthen the interaction between universities and the corporate world, including university-based science or technology parks. Some of them are national while others are initiated by regional authorities. Among the different initiatives, one that has been given special attention refers to a U.S. law called the Bayh-Dole Act, which was introduced in 1980. The basic idea was that the reason markets could not be established for research results was the absence of clear ownership. The Bayh-Dole Act specified that universities could patent important new ideas of use for industry. On this basis markets could be established and the knowledge could be sold to private firms (Mowery et al., 2001).

However, there is disagreement regarding to what degree this kind of institutional set-up actually has benefitted the innovation process at the national level. There was a growth in the frequency of patenting, especially at some of the largest and most research-oriented universities (e.g., Stanford University and Columbia University) in the wake of the law, but at closer inspection it turns out that other factors have played a more important role in explaining the growth. Actually, the upward trend had started before 1980, and much of the following increase in patenting reflected the expansion of research and patenting in one specific field of research: biotechnology.

Some experts in the private sector argue that the main impact has been to make the collaboration with universities more cumbersome and to make researchers more reluctant to share knowledge. The principle of openness within science has been crucial for the long-term progress in knowledge and technology. When the production of science becomes more involved in marketing, there is an obvious risk that this principle will be undermined.

STI Policy Instruments in Broadly Defined Innovation Systems

The broad definition of the innovation system includes all organizations rooted or located in the national economy engaged in innovation and in the use of new technologies, and it takes into account the role of learning among workers and consumers. This implies a wider agenda for innovation policy and a large variety of STI policy instruments (Edler et al., 2017; Flanagan et al., 2011; Georghiou et al., 2014; Martin, 2016), including (a) public procurement (Edler et al., 2007; Gee et al., 2013; Georghiou et al., 2014); (b) the support of the formation of formal and informal innovation networks (Laranja et al., 2008); (c) promoting learning organizations (Nielsen & Lundvall, 2003); (d) promoting consumer learning (Boon et al., 2018); (e) education policy; and (f) labor market policy (Lorenz et al., 2016).

This list shows that people and their networks are important for the performance of national innovation systems. Upgrading the knowledge and skills of workers and consumers contributes to innovation performance. But it is also important that education and training systems, as well as labor market experiences, shape the relationships and the human interaction in the national innovation systems. Elite education and large social distance will for instance reduce the interaction and slow down processes of innovation and diffusion. Rigid definitions of professions and high barriers between knowledge fields undermine functional flexibility within organizations.

Innovation processes will succeed when there is high degree of diversity among agents but where the agents in spite of differences easily can combine their specific skills with others with radically different skills. Education systems, work-life experiences, and labor market dynamics will all contribute to shape people. There are significant differences between national systems in these respects, and to a significant degree they reflect differences in labor market, education, and social policies. For instance, the introduction of modern participatory education principles may foster new generations of workers who are more apt to contribute to innovation.

Directions for Future Research

STI Policies in the Era of Globalization

One reason for the increasing focus on innovation policy is that the room for traditional economic policy as well as industrial policy has become narrowed by the increasing openness of the economy and by world trade regulations. But the national innovation system is also quite open, and this openness has important implications for national innovation policy. No single country is completely self-reliant when it comes to the knowledge needed to remain competitive to the extent that some authors talk about global innovation systems (Binz et al., 2017). Firms and other organizations operate across national borders and draw upon knowledge developed abroad through global innovation networks (Barnard et al., 2017; Cano-Kollmann et al., 2018), and some of this knowledge has been created by national public investments.

The mechanisms through which these STI global networks may have positive or negative impact upon the host economy’s innovation system is an important area of future research (Pietrobelli et al., 2009, 2011). Recent research suggests that openness will favor strong economies and firms with a strong knowledge base, while weak economies may experience negative effects with a further weakening of the knowledge base as local production capacity is destroyed by foreign competition (Fagerberg et al., 2018). An important area for future research relates to understanding how STI objectives and instruments need to be designed and implemented when knowledge is globally distributed.

As governments become increasingly aware of the issues of openness, conflict can ensue. The World Trade Organization rules on intellectual property rights reflect negotiations between governments with the U.S. trade representatives being closely coached by U.S. multinational companies. As knowledge and innovation become seen as crucial for economic development, there is a tendency for international disputes to appear not only in relation to the intellectual property issues. The definition of what constitutes strategic technologies that need to be kept outside the reach of other countries and treaties has been widened in the United States. China has developed its strategy for “independent innovation” in order to reduce its dependency on foreign multinational firms.

Another specific major challenge for research on innovation and STI policy is to deepen the understanding of how global network giants such as Google, Apple, and Amazon with roots in the United States and Ali Baba and Tencent in China shape the global context for national regional STI policies. Growing rivalry may lead to increased international tension, but it might also offer windows of opportunity for other regions including Europe, Latin America, and Africa.

Understanding the global governance of STI is thus an important area of research that requires collaboration between different disciplines including, among others, innovation studies, international business, development studies, and political economy.

STI Policies for Developing Countries

There is a standing debate on what governments and firms in poor countries can do to upgrade products and processes and hereby contribute to job creation and economic development. The Bretton Woods institutions (World Bank, International Monetary Fund, and OECD) dominated by the most developed countries and most significantly by the United States tend to recommend simple recipes for the poor countries. They recommend full respect for private property, free trade and public policies that attract foreign capital, and public policies that stimulate local firms to join global value chains. It is assumed that by joining global value chains the local producers will more or less automatically enter a process of incremental innovation and upgrading. Recent research results do not support this kind of strategy.

There is a tendency that countries that increase their participation and global value chains through importing components and re-exporting commodities with high import content are characterized by lower rates of economic growth, and this tendency is strongest for the poorest countries (Fagerberg et al., 2018). On the other hand, there are strong indications that countries that build stronger national innovation systems are characterized by higher rates of economic growth (Fagerberg et al., 2009; Lee et al., 2014). This indicates that openness to trade or foreign investments cannot substitute for active innovation policies.

STI Policies Beyond Economic Growth and National Competitiveness: STI Policies Addressing Grand Challenges

The economic competition between nation-states reflects the concern of national governments to promote the well-being of their own national citizens. Competing through building stronger national innovation systems may under certain circumstances contribute to the increase in global wealth and well-being.

But the realities of world development do not reflect such optimistic scenarios. Massive investments in the development of weapon technology have not led to more security for the world population. Financial innovations have increased the risk for worldwide economic crises. While there has been economic growth on the basis of application of knowledge, the economic gains have gone mainly to the superrich and domestic, and, with few exceptions, international inequalities have increased. While a few countries, such as Japan, South Korea, and China, have succeeded in catching up economically with the rich countries through building stronger national innovation systems, most African and several Asian countries remain poor and with very limited access to the global knowledge base. Technologies aimed at increasing productivity and growth have created great social inequalities and irreparable losses in the environment (Steffen et al., 2011). The knowledge-based competition and technological advancement has led to environmental crises where weather conditions and climate changes threaten the livelihood of people around the world.

The current economic, social, and environmental challenges require a shift in the focus of STI policies. Instead of focusing on economic growth and national competitiveness, STI policies should address wicked global social and environmental challenges (Chaminade et al., 2018; Grillitsch et al. (2018); Schot et al., 2016). This requires a deeper reflection on the role of demand for system transformation (Boon et al., 2018) and structural change (Fagerberg, 2018) as well as moving beyond the current focus on the transition in particular technological systems to transformations of the national innovation system.

Throughout history there are many episodes where government interventions have played a crucial role in giving technological development radically new directions. While it is important conceptually to make a distinction between innovation policies aimed at incremental innovation and policies focused on transforming innovation systems, the fact is that both might be needed at different points in time and in different contexts.

Toward Transformative and Transnational STI Policies

STI policy may be driven by different motives. Conventional policies rooted in neoclassical economics ideas typically aim at bringing the economy closer to the ideal market economy. They install private ownership to ideas and try to establish markets where those ideas can be bought and sold. They allow governments to invest in research in cases where the market does not work and where intellectual property rights cannot or should not be established.

However, the neoclassical ideal of pure markets is incompatible with an economy where knowledge and innovation is a ubiquitous phenomenon. Knowledge includes tacit elements and cannot be reduced to information. Cooperation is as important as competition, and the quality of relationships between firms and between firms and the knowledge infrastructure is as important as the knowledge base of each single organization. Here innovation policy needs to give attention to the demand side and to the formation of linkages between users and producers of knowledge and innovation.

The concept of national innovation system helps policymakers to implement STI policies targeting actors, their networks, and the institutional framework in which interactive learning takes place. It is important to make a distinction between narrow and broad definitions of systems and related STI policies. Broad definitions give special attention to people and to how experience-based learning emanating from education, labor markets, and working life shape people and the way they interact in national networks. Historically, the change in the socio-techno-economic paradigm may undermine the position of world-leading national systems and result in new countries becoming world leaders.

Looking ahead, the current mode of governance and development where national economies compete through building stronger innovation systems while the major economic players operate at the global level is unsustainable. Equally unsustainable is to put STI policies exclusively at the service of economic growth and competitiveness, ignoring broader societal and environmental concerns. STI policy needs to become transformative and transnational. The following general principles will characterize the new transformative and transnational innovation policy:

New direction: moving from economic to wider objectives and specifically addressing environmental sustainability and social well-being.

Open knowledge: sharing knowledge across national borders and counteracting the inherent tendency of concentration of knowledge.

Pragmatism: counteracting ideological nationalist as well as extreme pro-market or pro-state ideas.

New governance: combining local citizen participation in innovation with more, and more balanced, supranational governance of finance and knowledge production.

In an era of growing nationalism, these principles seem to be far away from reality. While this is true, it is difficult to see how the current mode of governance of knowledge and innovation can avoid running into crises that might give wake-up calls and start painful forms of policy learning. Science and technology has a great potential when it comes to solving problems in the world, but it can only be done if the institutional setting allows the wise use of knowledge. Understanding how STI policies can be developed in this new global context is the subject of current policy and academic debates and where new developments in the field of STI policies will occur.

Browse All

Share Link

Article contents

Science, Technology, and Innovation Policy: Old Patterns and New Challenges