impact may concern the modification of market structure, the creation of new markets or the rendering of existing products obsolete. In reality, however, in Schumpeter’s analysis of cycles, new combinations appear in clusters, thus combining major and minor innovations. Radical innovations initiated by entrepreneurs begin the cycle. The creation of profit opportunities attracts mimicking entrepreneurs who propose minor innovations and thus extend the growth trajectory at a slower pace until the cycle turns around. Researchers today refer to a third category, “disruptive” innovation (Christensen 1997, 2003). Its characteristic is to introduce new performance criteria by targeting different users. It is opposed to continuous innovation and favors new entrants who adopt a different business model. Thus, the notions of disruptive and radical innovation are close, but radical innovation is more associated with new technologies, stemming from scientific and technical progress, while disruptive innovation can also be associated with non-technological changes. Products may be simpler or offer new features that will appeal to new consumers.
The fourth argument is that innovation corresponds to an “economic function” embodied by certain individuals. For Schumpeter, these are the “entrepreneurs”, whose function is to execute new combinations (Schumpeter 1981). We will return to its characteristics later. By emphasizing this function of commercialization or introduction into production, Schumpeter highlights the essential difference between novelty or invention (in the technical field) and innovation. If the invention is defined as a technical solution to a technical problem, innovation consists of productive and commercial exploitation, with the aim of making a profit. The characteristic that distinguishes a mere novelty from an innovation is that the latter involves implementation, whether it is a market launch for a product or service or a productive use for process, marketing or organizational innovations. The objectives of innovation are always economic: to increase sales, to open new markets, to reduce the costs of production, of internal organization, or of internal and external transactions, and to increase labor productivity.
However, the creative power of entrepreneurship and the arrival of the “troop” of entrepreneurs, at the heart of his analysis in The Theory of Economic Development (Schumpeter 1981), depersonalizes itself in the course of his work, showing Schumpeter’s awareness of the nature and scope of the transformations that took place in the structure of capitalism since the beginning of the 20th century. The observation of the existence of what he calls “trustified capitalism” in Business Cycles (1939) will receive increasing attention to the point of becoming the essential cause of the historically determined character of capitalism in Capitalism, Socialism and Democracy (2008). The planning of technical progress by large companies and the development of private research laboratories with the aim of strengthening the potential for innovation mark the “bureaucratization” of technical progress. Technical progress is becoming depersonalized and automated (Schumpeter 2008). This bureaucratization, which is necessary to face competition, is a sign of the strengthening of monopolistic structures, which will, according to him, overwhelm both entrepreneurial spirit and capitalism (see section 13.4). Contemporary economists, especially neo-Schumpeterians, refer to this evolution by evoking a “Schumpeter Mark I” in which it is small firms that innovate, and a “Schumpeter Mark II” in which it is large firms that play this role (Malerba and Orsenigo 1995). In fact, as we show below, systemic relationships unite actors within innovation systems.
1.3. How can we measure innovation, in all its forms?
Innovation is a multifaceted phenomenon and is not easily confined to a typical indicator. This is also the case for the measurement of economic growth, which is expressed, despite the limitations of this indicator, by the evolution of gross domestic product (GDP). Research & Development (R&D) and patents have long been considered as key indicators of innovation, the former measuring the resources allocated to innovation and the latter evaluating the results of the activity.
The measurement of R&D is based on data collected from companies and research organizations, according to the codification carried out by the Frascati Manual, the seventh edition of which was published in 2015. Gross domestic expenditure on R&D (GERD) refers to the total expenditure on R&D performed by businesses, the government, higher education and the private non-profit sector at the heart of the economy. These expenditures include R&D financed from abroad, but exclude the financing of the R&D activities of foreign businesses. R&D intensity is established by the ratio of R&D expenditure to GDP for a country or the ratio of R&D expenditure to turnover for companies.
Global R&D capacity, as measured by public and private investment, has doubled since the mid-1990s. This increase in global R&D capacity is due in particular to the growth in corporate spending, which accounts for about 70% of total R&D spending. While the financial crises that marked the period (the crisis in emerging countries in the early 1990s, the crisis of new economy start-ups in 2001, the financial crisis of 2008) have led to cyclical reductions in R&D investment, companies are also relying on innovation to boost the growth of their activities. This can also be explained by the strong increase in spending as a percentage of GDP in emerging countries (China, Korea and Israel), compared with slower growth in the EU-28, the United States or Japan.
One of the major limitations of this indicator is that R&D appears to be mainly focused on science and technology and has difficulty capturing the expenditure incurred to bring about other forms of innovation, whether organizational or commercial. The services that make up an important part of the new solutions offered by companies are better evaluated by marketing expenses, which are not included in the R&D measurement. As a result, the expenditures made by small firms (which rarely employ researchers) to innovate organizationally or commercially are poorly understood and, therefore, not considered as being very innovative.
A patent is an industrial property title granted to an inventor for a period of 20 years, often used as an indicator of results and innovation performance. It has the advantage of being an available and reliable indicator, whose databases are public. Worldwide patent registration statistics show a strong growth in patent registrations from the 1970s onwards. Innovators around the world filed some 3.3 million patent applications in 2018, an increase of 5.2% for the ninth consecutive year of growth. At that time, approximately 14 million patents were in force worldwide. The largest numbers of patents in force were registered in the United States (3.1 million), China (2.4 million) and Japan (2.1 million) (WIPO 2019).
However, as an indicator of innovation, the patent also suffers from many limitations. On the one hand, it only measures registered “inventions”, thus leaving aside all other possible forms of innovation. On the other hand, many inventions are not patented, especially if they do not meet the criteria of novelty, inventive step and industrial application. An invention that is useful, but not innovative (so if it is already part of the state of the art, in other words everything that has been made public at the filing date), will not normally be able to pass the stage of the research report carried out by the industrial property institute. A registered patent does not always result in an innovation, i.e. a new product launched on the market or a new process integrated into the production process. Many patented inventions remain unexploited, often for strategic reasons. For example, some patents are intended to deceive competitors about the technological trajectories followed. Sleeping patents are also very common. In this case, they are not exploited because the profit prospects are lower than the costs of bringing them to market. The company holding the patent may not have the resources necessary to exploit the invention or may prefer to wait for the previous invention to become fully profitable before launching a new one (in this case, one can speak of “technological Malthusianism”).
However, criticisms have been made about the quality of patents, particularly those granted in the United States in the years 1990–2000, given the favorable attitude of this country towards inventors. In reality, the criteria for patentability have, in many cases, been little respected, leading to the multiplication of low-quality patents and numerous challenges and lawsuits. The statistics would thus be distorted by the multiplication of these “rotten” patents. Another difficulty related to the measurement of patents lies in the fact that they are national titles valid in the countries where protection has been claimed (apart from the unitary patent in Europe,