|
| |
Innovation and Information: The Economics of Knowledge
From: Cambridge University Press
| By:
John Cantwell |
EDITOR'S INTRODUCTION |
In a global world of lightning-speed communications, it may appear difficult to reap the full rewards of investment in successful innovation, as ideas and inventions can quickly be appropriated by others within the public domain. However, in the light of an emerging consensus among economists, John Cantwell of the University of Reading, UK, argues that innovation and imitation are closely related: those who make the best use of information being generated ultimately tend to make a greater contribution to knowledge. |
 y definition of innovation and of technology is a rather broad one. Until quite recently, most of the post-war literature had accustomed economists to thinking of technology in much narrower terms, as akin to knowledge or information, often generated in large firms through the research and development (R&D) function. When the exclusive focus of attention is on this element of technology, the 'appropriability' argument comes immediately to the fore, as to whether firms (and the countries in which they are situated) can appropriate a full return on their investments in the creation of such knowledge or information which can be traded and otherwise dispersed into the public domain. |
 | |
| Even experts can find it difficult to predict the commercial value of inventions produced by research and development. | |
This information-like element of technology thus has the characteristics of a latent public good, since information is costly to create, but once created it may be transferred to others at close to zero marginal cost, and it is non-depletable no matter how much more widely it is used. Since in most cases the patent system provides only a rather ephemeral way of protecting knowledge and ensuring that it is exploited solely privately, this gives rise to the classic problem identified by R. R. Nelson (Journal of Political Economy, 67, 1959, pp. 297-306) and K. J. Arrow (in R. R. Nelson, ed., The Rate and Direction of Inventive Activity, 1962)--that the social rate of return on investments in research and knowledge creation may exceed the private rate of return, leading to an underinvestment by private firms in research and hence in innovation. This problem would be accentuated if knowledge were also to flow increasingly easily and cheaply across national boundaries, since there would then be little point in national governments trying to correct for the underinvestment as they would end up simply subsidising free riders in other countries (the problem would recur at an international level). |
The prevalence of uncertainty
In the standard account, the difficulties of appropriating a full economic rent from investments in knowledge creation have to do firstly with uncertainties over the outcome of research, uncertainties over the ultimate value of knowledge or the associated devices at any given stage of development, and the ease with which imitators can replicate the results at lower (and perhaps at low marginal) cost. The difficulty that even experts in the field may have in forecasting the results of R&D projects, and in predicting the commercial value of inventions, implies that it is troublesome for those engaged in research or invention to ensure that they are fully rewarded for their efforts through contracts with other parties who, prior to the completion of a deal, lack information on the true value of what they are purchasing. Moreover, secondly, because of technological interrelatedness, even if the private value of a particular invention could be established in advance, its creation will contribute to a wider pool of knowledge, and hence there are likely to be external benefits from the wider usage of this latent public good element of any new technology. |
Hence, there are both microeconomic and macroeconomic perspectives on the 'appropriability' problem. From the microeconomic or business perspective, the difficulties that commercial inventors have in realising a full return on their investments in new knowledge creation (due to the difficulties in ascertaining the value of an invention in advance), may give rise to a closer vertical integration between R&D and downstream production facilities within the firm. Yet from the macroeconomic or public policy perspective, because the social benefits of R&D exceed the privately appropriated benefits (due to external spillovers associated with the latent public good properties of knowledge, as well as the problems of valuation at the firm-level), there may be an underinvestment in new knowledge creation from a social point of view. |
While no doubt there is an issue here, once technology is defined more broadly to include in addition a second element--namely, tacit capability--the 'appropriability' problem appears to have been greatly exaggerated, and it needs to be understood in a rather wider context. In a more general (non-neoclassical) framework, the concept of the 'ease of appropriability' relates not mainly to restrictions on knowledge flows between firms (the efficacy of the patent system, the use of secrecy etc.), but instead to the difficulties of learning in a locally specific environment, and of imitating learning processes when moving between different institutions and between different technological traditions. The new literature on technological change emphasises that it is an evolutionary and path-dependent process, which takes the form of the steady accumulation of tacit capability (through learning processes in production), rather than a sequence of discrete acts of knowledge creation (subsequently capitalised upon through changes in production). |
Thus, what is appropriated by firms is essentially a return on advances in tacit capability acquired through learning, rather than a return on the creation of particular items of knowledge. The knowledge generating function is instead to be seen primarily as an input into the learning process, the benefits of which depend mainly upon the contribution made to the problem-solving activity through which the firm consolidates its tacit productive capability, as opposed to the (uncertain contractual) value of any inventions. |
The inspiration for this new view of technology is essentially twofold, and comes firstly from Rosenberg's thorough investigation of the history of technological change, which has shown it to be a cumulative and incremental process of problem-solving in production (a view that again can be traced back to Marx, Smith and the classics: see N. Rosenberg, Perspectives on Technology, 1976; Inside the Black Box, 1982). Secondly, the theoretical articulation of the new approach was provided by R. R. Nelson and S. G. Winter (An Evolutionary Theory of Economic Change, 1982), who suggested that the evolution of tacit capability was encapsulated in the organisational routines of firms' production teams. |
Investing in capability
In the evolutionary perspective on technological change, the two elements of technology are strictly complementary, in that they cannot be used in an operational production system other than in combination with one another. Thus, potentially public technological knowledge and skills (which also has a tacit component, which is understood only by individual practitioners trained and experienced in the activity in question, and so this component of know-how overlaps with the economist's notion of individual 'human capital') can only be effectively exploited by firms that accumulate the requisite tacit capability (which is acquired through a collective learning process within the firm, and thus becomes embodied in a form of social organisation). |
For this reason, knowledge cannot be used at zero marginal cost by firms that did not create it, as they must make their own investments in tacit capability. Where one firm imitates another, or borrows knowledge from another by agreement (as in cross-licensing arrangements), the recipient firm will incur costs that are an inverse function of the degree of relatedness between the technological profiles of the companies in question. When a firm imitates another in a different field the costs of learning and establishing its own tacit capability in a new area are sometimes sufficiently high that the costs of imitation exceed the original costs of innovation in the initial firm, which might occur if the imitating company had a very different technological history and an experience that was distant from or unrelated to that of the original innovator. |
As in the conventional story, in the new evolutionary approach too, we can distinguish between the microeconomic and the macroeconomic perspectives on the problems of technological learning (of which the 'appropriability' issue is just one). At the firm level, the cost of learning rises the more unrelated is a new technology to the existing capabilities of the firm. Thus, the cost of imitating a rival or even a collaborator is higher, the greater is the difference between the technological traditions of the companies concerned. Since technological traditions are always to some extent differentiated, each firm can exploit some sphere of competence relative to others, but firms whose achievements are particularly distinctive will be able to capture the highest returns. Firms invest in knowledge creation (in large companies, formalised in R&D), so as to be able to tailor the search for new knowledge to the specific needs that arise from their own problem-solving in production, and so as to be able to better understand what is relevant to their own requirements from the knowledge being generated outside the firm, and then to be able to properly assimilate it. |
At the public policy level, if firms have limited learning capabilities (say, in a given field outside their existing areas of competence or specialisation) then they will be reluctant to invest in new knowledge creation, which they would find difficult to incorporate into new tacit capability, and hence difficult to capture a return on. However, using this argument the ease of knowledge spillovers between firms is likely to be inversely related with the extent of underinvestment in knowledge creation, and not positively related as in the conventional account. The greater the availability of external knowledge sources from other companies and from the science base (supported by the local government through universities and other public research establishments), the more likely that an individual firm's investment in commercial research will succeed in providing (or suitably identifying and refining) the knowledge inputs needed to help extend and improve its own tacit capability. From this viewpoint, apart from supporting basic research as well as education and training, governments have an important role in actively encouraging knowledge diffusion through intercompany collaborative agreements, licensing and the like. |
Note further that these different views on the 'appropriability' issue imply a significant difference of opinion over the form of the connection between appropriability and profits, which relates to arguments about the origins of profits. In the orthodox market-based interpretation, the 'ease of appropriability' is a variant of the market power notion of the origins of profits, and has to do with the ability (or inability) of the firm to earn a return on its investment in new knowledge creation through contracts with downstream users of inventions. In contrast, in the alternative evolutionary formulation the 'ease of appropriability' of a return on technology creation has to do with the difficulties confronting problem-solving activities in production, which are the learning activities that lead to the formation of firm-specific capabilities, and thus to profits. The more difficult it is to raise productivity and improve product quality the higher the potential return to innovative firms, but also the more such firms come to depend upon an innovative environment of intensive knowledge flows between firms and (at least in the science-based industries) knowledge flows between firms and universities. |
Complements, not substitutes
From this new perspective, innovation and imitation are complements rather than substitutes, since in order to imitate others firms must innovate by creating their own tacit capability, and by adapting any publicly available knowledge to suit their own specific requirements, including the adjustment of methods to the distinctive and unique aspects of their tacit capability. In addition to the costs they must bear, imitators are not free riders, because in the course of the learning process by which they create new tacit capability, they will also normally generate new knowledge of their own, and thereby make their contribution to the public pool. Even in terms of the R&D function as such, firms (at least large firms) that wish to imitate others usually cannot escape their share of the costs. |
For one thing, because the two elements of technology are complementary, when they are absorbing knowledge production companies need knowledge inputs that are tailored to the specific nuances of their own problem-solving activity, and this is normally most easily obtained from an R&D facility under their own direct control (and in turn, as a result, the knowledge and skills generated by that R&D facility are generally most relevant to the development of production in the same firm). For another thing, corporate research has a dual role--apart from creating new knowledge and skills, it is necessary as a means of monitoring knowledge created in the environment outside the firm, and in understanding what parts of that knowledge may be relevant and useful to the firm. |
Innovation and imitation
The most innovative firms are also generally the best imitators, and those that make the best use of the knowledge being created in the external environment generally make the greatest contribution to knowledge themselves. Indeed, the very distinction between innovation and imitation is really quite blurred. Thus, the model that focused upon the returns to single and discrete acts of knowledge creation overplayed the appropriability problem. Another way of looking at the new approach is as a criticism of the so-called 'linear model', which depicts a simple causal chain running from invention (knowledge creation) to innovation and then to diffusion. In practice, there is a continual interaction between learning in production (innovation) and research and science, but on the whole there tend to be more linkages that run from technology in production to research and to science than the other way round. An illustration is the way in which new instruments and computer technology have completely transformed the nature of scientific enquiry in recent years (such as in the use of more powerful telescopes in astronomy, the use of instruments which enable experiments to be done at a microscopic level, and the use of computer simulations and graphics in virtually every branch of science). Moreover, practical technological advances have sometimes led to new areas of science. Thus, Edison's work on light bulbs led to the discovery of the electron, Marconi's practical discovery of the possibility of international radio transmissions encouraged research on electromagnetic waves, while the Wright brothers' success in demonstrating the feasibility of aircraft construction led subsequently to scientific theories of aerodynamics which explained how their technology worked. |
Of course, the mutual interaction between the formation of technological capabilities through problem-solving activity in production and the creation and diffusion of knowledge and skills still does not quite explain the incentive for investing in a local research system and a science base as a means of encouraging a higher level of private investment in corporate R&D and innovation. Instead, it might be argued that in a global economy research and production may become locationally separated, with the linkages between the two sustained at a greater geographical distance--with some of the knowledge inputs into learning coming from afar, either within the multinational firm, or when monitoring external sources of potentially relevant new knowledge. There is a grain of truth in this argument, at least with respect to basic research and scientific advance, which might feed into (and draw upon the fruits of) applied research, development and production in a different location. However, a good deal of recent empirical work has now shown that the intensity of linkages between science and technology still tends to decline with geographical distance. The linkages between science, research and technology have remained essentially localised owing to the importance of face-to-face contacts in communicating the results of complex learning processes which embody a tacit element. |
Conclusion
There is one other point that should be mentioned when considering the ability of firms to capture the returns on innovation. That is, to exploit the knowledge and skills created by research, firms require not only suitably matching tacit capability in production, but also complementary or co-specialised assets which lie downstream in product distribution and marketing. The direction taken by problem solving in production, and the consequent agenda set for allied research facilities, may well depend upon the nature of these co-specialised assets, and the types of product diversification they would either facilitate or restrain. |
This is an extract from pages 229-235 of "Innovation as the principal source of growth in the global economy," by John Cantwell, in Innovation Policy in a Global Economy, edited by Daniele Archibugi, Jeremy Howells and Jonathan Michie, published by Cambridge University Press. Copyright Daniele Archibugi, Jeremy Howells, Jonathan Michie, 1999. |
|
| |
