The growing knowledge intensity of production and competition is often understood narrowly as pressure towards greater efforts in R&D both inside firms and in universities and institutes. But, the discussion so far shows that it is in the combination and interaction of technology, organization and human capital that the increased knowledge content of production is realized. This has immediate consequences for the direction of technological efforts within firms and for national science and technology (S&T) policies. .
        

Traditionally S&T policy has focused on strengthening the supply side of the system, especially in the hope of supporting the development of new products and processes. These were expected to lead to breakthroughs for the local firms which might eventually use such results. In doing so, as the OECD-TEP report remarks, these policies "have probably insufficiently paid attention to the capacity of the economic and social system to incorporate such technological changes and transformations[44] ". While a supply-side focus was perhaps acceptable up to the recent past, it is becoming increasingly necessary to deal with the complexity of the innovation process from the perspective of the competing firm. This implies enhancing the firm's capability to absorb and generate technical change and raising its ability to use technologies from wherever they are generated. The time thus has come to shift the balance of concern "from R&D to innovation and diffusion policy[45] " .

This change of focus is based in part on a fuller understanding of the role that users play in the process of innovation. Rather than being passive receivers of technology, users, it is now acknowledged, play a very active role in the generation of directly applicable technical change[46] . Indeed, as the MIT Comission pointed out, "simultaneous improvement in quality, cost and delivery" as well as "closer links to customers and closer relationships with suppliers" figured prominently among the factors contributing to competitiveness in successful companies[47] .

One consequence of this is for governments to shift from financing innovation outside industry to helping strengthen the processes of technological mastery inside firms and to promoting and facilitating the inter-firm linkages, the exchange of information and all of the other conditions that favor technological interaction between users and producers. These networks of technical collaboration which are woven in a technologically dynamic economy are the core of what has been termed by Lundvall and Freeman the "national system of innovation[48] ". The quality of such a system and its degree of specialization are crucial in determining the competitive advantage of a national economy.

A second is for firms to reconceptualize the R&D function. The traditional model of innovation within the firm was linear and sequential. First, people from the research and development department would bring the product or process up to the prototype stage. Then the functions of product engineering would take place and manufacturing and marketing would follow on. This form of compartmentalized effort takes much longer and is much less effective than what is now being called "simultaneous engineering." Researchers, designers, product, process and manufacturing engineers, marketing and salespeople work together in "self-organizing development teams". Work is integrated towards a roughly defined product goal, development phases are simultaneous or have broad overlaps. The result is a much quicker response to markets, a significantly shorter time to innovation and increased learning from interaction[49] .

Another important change in the traditional notions of technology within the firm is a shift in the focus of attention away from the product or process and towards the whole production system, beginning with the tangible and intangible inputs at the beginning of the chain, through the transformation, packaging and distribution process to servicing the product at the client's premises. Every aspect of the production system -and of the administrative procedures as well- is seen as a target for improvement. The overall result is a visible increase in quality, efficiency and adaptiveness resulting from the additive and systemic effect of minor, medium and major improvements being constantly incorporated at different points in the system.

This incremental attitude to technical change gradually leads to a deeper understanding of the potential and limitations of the technologies in use and points to the direction of more radical change. In short, changes such as this mean that competition is increasingly based on the mastery of technology in an everyday sense; on having production organizations that behave like learning machine.

For each firm, the production system to be improved does not end at its own doors. Technical interaction for the purpose of continuous improvement, system upgrading and new product development requires links outside the organization to other companies and institutions. Cooperation with suppliers, for example, takes many forms, from the basic exchange of information, through personnel training to collaborative engineering or R&D and joint investment[50] . Clients are also seen as technical partners. They are the source of valuable information about user needs and about the performance of existing products. Direct interaction with customers has been found to be one of the most effective means of guiding technical change[51] . Finally, networks with competitors, involving partial collaboration, patent agreements, joint research, and multiple arrangements for access to complementary assets or for sharing the high cost of some activities, are also becoming a feature of the modern competitive firm[52] .

Universities and research institutes are among the most valuable suppliers firms can count upon. Their capacity to provide training, information, technical services and research is important. But, for a wide range of reasons there are barriers which prevent a truly intense and fruitful interaction to take place between these institutions and productive units or networks. 
             

In most developing countries barriers to business collaboration are particularly strong. In relationships between producers and their suppliers there is a long history of confrontation which might need great doses of good will and imagination to overcome.

In order to advance from the mere final assembly stage, most ISI policies included mechanisms to force end product manufacturers to incorporate domestically produced inputs, parts or capital goods, however high the cost or low the quality. The goal of deepening the industrialization process was often achieved but with the unwanted side-effect of creating mistrust and adversarial relations between producers and their suppliers. It is often said in the developed world that firms have to learn to go from "arm's length relations" with suppliers to stable, technology centered, collaborative links. In developing countries the starting point can be better described as "gun-point confrontation". Overcoming these negative attitudes -reforming their institutional embodiments in ministries and industry associations- is a task facing both the private and the public sector on the road to modernization and competitive restructuring.

The experience of technical cooperation with R&D institutes was not very encouraging either. For most countries pursuing an ISI strategy, technology was an input purchased from the foreign originator and put to use by learning established and proven routines. In contrast, the attitude taken by the Japanese and the "catching-up" Asian NICs was to go beyond the mere use of imported technologies according to instructions, and to the creation of a type of organization and a level of technical skill in the workforce that would purposely lead to full mastery and continued improvement of such technologies.

Elsewhere governments in the developing world tried to compensate for the passive attitude towards technology that characterized local industry by building technological capabilities outside firms, establishing research and development institutes and allocating funds for science and technology to universities. These two worlds remained mainly apart despite constant efforts to build a "bridge" between supply and demand. Yet, as Martin Bell says, "supporting institutions can rarely generate technical change on behalf of industry without significant innovative activity on the part of industrial firms themselves[53] ". In other words, a bridge can only be built if there is a support on each side. But for most firms under the typical ISI regime innovation was not a source of profits. Exogenous factors such as the level of protection or subsidies had a much greater impact on profits than technology, productivity or quality of products.

All this led to another shortcoming: a generalized disdain for everyday engineering. Neither firms nor researchers saw incremental innovation as their concern. With a few notable exceptions[54] , the great majority of engineers in developing countries have little experience in the constant improvement of products and processes which is so crucial today. Until recently, engineering graduates were often faced with a stark choice: either the passive operation of foreign technology or research isolated from production.

Competitive restructuring requires activating the links that create a national system of innovation. This depends on establishing a mutually fruitful relationship between industry and technological institutions on new grounds. Such a linkage could begin by the recognition of continuous improvement as an important joint job and of "learning by interacting" as the way to go about it[55] .

Those countries that do not show themselves capable of establishing cooperative links among domestic firms and between these and technical institutes are likely to find it equally difficult to collaborate across borders. The experience developed through collaboration between firms within each country can become an invaluable asset for attempting cooperation between firms from different developing countries and making it successful.

The emphasis here on cooperation for incremental technical change is not intended to diminish the importance of R&D for new products and processes. Rather it is a question of timing in the process of learning to be technologically active. As firms and networks acquire greater mastery of technology in their daily practice, they become more capable of pursuing radical product and process change; as they move to more demanding market segments, they are pushed by competition to do so. This means that strengthening the quality and the capacity of the science and technology infrastructure to prepare for the expected increase in requirements is a wise -and indeed an indispensable- exercise in foresight on the part of both firms and governments. Regional cooperation programs and regional consortia set up to do research about a shared natural resource, product or eco-system is one obvious way to cut the costs for each while increasing the benefits for all. We will return to this below.