In industrial countries, mill new information and communication technologies produce tremendous gains in productivity? A look at the role of institutional change
Why is it that when computers proliferate and scientists constantly emphasize new discoveries and inventions, rates of measured productivity growth have been lower in the 1970s and 1980s, in most industrial countries, than in the 1950s and 1960s? In Denmark, for example, productivity growth has been zero or negative over the last few years even though computerized machinery has been widely diffused in Danish industry and offices. In the United States, the growth of both labor and total factor productivity slowed down markedly in the 1970s and early 1980s and although manufacturing productivity has recovered somewhat, services productivity continues to stagnate.
This article is drawn from a longer paper, “Technical Change and Long-term Economic Growth,” October 1988, prepared for a 1988 World Bank seminar on Technology and Long-term Economic Growth Prospects, and available from the author. See also “Technological Change in Industry in Developing Countries,” by Carl Dahlman in Finance & Development, June 1989.
Some economists have been so impressed by this anomaly that they have concluded that the measurement system must be at fault. This is too complex an issue to discuss in the space available here. The assumption will be made in what follows, however, that there has been a real slowdown in the growth of productivity in most industrialized countries.
Can the productivity decline be explained by a falling-off in the outputs of research and development work, or of other sources of scientific discovery, invention, and innovation? The evidence generally shows no such trend. Indeed, the microelectronics revolution has led to an upsurge of inventions and innovations which affects not only manufacturing but most service industries as well.
The explanation seems more likely to lie in the processes of structural change and the new investment and training that are needed to diffuse radically new technologies. Thus, it may be that the rate of technological change, in the sense of output of potentially useful discoveries and inventions, has been increasing during the 1970s and 1980s, while technical change—that is, the introduction, diffusion, and efficient exploitation of these discoveries and inventions, has not progressed as fast.
In analyzing the causes of changes in productivity, this article focuses entirely on technical change, and not on the effects of macroeconomic policies, which themselves have an important influence on productivity.
Change in paradigm
A change in “technoeconomic paradigm” takes place when new technology systems change the dominant pattern of production and management throughout the economy. The introduction of steam power or electric power are examples. Change of this kind not only gives rise to new ranges of products, services, systems, and industries in its own right; it also affects the costs and nature of production and distribution throughout the economy.
The very big productivity gains of the 1950s and 1960s were achieved by scaling up the use of flow and assembly-line technologies and, in the European Economic Community and Japan, by catching up with the United States, which pioneered these technologies. Gains from these sources have by now been largely exhausted. During the 1970s the emphasis in the development of industrial technology began to shift away from mass and flow production, intensive in its use of energy and materials, to a much more flexible, information-intensive pattern of production, associated with computerized control and communication systems. The slowdown in productivity over the 1970s and 1980s seems likely to reflect, to a large extent, the problems of structural and institutional adjustment to a new technoeconomic paradigm, based on information and communication technology (ICT).
The four technologies illustrated in the table are all sometimes described as “generic” or “enabling” technologies, since they offer scope for innovations and productivity improvements not just in one product or process but in many. The table suggests that, because of its potential influence on productivity, information and communication technology will prove to be by far the most important and pervasive of the generic technologies in the 1980s and 1990s.
Characteristics of ICT
Both within the ICT industries and in a wide range of applications of these technologies, technical change is very rapid. Underlying this rapid change is the continuing dramatic improvement in the large-scale integration of electronic circuits, and the continuing fall in costs that this permits. In communications technology, fiber optics has made possible similar drastic improvements in costs and performance and, together with computerized telephone exchanges, is leading to a new telecommunications infrastructure, permitting very cheap worldwide image, data, voice, and other types of communication.
The advances in information and communication technology have consequences for the integration and control of production as well as for the use of materials and labor. First, for example, ICT makes possible greater flexibility and speed in changing tooling and dies, and hence in changing product models and designs. This means that small production runs become more economic than before. Since economies of scale are still very important in marketing, finance, and design, the rise of ICT is unlikely to displace large firms from their dominant position, but it is leading them to surround themselves with a penumbra of nearby specialized satellite firms.
Second, the use of ICT to link firms with their suppliers of inputs, and to link producers, wholesalers, and retailers, permits big savings in inventories at all levels in the system, and a far more rapid and sensitive response to even daily changes in consumer demand.
|Scale of investment||1||1||2||1||2-5||1-5||2||2|
|Potential range of applications||3||2||3||2||5||5||3||4|
Third, with the advances in the transmission of information, industries, services, and markets are becoming much more integrated internationally. Faster international transfer of technology and more mobility in the service industries are among the consequences. The financial services industry is already showing the far-reaching effects of this enhanced mobility and many countries and regions will ultimately be affected by major changes in the international division of labor.
Fourth, in many types of machinery and electronic goods, the redesign of products and processes permitted by ICT makes possible a reduction in the number of electromechanical components and a telescoping of some of the stages of component transformation. This permits substantial savings in materials and energy. It also promotes a process of structural change, involving the loss of jobs in some metals and metal goods industries and an increase in jobs in parts of the electronics components industry, electronic products, and producer services.
Fifth, the use of ICT to communicate and store vast amounts of information on sales, inventories, and financial transactions has had very far-reaching effects on banks, insurance companies, retail chains, and supermarkets. It has made possible savings in capital and materials as well as labor-saving organizational and technical changes.
Sixth, the use of ICT to integrate design, manufacturing, procurement, sales, administration, and technical service in any enterprise has had significant effects on the skill profile of industry, affecting drawing offices and designers as well as clerical labor and machine shops. While undoubtedly displacing some labor, these developments also generate a demand for new skills, which are everywhere in short supply, and for new types of information within firms and from specialist firms in the provision of business and computer services.
Given all these characteristics, it is clear both that the new technoeconomic paradigm differs from that prevailing in the 1950s and 1960s in many important respects, and that it entails revolutionary changes in the capital stock, in the skills needed by the economy, and in management systems. How are different industries adapting to this change?
Productivity in industry
Labor productivity has been growing fastest in the electronics industries, especially the computer industry and the electronic component industry. These are the industries that make greatest use of their own technology for design, production, stock control, marketing, and management. They are also the only industries to show a significant rise in capital productivity.
In industries—such as scientific instruments, telecommunications, and watches—that have already been heavily penetrated by microelectronics, both in their product and process technology, labor productivity has risen considerably and recently even capital productivity has advanced.
Institutional change and productivity growth in Japan
Japan’s productivity, in both manufacturing and services, has continued to grow faster than that of most other OECD countries. It is no longer possible to attribute this growth to the gains from imitation: for a long time Japan has itself been the acknowledged leader in consumer electronics; it is now pulling ahead in electronic components, telecommunications equipment, and office equipment and is close to world frontiers in computers and software. The increasing strength of Japanese research and development, as measured in patent statistics and technical innovations, is only part of the story. At least equally important is the institutional adaptation of the Japanese economy to the new technoeconomic paradigm.
First, technology policy has long been a central feature of Japanese industrial policy, and even before the oil price rise of 1973, ICT was explicitly recognized as crucial to Japan’s long-term future. Many measures were taken to strengthen the performance of the Japanese computer and electronic component industries and to diffuse the new technology throughout other parts of the economy.
Second, Japanese firms have proved quite flexible in adapting their organization and management systems to the potential of ICT. One of the most important institutional changes has been the horizontal integration of information flows between research, development, design, production, and marketing. This leads to shorter lead times in Japanese firms than in comparable US firms, in which vertical integration is still more typical, and there is increasingly strong evidence that it also raises the quality of new products.
Third, though the Japanese are themselves critical of their educational system and there are many proposals for its reform, it is clear that the high level of general education for a very large proportion of the total workforce, combined with the intensive training given in the main enterprises, has given Japan a strong capability for introducing and assimilating technical change.
It has also made possible a rather high level of participation in decision making at all levels of the work force, despite the essentially hierarchical and deferential modes of organization. Other major social changes in post-war Japan, such as the elimination of status distinctions between manual and professional workers, have contributed to this widespread participatory approach, which lias has very beneficial results in such areas as quality control and incremental innovation.
Institutions adapt to new technological possibilities in ways that depend heavily on cultural and political factors, so it would be unwise to presume that the Japanese model should be followed by other countries. In Europe, Sweden and Finland have shown the capacity to make institutional changes that effectively exploit the new technoeconomic paradigm while preserving a level of social services and public sector involvement more characteristic of European countries.
In industries still dominated by older technologies, some firms have achieved high productivity increases, rejuvenating themselves through innovations in processes and through redesigning their products. In other firms productivity has stagnated or declined. In many cases ICT is introduced piecemeal, and not yet as part of an integrated system. Under these circumstances productivity may fall temporarily, for lack of the necessary skills in design, software, production engineering, maintenance, and management generally.
Industries such as petrochemicals and plastics, producing standardized commodities on a flow production basis in rather large plants, have made considerable use of information technology in their process control systems and in various management applications. This has helped them to save energy and materials, but in general their capital productivity has declined. They also have suffered from a decline in product innovation. Rationalization and closure of low-productivity surplus capacity, combined with further process innovation in the more advanced plants, can enable these industries to resume their productivity advance as well as improve the quality of their products and control over inventories.
In most service industries, ICT has been diffused relatively little and productivity gains have been small or nonexistent. A group of service industries completely based on ICT, however, are among the fastest-growing industries in the world economy. They include especially the software industry and computerized information services and data banks. In financial services and insurance, where ICT is now extensively used, there is evidence of significant gains in labor productivity. This is important because it has often been thought that the service sector of the economy could not achieve such rapid labor productivity gains as manufacturing.
The advances in information and communications technology call for many organizational and institutional changes. Older management systems, whether in industry or government, are geared to older types of technology. Their resistance to technical change can be overcome only by a prolonged process of social change. What is involved is the transformation of the skill profile of the workforce, the renovation of the capital stock, the redesign of innumerable products and processes, a vast new international infrastructure and many changes in standards, regulation systems, and government institutions.
It is not surprising that despite a high rate of invention and innovation, there should be some slowdown in productivity growth associated with the introduction and assimilation of these institutional and social changes. Nor is it surprising that there should be greater caution in undertaking investment. In most industrial countries, the process of adaptation is likely to take decades. There are long lags in the adaptation of the educational system, and even longer lags in producing the legislative and infrastructural support for these changes.
Shortages of skills are the most critical bottleneck in the diffusion of the new technologies. A recent survey of more than 20 research projects in all the major OECD member countries showed great differences in conclusions about the employment effects of ICT, but universal agreement as to the critical importance of education, training, and retraining of the labor force.
New skills are needed everywhere, both for the rejuvenation of old industries and the growth of new ones. Without enough skilled people it is impossible to embark on the design and redesign of products and processes needed to use ICT efficiently; investment in new capital equipment may lead to a fall in both capital and labor productivity, because of an inability to maintain the equipment, use it to full capacity, or integrate it with other parts of the system. It is not just a question of high-level skills, critical though these are, but of the transformation of the skill profile of the labor force at all levels.
It is not clear what patterns of institutional and social change will ultimately be the best match with ICT. Nor is it yet clear whether the new national and international systems of regulation that are emerging, or, in many industrial countries, current educational policies and practices, will bring back the world economy to the very high rates of growth achieved in the 1950s and 1960s. This article has argued that the potential for high rates of technical change and productivity growth is there and is likely to last for many decades. There is no slowdown in science and technology. But whether social and political systems prove capable of adapting to this potential is still an open question.