Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
3.3.2 Impact of adoption on non-adopters of new technologies
Section 3.3.1 stressed that differential adoption and diffusion of new technologies in the Third World could give rise to a correspondingly diverse pattern of economic impacts at the sectoral and economy-wide levels. This section addresses the general issue of the impact of adoption of these technologies on non-adopters (whether these are individual units or aggregations of units into regions/countries, etc.).39 It is useful on analytical grounds to approach this issue in terms of a broad taxonomy that incorporates the following elements:
1. Different types of adopting and non-adopting units (for example, consuming versus producing units);
2. The location of these units within or between different regions or countries; and
3. The different possibilities that arise from the combination of points I and 2.
For example, the effects on the welfare of consumers in one country arising from adoption of consumer electronics goods in another country could be analysed in terms of the phenomenon of international demonstration effects (see James, 1987b). Or one could evaluate the effects on non-adopting firms in a given region arising from the adoption of the new technologies by other firms in the same region.40 At the level of countries, questions are posed for the impact on exports and imports by non-adopters due to adoption in other countries. At each of these levels - national, regional and individual-important distributional consequences are likely to arise. In particular, to the extent that adopters have high initial levels of income relative to non-adopters, the impact of new technologies on income distribution will tend to be inegalitarian.
This possibility has been raised in the literature almost exclusively in one particular context: the effects of the adoption of innovations in microelectronics by firms in the advanced countries on the comparative advantage of developing countries. Indeed, this theme has been so dominant that special issues of two development journals were devoted specifically to its exploration (see Hoffman, 1985a; Kaplinsky, 1982b). In what follows we first review this particular aspect of the effects of microelectronics on the relationships between adopters and non-adopters and then, in the light of the more general taxonomy above, describe areas that seem to us to demand further attention.
3.3.2.1 Comparative advantage reversal
The central concern of the numerous authors who have addressed this question is that manufactured exports from the Third World will be undermined by the introduction of microelectronics in the developed countries from where, in general, the innovations originate.41 Although rarely spelled out in any detail, the theoretical rationale of this concern seems to be that the nature of new technologies in general (and microelectronics in particular) is heavily conditioned by their origin (overwhelmingly) in the developed countries and that this, in turn, imparts a systematic tendency for diffusion of new technologies to take place in rich rather than poor countries.
The first part of this argument-that dealing with the endogenous nature of technical change-can be expressed concisely in terms of the theory of induced innovation. According to this approach, as espoused by Binswanger and Ruttan (1978), the direction of technical change in any economy tends to be based on prevailing factor scarcities. In developed countries, where labour is relatively scarce and the wage: rental ratio is accordingly relatively high, technical change will tend towards saving labour rather than capital. Thus, insofar as global technical change is concentrated in these countries, it will tend to be poorly suited to the generally very different factor endowments of the developing countries. Even if this tendency is acknowledged, it does not follow that the pattern of diffusion to which it gives rise will automatically induce trade reversals. The reason is that a general tendency to save labour is consistent with two different types of new technologies and these differences have contrasting implications for rates of diffusion between rich and poor countries.
In the first case, new technologies widen the range of technological options in a labour-saving direction, without dominating older technologies for similar products. That is, these technologies systematically save labour per unit of output while requiring more capital per unit of output in relation to the latter. Consequently, the cost-reducing effects of the new technologies can only be realized in countries where labour is sufficiently highly priced and capital sufficiently cheap. This condition will not necessarily hold in all advanced countries, and it may not hold in any. But it is evidently more likely to hold, in general, in rich than in poor economies.42 To this extent, trade reversals might possibly arise; and for some forms of microelectronics technologies this possibility already appears to have been realized to some degree. For example, the differential in the diffusion of robots (which save mainly unskilled labour) between OECD and newly industrializing countries appears to reflect largely differences in wage costs in the two groups of countries (Edquist and Jacobsson, 1988).
The second form in which new technologies save labour, however, does not give rise to any such presumption of trade reversals. For in this case, the new technologies require no more capital per unit of output and for this reason are said to dominate older techniques.43 That is, because they use fewer inputs (strictly, less of one and no more than the other) to produce any given level of output, the new techniques will be more profitable at any set of factor prices. Consequently, they will tend to replace the old in all new investments in the relevant lines of production in all countries. Insofar as adoption patterns are actually determined on this basis, there is accordingly no basis for expecting trade reversals to occur.
The question that is then immediately raised is which of these two forms do new microelectronics technologies tend to assume: that is, do they supplement or dominate the range of existing techniques? Some of the evidence cited above (for example, from the studies by Mody and Wheeler, 1990; and Edquist and Jacobsson, 1988) suggests that dominance is not a pervasive feature of new technologies. As yet, however, the debate on this question remains inconclusive (see Kaplinsky, 1984; Soete, 1985), and it is therefore difficult to assess the empirical basis of the reversals argument. But what can and should be emphasized is that the origin of microelectronics technologies in the advanced countries carries with it no general presumption in favour of this phenomenon.
This conclusion may, however, be challenged on the grounds that it is based on a simplistic two-factor model and that only in a more realistic 'three-factor' world do the systematic links between the origin of new technologies, the features that they embody, and the rates at which they diffuse in rich and poor countries become fully apparent.
Only in this broader context, therefore, is it possible (in this view) to establish an adequate rationale for the possibility of trade reversals. In the following discussion, accordingly, we shall examine three factors that constitute highly important aspects of microelectronics technologies, but which are not captured in the simple two-factor model: namely, skills, the organization of production, and flexibility of output.
Availability of skilled labour In relation to skills, the implicit assumption of some versions of the reversals argument seems to be that these resources are available only in the developed countries, where adoption of new technologies will consequently tend to be concentrated. Unable, because of an absolute scarcity of relevant skills, to make use of these technologies, developing countries will then tend to be confronted with an erosion of their comparative advantage in international trade.
At the level of particular skills and countries, this argument may carry considerable force. But it would seem to have only limited relevance to the developing countries that are currently most heavily engaged in the production of manufactures for export, namely, the newly industrializing countries of East Asia.44 There are several reasons for this. First, these countries do appear to possess at least some of the relevant skills; indeed, the price at which they are able to supply this factor of production may serve to attract foreign firms. For example, in the semiconductor industry, Ernst has observed that
Even for increasing levels of automation, it still matters that, by and large, operators and maintenance personnel in developing countries receive significantly lower wages than their counterparts in the US, Western Europe and Japan.... Caught in a severe 'human capital trap' and in order to reduce the cost of R&D personnel and highly skilled technicians, electronics firms will have to rely increasingly on cheap human capital available in some of the more advanced developing countries.
(Ernst, 1985, p. 30)
Second, it is now becoming apparent that certain new technologies may actually save skilled labour. For example,
CNC lathes can be tremendously skill saving. Proportionately speaking, the savings in skills is greater than that of undifferentiated labour which in turn is far greater than that of capital. This means that the CNC lathe represents a technical change, the appropriateness of which is a function of the degree of scarcity of skilled labour compared to semi-skilled labour and to capital. If we assume that the scarcity of skilled labour in relation to semi-skilled labour is greater in the NICs than in the developed countries, we would be forced to the conclusion that we are witnessing a technical change which has its origin in the developed countries but which is of a greater potential benefit to the NICs.
(Jacobsson, 1986, p. 24, emphasis added)
On the other hand, as noted earlier in connection with adoption decisions, this 'potential benefit' of CNC machine tools (and also CAD) has not been fully exploited in practice by the NICs. For a variety of reasons - such as government policy and information imperfections-the rates of diffusion of CNC machine tools and CAD in the engineering industries of these countries, have lagged quite substantially behind the OECD countries (Edquist and Jacobsson, 1988).
In judging the extent to which this situation actually threatens the comparative advantage of the NICs, however, two things need to be considered. The first is that the product lines which have been most affected by the diffusion of flexible automation techniques are generally not those in which the NICs have been most successful in exporting (ibid.). The second is that (as noted previously) the new techniques often seem to be associated with changes in product characteristics and the implications of these changes for competition between adopters and non-adopters have still to be worked out (as we suggest in more detail below).
Finally, even if skilled labour is neither available nor saved by the new technologies, it may be supplied to the newly industrializing countries, either by foreign firms (seeking to make superior use of the factor abundances in these countries by supplying complementary factors, which are in short supply there), or by the governments of these countries themselves. The evidence presented below (when considering the acquisition of technological capabilities in microelectronics, Section 3.3) of the rapidity with which some of the relevant skills have been built up through state intervention in some of the NICs suggests that the possibility of trade reversals cannot realistically be made to rest on the assumption of fixed factor endowments. On the contrary, in the semiconductor industry, certain East Asian economies appear to have acquired a new source of comparative advantage through the intervention of the state, which has 'created conditions conducive not to investment in unskilled assembly processes, but rather to investment in relatively high-skill, capital-intensive production processes' (Henderson, 1989, p. 159). As a result, what Henderson refers to as a distinct regional division of labour has emerged in East Asia. It is reflected in the facts that:
More and more of the investment in assembly plants for large-batch standardised outputs tended to go to an increasing degree to Thailand, Malaysia, the Philippines and so on, while Hong Kong, Singapore and now, seemingly Malaysia, have tended to be upgraded as to the quality and complexity of their production technologies and labour processes.
(ibid., p. 59)
Organization of production A second aspect of new technologies that is not captured in the two-factor model is the radical change in the organization of production and the work practices that these technologies often seem to require. Exemplified by the Japanese ('just-in-time') system of production, the labour processes associated with best practice levels of productivity in microelectronics confers considerably more autonomy on workers (with respect, for example, to quality control) than is the case with mass-production methods based heavily on a rigid division of labour (see Kaplinsky, 1986, 1987, 1990). Because worker collaboration tends, on this account, to become more important in the new labour process, new forms of control over labour need to be exercised to replace methods (notably, machine-pacing and supervision of labour at the point of production) that were formerly relevant. In particular, as Kaplinsky (1986) has emphasized, social forms of organization, or what is known as 'habituation', come to assume much greater importance. 'Insofar as "culture" or politics determines the success with which labour is habituated to work in this way, the social context of innovation becomes increasingly important in the analysis' (ibid., p. 39).
This social context, however, is unlikely to be generally more favourable in the developed world than in the NICs. On the contrary, for many of the East Asian countries it may be easier, both politically (because of the strong control over labour that is normally exercised there) and culturally (because of an 'ethos' that is relatively close to the Japanese model), to introduce the new labour process than it is in some of the advanced countries. To this extent, far from precipitating a trade reversal, the new technologies may be more likely, in fact, to enhance the comparative advantage of certain NICs vis-ā-vis some of the industrialized countries.45
Flexibility of output A third aspect of new technologies that is neglected in the basic two-factor model is flexibility of output, which, as noted in a previous section, has come to assume increased importance in sectors where demand is prone to change rapidly. Insofar as this factor becomes the decisive element of competition in such sectors (of which garments and textiles are good examples), the location of production would tend to take place entirely in countries that constitute the final markets for the relevant products (which, for the most part, is likely to mean the developed countries and the rapidly growing markets of some large developing economies).
The actual behaviour of producers, however, suggests that this is not the case; that is to say, the time factor is not always (or even generally) the only competitive consideration, even in sectors especially prone to rapid alterations in demand. What firms often appear to seek, instead, is a combination of low wage costs and short turnaround time. In garments, for example, 'there are clear signs that the Caribbean is becoming an increasingly important site for production aimed at the US market' (Mody and Wheeler, 1990, p.61). What appears to be happening is that U.S. producers preassemble the garment (using relatively capital-intensive techniques), which is then sewn in the Caribbean (and exported to the United States under section 807 of the tariff code). In this way, producers are able to combine low production costs and quick turnaround time. A similar pattern can be discerned in the investment by West European firms in Mediterranean Basin locations close to the domestic market (ibid.).
Correspondingly, other locations in the Third World, notably those in the Far East, have tended to lose attractiveness from this point of view.46
Conclusions The conclusion of our discussion of trade reversals is that this phenomenon is a necessary outcome of neither the simple two-factor model nor of extensions to this model which embrace a number of 'third factors'. We have shown that while there may indeed be cases in which reversals seem likely (such as when the new technologies are profitable only at factor price ratios prevailing in developed countries), there are also cases in which the opposite effect may occur [such as when the new technologies save skills that are relatively scarce in the NICs; when they require a labour process that is more easily implemented in these countries than in developed societies; or when there is no systematic bias in adoption patterns between rich and poor countries (which will be the case when the new technologies are more profitable than the old at all price ratios)]. This variety of outcomes may help to explain why no substantial relocation of investment (to the developed countries) has yet taken place.47 But this fact may also be explicable, in part, by the notion that reversals take time to be realized. For example, firms are not likely to scrap equipment in existing offshore locations immediately. Rather, they may behave in a manner consistent with the predictions of the vintage model of diffusion:
It is perfectly rational for entrepreneurs to use old technology even when new best-practice techniques exist. Essentially, old machines can still yield a contribution to profits if price covers operating costs, and they are therefore worth using while this condition holds. Only as the appearance of new machines drives price below operating costs are they replaced.
(Stoneman, 1983, p. 1 15)
For this and other reasons (that are given in the general literature on the diffusion of innovations), locational changes expected to follow from reversal of comparative advantage will tend to be more of a gradual, long-run process than a sudden, radical departure from past tendencies. This is the principal finding, for example, of Hoffman and Rush's study of the impact of microelectronics on the clothing industry, which concludes that:
Earlier predictions that the North's use of microelectronics would lead to an imminent and massive shift in competitive advantage against the Third World must now be tempered considerably. The requisite structural and technical changes for such a transformation in the international division of labor are far too complex to occur in only a few years. However, the equally important conclusion that stems from our analysis is that the clothing industry in the OECD will eventually be altered fundamentally by the developments and trends we have observed. One result of this transformation could indeed be a threat to the competitive advantage of developing countries.
(Hoffman and Rush, 1988, p. 224)
An industry study of automobiles and components draws a similar distinction between short- and longer-term locational changes (Hoffman and Kaplinsky, 1988). Longer-term changes seem likely to comprise a reduction in both export-oriented foreign investment and in offshore sourcing by foreign-owned automobile and components firms (ibid.). The former changes, in contrast, present a much more complex picture. Indeed, what has to be explained is the phenomenon observed during the 1980s, of 'a sharp rise in the volume and value of components and whole vehicle exports originating from DCs [developing countries] and destined for the United States and Western Europe, often tied to TNC activity' (ibid., p. 294). Hoffman and Kaplinsky explain this seemingly paradoxical finding mainly in terms of the competitive response of U.S. firms to the Japanese success in adopting those elements of the new technologies that have been described as 'systemofacture'. In particular, 'fearful that they would not be able to reduce costs quickly enough in the short run, all United States assemblers have since the late 1970s and early 1980s been exploring a variety of conduits and mechanisms to source both finished units and components from offshore suppliers' (ibid., p. 298).
3.3.2.2 Neglected aspects of the relationship between adopters and non-adopters of new technologies
The nearly overriding concern with comparative cost reversal in the literature has led to the general neglect of several important aspects of the relationship between adopters and non-adopters of new technologies. These neglected aspects arise from the focus of this literature (1) on countries (rather than households or firms) as the relevant trading units; (2) on costs (or supply) as the sole determinant of comparative advantage (thereby neglecting demand-side considerations of product differentiation); and (3) on the effects of export contraction rather than changes in imports on the developing economies.
Firms competing within developing countries Taking the relevant unit to be the firm rather than the national economy raises a host of important questions for trade and inequality within the developing countries that deserve much more attention than they have hitherto received. There is a need, for example, to examine the various mechanisms through which losses may be imparted to non-adopting firms. One such mechanism may arise through the product enhancement effect of new technologies (which was emphasized in Section 3.2.2.2). In general, and depending on the extent of substitutability between them, product innovations tend to diminish the market for existing products (as when, for instance, the appearance of improved-quality colour televisions hastens a decline in the market for black and white televisions) (see James, 1985). Moreover, the initial entry of new products may set up barriers to the entry of (often smaller) firms to the market.48 In instances where, on the other hand, innovations take the form solely of changes in processes, non-adopting firms will tend to be unfavourably affected by the decline in the price of existing products that the innovation may occasion. If these firms happen to be small relative to the adopting units, the result, again, may be to increase inequality in the developing country.
The relationship between firm size and adoption seems in fact to be rather complex. On the one hand, some new technologies (such as CNCMTs and CAD/CAM) are in important respects well suited to adoption by small-scale enterprises. On the other hand, questions of risk aversion and information availability tend to favour early adoption by large-scale firms (Edquist and Jacobsson, 1988). The nature of the relationships between the firms themselves must also be considered. For example, small-scale subcontractors may be encouraged to adopt new technologies by the requirements of their large-scale manufacturing clients for inputs of a particular quality (One, 1989).
Households within developing countries Because they are designed mostly for the high average-income levels prevailing in the advanced countries, product innovations in electronics (especially durable goods) tend to be adopted by the relative affluent minority of households in developing countries. The resulting 'demonstration effects' on non-adopting households in these countries may be very important, although they are totally neglected in the literature. For example, adoption by others of new products may induce non-adopting households to supply more labour, with favourable effects on economy-wide output.
Alternatively, non-adopting groups may increase consumption spending out of existing incomes, with corresponding negative effects on savings and economic growth.
The role of product characteristics in international trade Discussion of comparative advantage reversal tends to focus on supply and cost to the neglect of demand and product characteristics.49 Yet product and demand factors are likely to have the most important bearing on the outcome of the relationship between adopters and non-adopters in international trade. It is possible, as noted above, that as the characteristics of products traded internationally change in a 'high-income' direction as a result of microelectronics (because of more consistent or higher quality), the developing countries are enabled to maintain (or even increase) their advantage in 'low-income' trade with other developing countries (see Boon, 1986b). Or again, certain forms of trade between developing countries (such as in cotton textiles) may be based on preferences50 that are not readily eroded by product innovations in microelectronics in the developed countries.
It is worth stressing that if these hypotheses are to be tested, much more disaggregated data will be required than is available in international trade statistics. For even in their most disaggregated form, these data tend to conceal the subtle differences in product characteristics that in our view are so important to understanding the impact of microelectronics on global competition.
Imports, decline of local industry, and effects on employment As noted above, the literature on reversal of competitive advantage is concerned with only one side of the balance of payments problem posed by the adoption of microelectronics-based innovations in the developed countries: namely, the potentially reduced competitiveness of manufactured exports from the Third World. This literature largely ignores the change in imports that might also be occasioned and the effect that this change, in turn, could have on the domestic economy. That the implications of such neglect may be considerable is suggested first by the fact that 'developing countries' imports of manufactured goods exceeds their exports of manufactured goods by a factor of more than three. Nearly 85 % come from developed market economies (Keeling, 1979, p. 17). And insofar as these imports from the rich countries gain in competitiveness as a result of microelectronics-based innovations, the potential for displacement of competing local production becomes obvious, a problem that, in other literature, has not gone unnoticed. Indeed, as Hagen (1982, p. 176) points out, 'the disastrous effect of imports on traditional handicrafts has frequently been noted'. The consequences of displaced traditional handicrafts producers for unemployment and poverty in the Third World have been frequently noted too.
Much depends, of course, on the extent of tariff protection in different developing countries. But to the extent that such displacement does occur, it is important to recognize that unemployment is the probable outcome.
For example, if the introduction of video recorders, produced overseas, leads to expenditure being switched out of domestic products, in the absence of price changes, such a technological change may reduce labour demand. Only if reductions in domestic wages result and lead to import substitution for other goods can full employment be re-established.
(Stoneman, 1983, p. 260)
Much also obviously depends on the extent to which goods affected by new techniques are actually produced in the Third World. In cases where local production is slight or non-existent, consumers or producers may benefit from the importation of products made cheaper by the introduction of new techniques. (On the case of capital goods, see Edquist and Jacobsson, 1988.)
3.3.3 Acquisition of technological capabilities
In the first part of the survey we distinguished between numerous modes of transfer of microelectronics innovations to the Third World. This section examines how these different modes bear on the process by which domestic technological capabilities are acquired. Then, in Section 3.3.4, we take up the question of how (certain of) these capabilities are actually applied in the developing countries and ask, among other things, whether the result is a more appropriate pattern of innovation than would result from dependence on imports.
3.3.3.1 Acquisition of user capabilities
Much of the Third World (especially the poorest parts) relates to microelectronics innovations from the point of view of users (i.e., importers) rather than producers.51 It follows that an important part of any adequate analysis of the effects of microelectronics on local technological capabilities needs to be concerned with various user capabilities, such as 'to follow instructions, to learn how to service it, [an imported good or machine] and eventually to replace parts and repair it' (Teitel, 1984, p. 51).
Unfortunately, there is not a great deal in the literature that specifically addresses this question. What fragments of evidence that do exist point to the role of the supplier in determining whether, and to what degree, user capabilities are fostered by the importation of the new technologies. From this point of view, the role of the supplier is often painted somewhat negatively.52 Narasimhan (1984, p. 50), for example, observes that, in the Third World as a whole,
To a very large extent the needs of the information processing sector are met through imports; ... The import and sale of systems (hardware and software) are done at the of multinationals, their subsidiaries, or their agents. Quite often systems are over-sold, remain under-utilised or un-utilised due to lack of proper training, and after-sales service to maintain these systems in satisfactory operation.53
A similar point has been made by Carroll (1985, p. 110) in relation to microcomputers in development assistance projects:54
Microcomputer activities in projects for sector energy planning and management usually include ... an expressed intent that developing country staff will enhance their analytical skills and ability to use microcomputers effectively in decision analysis. However, under the pressure of time and resources, such good intentions often suffer. Many of the U.S., Canadian and European professionals interviewed ... feel 'more could have been done' either in training or analytical work with developing country staff. The complaint is heard that 'since we left, the computer isn't used much'.
Some authors have discerned much the same problem of inadequate supplier support in relation to electronic capital goods.55
Taken together, these examples point to what may be a quite pervasive problem with the importation of microelectronics technology.56 However, this same problem may, under certain circumstances, lead to substitution of domestic for foreign sources of supplier support. Future research in this area ought to identify the circumstances under which this form of substitution is likely to occur.
3.3.3.2 Acquisition of production capabilities
This section will discuss research on the roles of various technology diffusion mechanisms in the acquisition of production capabilities in the Third World. (The diffusion mechanisms were discussed in Section 3.2.)
Export-oriented direct foreign investment Much of the literature in this area addresses two related questions. The first is whether and to what extent local technological capabilities are fostered by export of assembly forms of production. The second concerns the transferability of any capabilities thus acquired to other areas of domestic production. With respect to the first question, it seems extremely doubtful (as one might perhaps expect) that any significant degree of local capability is generally engendered in the assembly of products for export. With respect to the semiconductor industry, for example, which, as noted earlier, is dominated by foreign investment, Ernst (1985, p. 16) argues that 'export-oriented chip assembly has played a very minor role in establishing an effective transfer of technology to developing countries'. He contends that this form of assembly production has fostered neither 'the skills, knowledge and learning capabilities' of the production workers, nor the engineering capabilities of the domestic economy (see also O'Connor, 1989). Even if this is not the case, that is even if capabilities do happen to be acquired, doubts have been raised about the transferability of these capabilities to other areas of production. In this respect, it appears to be necessary to distinguish the semiconductor from the consumer electronics industry.
Whatever technology and skills are transferred to developing countries through ... offshore semiconductor assembly may be applicable to other assembly-type industries but are not apt to foster the development of a more technologically sophisticated electronics industry per se. By way of contrast, the technology transferred through offshore investments in consumer electronics production would appear to embody certain skills and technical know how which are more directly applicable to the electronics industry. In part at least this can be explained by the fact that in consumer electronics a larger portion of the production process is transferred abroad. Certain developing countries which have perfected manufacturing techniques for components of consumer electronics products have been able to apply their experience ... to the production of peripheral equipment for computers.57
(O'Connor, 1985b, p. 22)
A third strand of literature, like the first, concerns foreign investment in the semiconductor industry. But whereas the latter emphasizes the generally limited contribution made by foreign investment to domestic capabilities, the former addresses the question of why, in countries such as Korea, Taiwan, and Singapore, foreign investment in this industry has been effectively used as the basis for indigenous technological development (as reflected, among other ways, in the emergence of locally owned production complexes and in higher levels of local expertise employed in foreign-owned firms). Henderson's study of this important question stresses a variety of explanatory factors, ranging from 'the changing structure of the world economy and markets and corporate priorities in those contexts, to the development strategies of the national states in question' (Henderson, 1989, p. 163).
Local-market-oriented direct foreign investment Hobday's (1986b) study of telecommunications in Brazil provides the most detailed evidence available related to this type of foreign investment (see also Göransson, 1984). Hobday's research (which we shall mention again below) is extremely important in demonstrating how subsidiaries of multinational telecommunication companies operating in Brazil imparted highly favourable impacts on Brazilian technological capabilities. Hobday adduced considerable evidence of '(a) strategic investments and efforts in local technology, (b) falling technology imports' and (c) developments in management and technological organisation at the firm level'. Hobday suggests that this evidence 'points to a major shift by the MNCs [multinational corporations], from the passive subsidiary role under electromechanical technology, to a positive technology developing role under microelectronic technology' (Hobday, 1986b, p. 301). These new and more favourable technological efforts appear to be explicable in part by government policies, and partly by competition among subsidiaries for the growing market in microelectronic exchanges.
Because they appear to conflict with the conclusions of other studies,58 Hobday's findings pose a fundamental empirical question for future research: Did the successful acquisition of technological capabilities in this case result from an inherent aspect of the telecommunications sector (as opposed to other areas of local-market-oriented investment in microelectronics),59 or is it due instead to some exceptional circumstances of the case?
Licensing Systematic evidence of the impact of arm's-length licensing on technological capabilities in developing countries seems to be confined to Tigre's (1983) study of the Brazilian computer industry. His findings suggest that the short-run impact of licensing may conflict with some of the long-run effects. On the one hand, 'licenses substantially reduced the time required to start up local production of data-processing equipment and helped to avoid mistakes both in product and process design' (Tigre, 1983, p. 113). On the other hand, 'product policy, including product changes and competition, is widely used by licensers as an instrument to restrict licensees' managerial and technical independence' (ibid., p. 169).
A case study of the 'super-mini' computer industry in Brazil indicates that the drawbacks of licensing as a mode of acquiring technological capabilities were not confined to competitive problems with the licensor, although these were still very much in evidence (Evans and Tigre, 1989). The study shows that the effectiveness of licensing in this respect also depends heavily on the specific institutional context in which it takes place: in Brazil's case the 'nationalist environment', based on focally developed technology, often seemed to be in conflict with efforts to acquire technological capabilities through licensing agreements. Evans and Tigre illustrate this conflict with examples that fruitfully combine political economy, industrial organization, and a thorough grasp of the technologies in question. They also use the same methodology in a most instructive comparison of Brazil's attempts to acquire 'super-mini' capabilities with those of Korea, which, for the most part adopted a very different (and generally more successful) strategy.
Loss of effective proprietary rights: role of the microprocessor In Section 3.2.1.2 we saw how the circumstances under which the microprocessor was invented enabled some developing countries to gain access to the new technologies without involvement by MNEs. As one might expect, this development also contained some profound implications for the development of indigenous production capabilities in these countries. Some countries seem to have received a considerable and favourable impact. Brazil, for example, 'has created locally controlled manufacturing capacity in a technologically complex industry. It has shifted its reliance on imports from final products to components. And, most impressive of all, it has generated local capacity for product innovation' (Evans, 1985, p. 27; emphasis added). A similar pattern can also be discerned for India (Grieco, 1984).60
These authors have described these two country cases in detail.61 It is essential, however, to recognize that in both cases, development of local capabilities did not result solely from changes that had occurred on the supply side. Rather, 'it was the emergence of relatively strong governmental institutions in the specific field of computing which in Brazil, as in India, made possible the initial domestic exploitation of these international opportunities' (ibid., p. 162).
Two factors related to the government's role warrant emphasis because both seem to have been necessary conditions for the successful interventions that occurred. One involves the configuration of political and economic forces that permitted the governments in these countries to undertake the policies that they did (among the most important of which were the protection of the local microcomputer industry from foreign competition).62 The second relates to the availability of what Evans calls the 'pre-existing infrastructure'. In the Brazilian case, the availability of this administrative infrastructure was made possible by the fact that 'the development of the people and skills necessary to take an initiative in the computer industry had been going on for at least a decade before the initiatives were taken' (Evans, 1985, p. 38). India, too, is well known for a long tradition of governmental controls and an unusually well-developed administrative structure. It is worth emphasizing that these two necessary conditions are unlikely to be satisfied simultaneously in many developing countries. For even if the political will to make the necessary interventions exists, the required infrastructure often does not.63