Industrial capitalism has been the most innovative economic system, but why? We have to assess the performance of economic systems not so much in terms of static allocative efficiency but in terms of dynamic efficiency of their innovative performance. The relevant question is, how can we achieve fast technological progress? We can also ask whether some types of economic organisations are more dynamically efficient, and whether policy makers can improve the innovative performance of their economies.

We shall discuss the limitations of the neo-classical model, and then discuss the evolutionary approach to technological change.

The neoclassical approach to technological change raises serious theoretical questions. The production function analysis assumes that production takes place on the production function. The analysis assumes acquisition of technology is just like acquiring any other piece of information, and assumes perfect knowledge. This conceptionalisation of technology has two implications:

• since firms are rational and technology is accessible to all, all firms should be adopting the same best-practice techniques – however, there are significant differences in the level of productivity of firms within most industries;
• the representation of technology as information is that choosing a technique of production becomes a matter of choice – however, technology requires learning, so that organisational norms and routines are important in how firms acquire technology.

Moreover, there are three reasons why the allocation of resources to innovative activity is likely to be non-optimal:

• indivisibilities: costs of producing innovation are large, involving large fixed costs, rather than discrete and infinitely divisible, so that the price of the innovative product is above its marginal cost;
• appropriability: depicting technology as information means that it has some characteristics of a public good, so that innovators may not appropriate all the benefits from innovations because competitors may imitate them. For this reason, innovators may underinvest - the patent system is one way out;
• uncertainty: innovation introduces uncertainty in a framework, in which the financial returns are unknowable – the response of the consumer market to the product is unknown.

‘Scientific research’ has been considered one of the main sources of new technology. It is codified knowledge that can be transferred to others through journal publications of theories. ‘Technological knowledge’, on the other hand, is more tacit, less general, more specific to the artifacts produced, and therefore less easy to copy and transfer. This kind of knowledge is kept private as possible to provide the producers with competitive advantage in the market. The differences between scientific and technological knowledge in their economic returns they generate mean that scientific knowledge has to be state-funded, whereas technological knowledge can involve private sponsors.

There are two linear models of the innovation process: the first, stresses the role of technology as the prime mover; the second, the role of economic growth.

Most of the research and development (R&D) performed in large industrial corporations is aimed at the introduction of new products and processes. Three types of activity can be identified:

• ‘basic research’ is directed towards the pure advancement of (scientific) knowledge;
• ‘applied research’ is also directed towards advancement of (technological) knowledge, but with direct commercial applications in mind;
• ‘development research’ is directed towards the refinement and improvement of new products and processes.

The stages of the linear innovation process runs from basic science to the innovation on to the market. In this model, innovation is not a single act, but rather a long and costly process. This model has been used by many policy-makers to support public sector scientific research and private R&D.

This model suggests that economic growth (demand) influences inventive effort. Since firms introduce innovations for economic gains, they tend to put their innovative investment into areas where demand is growing so that they can reap greater returns.

However, both linear models are too simplified. We require a framework that understands how demand and technological opportunities interact with each other and with the institutions that form part of the innovative process.

We can think of technology in two main ways: as a body of knowledge, or as an artifact (material product). These two definitions are interlinked since an artifact embodies the knowledge needed to produce it. Yet, information, the instructions necessary to transform raw materials, is only one side of technology technological knowledge also requires skills that are not codifiable and can only be acquired through experience – e.g., ‘learning-by-doing’. These tacit skills and knowledge are embodied in individuals and organisations. Such skills and knowledge include the way in which people interact within an organisational setting, so that each firm as its own knowledge and skills - capabilities and competencies - learnt through experience and practices.

The tacit aspect of technology affects its appropriability. The disincentive to innovate should be reduced when the output is tacit knowledge, which cannot be easily copied. The balance between tacit and codified knowledge, which varies across industries, affects the extent to which firms can appropriate returns to investment in innovative activities. We would therefore expect to observe differences across industries in the effectiveness of means of protecting the competitive advantages of new or improved processes and product (through patents, learning curve, secrecy, first mover advantages, marketing networks, and property rights to knowledge). The effectiveness of each mechanism depends on the technological characteristics in the industry. For example, patents are widely used in the biotechnological industry, where it is easy to specify the innovation; though secrecy and learning are key mechanisms in complex products and processes involving many components such as the car industry.

One important characteristic of technology is that it is systemic. One important consequence of conceiving technology as a system is that we have to look beyond the firm that ultimately introduces the innovation to identify the sources of new technology. Innovative firms do not have all the necessary knowledge to introduce new products and processes, and need to interact with other institutions. This constitutes the technology support system.

The list of institutions involved in the introduction of new technology includes scientific institutions, suppliers, financial institutions, customers, competitors and trade unions as well as profit-seeking firms and government.

Within the system of innovation, the way in which institutions interact during the innovation process is far from linear – constant improvements on existing products and feedbacks are key features. Technical change is an ongoing process of accumulation of knowledge. Each innovation is just a small step in the continuous advance of a technology that co-evolves with the institutions that are part of the system.

The importance of different activities and institutions in the innovation system varies across industries according to how knowledge is generated. In some industries, scientific discoveries are important. In other industries, learning occurs between innovators and users; while in others suppliers are a vital source of innovation.

The idea of technical change as a process of accumulation of knowledge brings a new way of looking at technical change.

In the evolution of many technologies, technical considerations indicate the ‘natural’ way for the next advance, as researchers focus on the gradual improvements of parts of the system: this can be termed as natural trajectory.

A trajectory occurs within a technological paradigm – a shared understanding of what are the core problems, how they are specified, which scientific principles are relevant and how solutions are to be searched for and applied. Technical change within a paradigm, thus, focuses on the improvement of some characteristics. However, changes alter the system’s balance – established practices become obsolete, new opportunities open up. In some extreme cases, it is possible to observe the emergence of new standard and general technologies (i.e., new paradigm) whose diffusion affects almost nearly the entire economy – e.g., informational technologies.

The accumulation of knowledge advances in certain directions. The evolution of technological paradigms can be characterised as path-dependent (where history affects the present and future), creating institutional rigidities so that system can be locked into specific paradigms.

As technology is partly tacit and embedded in systems of institutions, some countries will continue to specialise in specific technologies (e.g., German cars and US information systems). The technological division of labour evolves and develops, giving some countries sustained competitive advantage.