3.  Equalizing Technological Capabilities

The global distribution of technological capabilities is highly unequal, even more unequal than the global distribution of income or the global distribution of physical capital. Indeed technological capabilities are highly concentrated even among the developed economies. Except for a handful of newly industrializing countries, the Group B and Group C countries have few technological capabilities. Hence the task of international policy should be to increase the technological capabilities of the developing countries as a whole, creating a broad base of scientific and technical knowledge among the population and building on this base so as to reduce gradually the gap between developed and developing countries.

It would be misguided for developing countries to attempt to incorporate the most sophisticated technologies into their present structures of production. An attempt to do so would require massive expenditure on research and development, a commitment to produce large numbers of highly trained scientists, engineers and technicians, and an entrepreneurial class capable of transforming scientific advances into commercially profitable activities. Even if such an attempt succeeded, it would warp the economy, accentuate inequality in the distribution of income and produce lower returns than alternative, less capital intensive investments.

The usual definition of a "high technology" industry is that expenditure on research and development (R&D) accounts for at least 10 per cent of value added. This requirement for heavy capital expenditure on research partly explains why R&D expenditures are highly concentrated even within the industrialized countries. Five countries--the United States, Japan, Germany, France and the United Kingdom--account for 85 per cent of all R&D expenditure in the OECD countries and about the same percentage of all persons engaged in research.26  Comparing the industrialized countries with the developing ones, it can be seen in Table 3.1 that the former have more than nine times the number of scientists and technicians per 1000 people than the latter and about 13 times the number of scientists and technicians engaged in research and development. The comparison with the low-human-development countries (excluding India) is even more unfavourable. Graduation statistics show that these differences probably are getting worse: the industrialized economies produce about 9.5 times more science graduates per age cohort than the developing countries. In fact, while the gap between the developed and developing countries has been narrowing in terms of adult literacy and primary education, it has been widening in terms of average years of schooling and overall enrollment ratios. 27

Source: UNDP, Human Development Report 1993, Tables 5 and 32.

If technical innovation is the ultimate engine of economic growth, as is widely believed, then the concentration of technological capabilities in the industrialized countries suggests that living standards between developed and developing countries will continue to diverge, global income inequalities will increase and the technological lead of the developed economies will grow. The Human Development Report 1992 contained evidence of the increasing inequality in the world distribution of income. The same trend is evident when one examines technological indicators. For example, it can be seen in Table 3.2 that in the United States, Japan and Germany, expenditure on research and development accounts for about 2.9 per cent of their GNP. Among the more industrialized of the developing countries, only South Korea comes close to this, allocating 2.3 per cent of its GNP to R&D. Taiwan and India allocate about one per cent to R&D and the other countries in the table significantly less than this.28  These percentages translate into enormous differences in absolute levels of expenditure because of the much larger size of the economies of the United States, Japan and Germany. The United States, for instance, spends about 30 times as much as South Korea and about 789 times as much as Thailand.

Sources: UNDP, Human Development Report 1993, Table 32; World Bank, World Development Report 1992, Table 1; Sanjaya Lall, "Technological Capabilities and Industrialization," World Development, Vol. 20, No. 2, 1992, Table 2.

Expenditure on research and development tends to be concentrated within certain industries--chemicals, microelectronics, telecommunications--and within these industries, R&D is concentrated in a small number of large transnational corporations. This high degree of concentration in large monopolistic or oligopolistic enterprises means that the firms are price makers rather than price takers. This in turn implies that most of the benefits of innovation are retained by the firm in the form of higher factor payments rather than distributed more widely in the form of lower prices. Intense competition in the computer industry is an exception to this pattern, and prices have fallen dramatically, showing the significance of competitive pressures. In the absence of intense competition, the developing countries (and customers in other countries) are forced to pay a monopoly rent to innovating corporations, a rent derived from intellectual property rights in advanced technology.29  In other words, the international market for technology is highly imperfect, competitive forces often are relatively weak and many barriers to increased competition are present.

The monopolistic structure of the international market for technology tends to result in a cumulative widening of differences in technological capabilities between developed and developing countries. This tendency is intensified by the nature of the new information and communications based technologies which are growing very rapidly, even relegating some of the industrialized countries to the status of a consumer rather than producer of these technologies. The countries which are moving quickly to base their economies on these new technologies have a high level of technological capabilities, a diverse educational and industrial structure, an abundance of capital and a large services sector. These are the opposite characteristics of most developing countries.

An advantage of the information and communications based technologies is that there appear to be no physical constraints to their continued growth since they rely on few non-renewable resources and produce little waste or pollution. The technologies potentially could be of great benefit to developing countries because they can cut costs substantially, reduce the optimal scale of production and lower the consumption of energy, but apart from some of the newly industrialized economies, most developing countries are unable to take advantage of the new technologies because they lack the capability to master their use.30

Public goods and property rights in technology

At the heart of the technological gap between developed and developing countries is the question whether firms should be allowed to claim intellectual property rights for their innovations. That is, should society allow knowledge to be privately owned? The issue is complex and the debate has strong ideological overtones. Firms which invest in research and development argue that they should be given monopoly rights to the innovations they produce in order to cover the costs of undertaking research. The justification for monopoly rights is that the market for information is inherently imperfect because of the free rider problem, i.e., the inability of the innovator to exclude imitators from appropriating the fruits of his effort. Because of this market failure, it is argued that government intervention in some form is necessary, e.g., by subsidizing R&D activities directly or by granting indirect subsidies through protection of patents. Without intervention in some form there will be a sub-optimal allocation of resources to innovative activity. This underinvestment is inefficient because it represents a failure to allocate resources where returns are highest to society as a whole.

The argument is correct as far as it goes, but it seems to imply that society should be indifferent as between direct and indirect subsidies to R&D activities. This, however, is not correct because once a new technology exists, efficiency requires that the technology should be available to all potential users at a price which reflects the marginal cost of dissemination. Since for all intents and purposes information is infinitely divisible, the appropriation of knowledge by one person does not exclude another from using it and hence the marginal cost of providing information to an additional user approaches zero. The costs of research and development are sunk costs and are irrelevant to the calculation of marginal cost.

Information, thus, is a classic example of a public good. If innovating firms are granted patent rights, the cost of information fails to reflect its marginal cost but instead reflects monopoly power. The resulting high price leads to static inefficiency in resource allocation and a deadweight cost to society as a whole. The defenders of the patent system argue that this deadweight cost should be regarded as a form of investment that society should be willing to bear in order to reap the dynamic efficiency gains from increased innovation. 31

It is not obvious, however, that the dynamic efficiency gains outweigh the static efficiency losses, although the conventional assumption is that they do. Moreover, competitive rivalries may provide a sufficient inducement to firms to innovate without a need for further incentives in the form of patent protection. Barriers to entry of various sorts may give oligopolistic firms plenty of time to reap monopoly rents from their innovations. Those firms which are the first to innovate have opportunities to build consumer loyalties and to differentiate their product from those of their competitors. Furthermore, unauthorized imitation may be costly, since learning requires expenditure and information does not flow perfectly. This is particularly true in developing countries where entrepreneurial abilities, engineering expertise and basic technical skills are scarce.

In general there is no convincing evidence that patents are necessary to stimulate innovative activity. Most of the patents issued are for relatively minor changes in already existing technology, not for genuine innovations.32  Most commercially important industrial patents either originate in large transnational corporations with considerable market power or are purchased by them. As a result, the granting of monopoly rights to technology accentuates the existing concentration of market power and diminishes global competition.

In reply to these points it is often argued, first, that if the patent system is not effective, the principal reason is that it is not sufficiently protective. The length of patent protection should be increased and the scope broadened in order to provide stronger incentives to firms to undertake risky expenditure on long term research and development. It is also argued, second, that one of the purposes of the patent system is to facilitate the dissemination of technological information by requiring detailed public disclosure as part of the patent application process. These arguments however are not very persuasive. The standard length of a patent in the industrialized countries is 15-20 years, surely long enough for a company to recoup its cost. And while disclosure is indeed required, as long as the patent is in force, it is likely to be quite costly for a competitor to develop a comparable technology. Quite apart from this, it is important to recognize that all firms which innovate have been subsidized to the extent that they rely on basic scientific research that was publicly funded and employ workers whose technical education was provided by the public sector. It is not evident that, beyond these general subsidies, firm-specific subsidies in the form of patent rights can be justified as being in the public interest.

An alternative to patent protection is selective, direct subsidies of applied research. In contrast to basic research, which is largely motivated by a desire to discover scientific principles, applied research is concerned with the development of new products and technologies which are commercially profitable. Direct subsidies of applied research are commonly part of an industrial policy and they have been used with success in Taiwan and South Korea to channel resources into industries which it is thought can be internationally competitive in future. Industrial policy thus discriminates among firms and industries by directing resources to some and not to others. Although historical experience suggests that some form of industrial policy can be useful in developing countries, it cannot be assumed that government planners are always or even usually more adept than private entrepreneurs at identifying future commercially profitable opportunities.

In general, the state is in a better position to finance basic research, an area which private enterprises are likely to neglect because of high risk and the long delay in reaping a return on investment. Expenditure on basic research, however, is likely to be a higher priority for governments in developed countries. Indeed it is likely to be a relatively low priority in most developing countries. This in itself need not be a cause for concern if the results of government financed basic research in the developed countries can be made available at low cost to potential users in developing countries. Unfortunately however the public good character of basic research in the developed economies has diminished in recent years because of increased collaboration between university research scientists and private corporations. While this collaboration is attractive to universities in a period of tight budgets, it has negative long term consequences arising from the fact that the profit interests of private corporations increasingly affect research priorities. In addition, universities on their own initiative have begun to apply for patent protection of their research, further restricting the free flow of knowledge.

The solution to these problems is for the governments in developed countries to resume their traditional role of financing basic research and placing the results of such research in the public domain, free of charge. Now that the Cold War is over, research budgets can be reallocated away from military research to favour research that will strengthen the civilian economy. This would be in the public interest in the developed countries and also in the global interest. Basic medical research (e.g., on AIDs) or research on agricultural technology (e.g., on nitrogen fixation in plants) could be especially valuable in the developing countries. A research policy of this type would be an indirect and disinterested form of foreign assistance, from which both developed and developing countries would benefit.
 

Building a broad base of technological capabilities

Within the developing countries there are strong arguments on grounds of efficiency and equity for placing primary emphasis on mass education as the way to increase technological capabilities. Conventional thinking visualizes technological progress in idealist terms, as a series of dramatic breakthroughs made by an elite group of scientists and engineers. Careful study of the historical record however suggests that this vision is incorrect. Technological progress can better be understood as a long, gradual process based on shop-floor improvements discovered or devised by ordinary technicians and workers while struggling to adapt new technology to their specific conditions. Often there is a smooth continuum from initial discovery of a scientific principle to development of its technical feasibility to transformation into a commercially viable innovation and finally to widespread diffusion of the technology. In the case of Watts' steam engine, for example, it took a full century after the initial development in the 1760s and 1770s, a century of design changes and gradual improvements, before this new source of power displaced the sail on ocean-going vessels and overtook water power on land.33

An important aspect of the continuum of technological innovation is the process of learning and the development of the human skills necessary for introducing and diffusing a new technology. These "learning-by-applying" efforts lie behind sustainable technological advances. They are the result not of chance or unpredictable brilliance but of a deliberate pattern of resource allocation: a pattern which favours not so much highly educated researchers and well-equipped laboratories and research facilities as primary and secondary education and, above all, popular scientific and technical education. The skills produced by such a pattern of expenditure are the ultimate source of innovations. These skills, however, are acquired informal aptitudes, general predispositions which are not registered on standard rating measures of formal technological capacities.34

These informal aptitudes embrace "learning-by-doing", minor improvements which are motivated by cost savings, enhanced maintenance, reverse engineering and "learning-by-using". As was pointed out in Kenneth Arrow's classic work on "learning-by-doing", many incremental improvements in technology occur as by-products of workers and technicians discovering how to carry out production more efficiently. 35  Nathan Rosenberg's concept of "learning-by-using" is similar, but includes a more explicit effort to understand the technology embodied in a product or a machine, often by taking it apart to identify its design.36  Both "learning-by-doing" and "learning-by-using" are low cost activities and hence it should be possible for developing countries to increase their technological capabilities quickly by giving priority to mass scientific education. Higher costs are associated with more advanced formal education or "learning-through-training". This involves some systematic instruction in the "know-how" of technology but it does not necessarily include an understanding of the "know-why" of technology or the basic principles of operation. Again, the historical evidence suggests that "learning-through-training" was crucial to the success of those developing countries which became internationally competitive in exporting manufactured goods.

Still more advanced is "learning-by-searching". This refers to the ability to investigate the available technology and to choose what is most appropriate for a given country's stage of economic development.37  If a country is able to engage in "learning-by-searching", it must have a larger stock of scientists and engineers than countries pursuing a broader based strategy and must devote a considerable volume of resources to research and development. Although "learning-by-searching" is important and ultimately will be essential, most developing countries would not be well advised to give top priority to creating a high level of capacity to investigate and absorb foreign technology. The resources required to nurture a domestic scientific and engineering elite and to mount a major effort in R&D activities would be better used to provide education and training in scientific subjects for the majority of the country's labour force.

In most developing countries indigenous technological capabilities can be increased most efficiently by giving priority to constructing the foundations, i.e., by concentrating on basic human capital formation. Broad literacy and numeracy come first. Next come education in popular science and elementary technical training. The acquisition of these essential skills by a large proportion of the population is the surest way to promote sustainable and steady technological change. An emphasis on "learning-by-applying" will ensure that a country is able to institutionalize continuous improvements in its technological capabilities. In the absence of this, even massive expenditure on educating scientists and engineers, hiring expatriate research personnel and on R&D will produce meagre results.

It takes many years to produce a stock of competent scientists and engineers. In Africa less than one per cent of the children entering primary school go on to study post-secondary science, engineering or technical subjects. The reason for this is that children in primary school are poorly educated in science and mathematics and this weakness of the school system makes it difficult to improve technological capabilities. To produce a scientist or engineer normally requires 5-10 years of education beyond secondary school, but in order to have a critical mass of students at the post-secondary level it is necessary to provide secondary school students with an adequate background in science and technology, and this in turn implies that the foundations at the primary level must be strong.38  Alas, in many developing countries, not just in Africa, the foundations are weak. Indeed in most developing countries a strong emphasis on basic human capital formation is essential both to ensure broadly based technical competence among the labour force and to produce an adequate number of scientists and engineers.

An emphasis on basic human capital formation also has the advantage that it helps to ensure that the fruits of technical change are widely distributed among the population. The alternative strategy, relying on an elite group of scientists and engineers, tends to increase the concentration of incomes. Again, the more broadly based strategy of emphasizing basic human capital formation is more likely to result in the adoption of technology that is appropriate to the country's existing structure of production and is designed to meet the basic needs of the population.
 

Technological capabilities and comparative advantage

As regards advanced technology, the present unequal global distribution of capabilities is unlikely to change in the near future. If anything, it is likely to get worse. The developing countries will remain importers of advanced technology, but they should develop some capability to become selective importers, able to modify and adapt the technology they have purchased to reflect their own resource endowments and their own needs.

Although the gap between developed and developing countries in "high-tech" industries may well continue to widen, the developing countries can reasonably aspire to close the gap, or at least reduce it substantially, in basic human capital. Specifically, if spending priorities are adjusted, it is reasonable to anticipate that within a generation the developing countries could sharply reduce the differentials in primary and secondary school enrollments and they could narrow the gap in technical school graduates and in the general technical competence of the labour force. At the end of one generation all developing countries could possess the fundamental technological capabilities.

If this objective were achieved, it would alter the comparative advantage of the developing countries. Indeed if a human development strategy is pursued consistently, then in the long run human capital will emerge as the main source of a country's comparative advantage. It is often argued that the comparative advantage of developing countries in the international division of labour rests on their relative abundance of unskilled labour. In a static sense, at a given moment in time, this may be true, but in a dynamic sense human capital formation can shift a country's comparative advantage from labour as such to skill-enhanced labour. That is, as the stock of human capital increases and it becomes relatively more abundant, the composition of output will change in favour of more human capital intensive products. This will be accompanied by an increase in the productivity of labour and a rise in living standards. Moreover, the broader is the base of skills and education, the more mobile and adaptable will be a country's labour force and the more flexible will be its economy. This will increase the ability of the economy to withstand shocks originating abroad and thus increase the benefits of integration into the world economy. In other words, a human development strategy not only has direct implications for the well being of people because of its domestic economic consequences, it also has implications for the gains from international intercourse and hence indirectly for the benefits people obtain from participating in a global economic system. The long term comparative advantage of developing countries that pursue a human development strategy will rest on technology embodied not so much in its stock of physical capital as in its human capital.

26  Ann Johnston and Albert Sasson, New Technologies and Development, Paris: UNESCO, 1986, p. 24.

27  UNDP, Human Development Report 1993, New York: Oxford University Press, Tables 6 and 7.

28  For the developing countries in Table 3.2, the expenditure on R&D as a percentage of GNP refer to various years between 1982 and 1987. See Sanjaya Lall, "Technological Capabilities and Industrialization," World Development, Vol. 29, No. 2, 1992, Table 2.

29  See Keith Griffin, International Inequality and National Poverty, London: Macmillan, 1978, Ch. 1.

30  Ann Johnston and Albert Sasson, op. cit., p. 70.

31  Keith E. Maskus, "Trade-Related Intellectual Property Rights," European Economy, No. 52, Commission of the European Communities, Directorate-General for Economic and Financial Affairs, Brussels, 1993.

32   Ibid., p. 172.

33  Nathan Rosenberg, Perspectives on Technology, London: Cambridge University Press, 1987, pp. 192-3.

34  Aaron Segal, "From Technology Transfer to Science and Technology Institutionalization," in John R. McIntyre and Daniel S. Papp, eds., The Political Economy of International Technology Transfer, Westport, Connecticut: Quorum Books, 1986.

35  Kenneth J. Arrow, "The Economic Implications of Learning by Doing," Review of Economic Studies, No. 29, 1962, pp. 155-73.

36  Nathan Rosenberg, Inside the Black Box: Technology and Economics, Cambridge, Massachusetts: Cambridge University Press, 1982.

37  Edward J. Malecki, Technology and Economic Development: The Dynamics of Local, Regional and National Change, New York: Longman Group, 1991, chapter 4.

38 Aaron Segal, "Africa: Frustration and Failure," in Aaron Segal, ed., Learning By Doing: Science and Technology in the Developing World, Boulder, Colorado: Westview Press, 1987.