Innovation in the Chemical Processing Industries

Ralph Landau and Nathan Rosenberg

Chemicals and allied products (Standard Industrial Classification 28) is the high-tech sector about which the general public probably has the least knowledge. Yet, judged by criteria that are generally regarded as socially and economically worthwhile, this sector should be ranked at the top of the high-tech scale. A common criterion for "high tech" is an industry's expenditure upon research and development (R&D). Chemicals and allied products is at the very top when industries are ranked in terms of the share of total R&D that is actually financed by private funds. With respect to the composition of R&D expenditures, a far larger share of such expenditures in this sector consists of basic research, and basic research and applied research together represent a much greater share of total R&D than is the case in any other industrial sector (see Table 1). It is tempting to say that this sector has received so little public attention because its performance has, in certain respects at least, been so exemplary.

Clearly, chemicals and allied products have been heavily dependent upon the performance of scientific research. Having said that, it must be emphasized that such research is only the very beginning of the innovation process, and not the end of it. A laboratory breakthrough is, typically, very far from the availability of a commercializable product. Commercial success or failure in this industry, as in other industries, is largely a matter of what happens after a laboratory discovery. However significant the contribution of science to human welfare in general, the question of who will benefit most from specific innovations generated by science will depend on factors far removed from scientific research capability.

In chemicals, and especially organic chemicals, the development of



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering Innovation in the Chemical Processing Industries Ralph Landau and Nathan Rosenberg Chemicals and allied products (Standard Industrial Classification 28) is the high-tech sector about which the general public probably has the least knowledge. Yet, judged by criteria that are generally regarded as socially and economically worthwhile, this sector should be ranked at the top of the high-tech scale. A common criterion for "high tech" is an industry's expenditure upon research and development (R&D). Chemicals and allied products is at the very top when industries are ranked in terms of the share of total R&D that is actually financed by private funds. With respect to the composition of R&D expenditures, a far larger share of such expenditures in this sector consists of basic research, and basic research and applied research together represent a much greater share of total R&D than is the case in any other industrial sector (see Table 1). It is tempting to say that this sector has received so little public attention because its performance has, in certain respects at least, been so exemplary. Clearly, chemicals and allied products have been heavily dependent upon the performance of scientific research. Having said that, it must be emphasized that such research is only the very beginning of the innovation process, and not the end of it. A laboratory breakthrough is, typically, very far from the availability of a commercializable product. Commercial success or failure in this industry, as in other industries, is largely a matter of what happens after a laboratory discovery. However significant the contribution of science to human welfare in general, the question of who will benefit most from specific innovations generated by science will depend on factors far removed from scientific research capability. In chemicals, and especially organic chemicals, the development of

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering TABLE 1 Percentage Composition of the R&D Expenditures in the Six Industries of the U.S. Economy where R&D is Mostly Concentrated   Chemicals and Allied Products Nonelectrical Machinery Electrical Machinery   BR AR D BR AR D BR AR D 1965 12.9 38.8 48.3 2.1 12.9 85.0 4.6 13.5 81.9 1966 18.0 39.0 48.0 2.1 13.0 85.0 4.0 12.0 84.0 1967 12.4 37.8 49.7 2.0 13.2 84.8 3.4 12.9 83.7 1968 12.5 37.1 50.4 1.9 12.8 85.2 3.3 13.7 83.0 1969 12.7 38.5 48.8 1.3 15.2 83.5 3.1 15.0 81.9 1970 12.0 38.7 49.3 1.3 15.2 83.5 3.3 15.0 81.7 1971 13.2 38.8 47.9 1.1 14.0 84.9 3.2 15.3 81.5 1972 12.9 39.5 47.6 1.2 13.6 85.2 2.9 16.3 80.8 1973 10.8 40.8 48.4 1.1 13.6 85.3 3.3 15.6 81.1 1974 11.3 39.8 48.9 1.0 13.0 86.0 3.3 15.6 81.1 1975 10.4 38.9 50.6 1.1 12.2 86.7 3.5 15.7 80.8 1976 10.1 41.0 48.9 1.6 11.3 87.1 2.9 17.4 79.8 1977 10.3 41.7 48.0 1.5 11.2 87.3 3.0 17.1 79.8 1978 n/a n/a n/a n/a n/a n/a n/a n/a n/a 1979 9.1 41.6 49.3 1.3 13.1 85.5 3.0 15.4 81.6 1980 n/a n/a n/a n/a n/a n/a n/a n/a n/a 1981 10.1 42.5 47.4 1.9 18.4 79.7 2.7 17.0 80.3 1982 n/a n/a n/a n/a n/a n/a n/a n/a n/a 1983 46.0   54.0 1.4 14.5 84.0 3.1 16.5 80.4 1984 8.4 37.4 54.2 1.6 14.0 84.4 3.0 16.7 80.4 new products depends on the findings of scientific experiments performed at the laboratory level. The initial stages in the development of new polymers, for instance, depend on the laboratory combination of individual molecules (monomers) to form a single composite molecule (polymers). Depending on the length, the shape, and the chemical properties of the individual monomers, one may create materials with different chemical and physical properties, such as plastics, resins, synthetic rubber and fibers, films, and foams. Of course the role of laboratory research becomes relatively less important at later stages of the development process when chemical engineering becomes the fundamental discipline for transforming the bench-scale reactions to production on a full industrial manufacturing scale. Yet the particular nature of the products and the production processes in chemicals accounts for the significance of scientific research at the early stages of the innovation development cycle, which sets closer ties between science and production than is the case in other industrial realms.

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering   Automobiles and Other Transportation Equipment Aeronautics and Missiles Scientific and Professional Instruments   BR AR D BR AR D BR AR D 1965 3.0 97.0 1.4 14.3 84.3 n/a n/a n/a 1966 3.0 97.0 1.0 14.0 85.0 n/a n/a n/a 1967 n/a n/a   1.3 12.8 85.9 n/a n/a n/a 1968 n/a n/a   1.2 11.9 86.9 n/a n/a n/a 1969 n/a n/a   1.1 10.2 88.7 n/a n/a n/a 1970 n/a n/a   1.2 9.6 89.2 n/a n/a n/a 1971 0.7 8.4 90.8 1.1 9.4 89.5 2.2 11.4 86.4 1972 0.6 7.9 91.6 1.1 8.5 90.4 1.9 11.7 86.4 1973 0.3 6.2 93.5* 1.0 10.1 88.9 2.4 10.5 87.0 1974 0.4 6.4 93.2* 1.0 11.6 87.4 2.2 11.0 86.8 1975 0.5 9.5* 0.8 11.2 88.0 1.2 9.6 89.2 1976 0.3 99.7* 0.9 10.5 88.6 1.7 11.6 86.7 1977 0.4 99.6* 0.8 10.7 88.5 1.6 13.0 85.4 1978 n/a n/a n/a n/a n/a n/a n/a n/a n/a 1979 n/a n/a n/a 1.1 10.9 88.0 n/a n/a n/a 1980 n/a n/a n/a n/a n/a n/a n/a n/a n/a 1981 n/a n/a n/a 1.1 12.4 86.5 1.1 9.8 1982 n/a n/a n/a n/a n/a n/a n/a n/a n/a 1983 n/a n/a n/a 1.1 25.0 73.9   13.4   1984 n/a n/a n/a 1.5 19.1 79.3   13.7   Legend                   * Does not include other transportation equipment BR = Basic Research; AR = Applied Research; D = Development n/a = Not available SOURCE: Percentages calculated on data published by National Science Foundation, R&D in Industry, various years. Furthermore, this preeminence of chemicals with respect to research performance is not a recent development. This sector has been the most research-intensive sector of the American economy throughout the twentieth century. If research intensity is measured by the employment of scientific personnel (scientists and engineers) expressed as a percentage of total employment, occasional surveys conducted by the National Research Council indicate that the chemical sector's research intensity was more than twice as great as any other sector between 1921 and 1946.1 An understanding of the present state of this industry, in terms of how individual countries rank with respect to performance and com-

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering mercial success, requires some historical perspective. America's considerable success in this industry in recent decades has to be understood against the background of international differences in natural resource endowments and the working out of what economists call path-dependent phenomena. Because the United States around the turn of the century already had an important domestic petroleum industry, and Great Britain, Germany, and France had essentially no petroleum supplies of their own, the United States readily, and at an early date, switched to a petrochemical base. The switch in resources was full of consequences, because experience with petroleum and petroleum refining led to the acquisition of many skills and capabilities that were, later, readily transferable to other chemical processing activities. This story, of the acquisition of skills and concepts that were acquired in petroleum, and their subsequent transfer to other large-scale continuous processing industries, is a central theme of the historical process by which America gained a position of world leadership. But this emergence had its base in differences in natural resource endowments and the consequences that flowed from that initial difference. This is where path-dependence became crucially important. The abundance of a particular resource at a particular point in historical time set in motion a movement, the direction and momentum of which had consequences that persisted even when the forces that gave rise to that movement had receded. On the other hand, an important aspect of the emerging discipline of chemical engineering is that it may also offer ways of exploiting alternative, lower-cost materials in the production of new or old products. The Haber/Bosch process, the first great milestone of chemical engineering, involved a new way of producing a very old product—ammonia. But it did so by shifting the underlying German resource base from a limited resource—the by-product ovens of the iron and steel industry—to an immensely abundant base—atmospheric nitrogen. There is an interesting counterpoint to these historical developments. On the one hand, the U.S. abundance of petroleum gave rise to a whole set of path-dependent phenomena by shifting U.S. industry to dependence on a resource, petroleum, that was available in abundance. On the other hand, the Haber/Bosch process, emerging in the second decade of the twentieth century, was a supreme instance of a country developing a new technology that enabled it to overcome the shortage of a critical industrial input—nitrogen. Thus, it is safe to say that, in these matters, history does indeed shape present capabilities very much, and many matters in which we have a current interest

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering can be accounted for only by recourse to path-dependency types of explanation. But neither is path-dependency the whole story, much less a simple story. What can be said is that Europe's lead in the chemicals industry, in the late nineteenth and early twentieth centuries, did not provide the most effective path for leadership in the chemical processing technology that later came to dominate the twentieth century. Whereas the United States was a distinct latecomer to the chemicals scene, its abundance of petroleum deposits and the experience that it had gained in continuous processing methods in exploiting these deposits, opened up a technology development path that provided an excellent entry into the chemical processing technologies of the mid-twentieth century. Nevertheless, that can only be a part of the story: Opening up a path in no way guarantees accelerated movement along that path. To put the matter in Toynbeesque terms, challenges sometimes generate vigorous responses; but sometimes they also overwhelm and prove to present insurmountable barriers.2 It is important to grasp the several separate dimensions along which productivity improvements are generated by innovations in chemical processing. There are the major, Schumpeterian innovations that occur relatively infrequently but, when they do, they open up a wide range of significant new opportunities at substantially higher levels of productivity. The Haber/Bosch process is an excellent example of such a major innovation. But chemical innovations not only raise productivity in the conventional sense. They may also offer products that are not only of better quality but are more precisely configured and differentiated to cater more effectively to specific categories of consumer needs. There is a flow of productivity and capacity improvements associated with the use of each of the major innovations. These improvements essentially involve a growing familiarity with a new technology once it has been introduced. Their impact is captured in the declining slope of learning curves or discussed in the vast "learning by doing" literature and popularized by the publications of the Boston Consulting Group (see Figure 1). However, a smooth movement down these learning curves may be interrupted by subsequent major innovations that offer the possibility of moving to drastically new, cost-reducing technologies. There is also a continual flow of individually small design improvements and modifications within the basic framework of indi-

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering vidual Schumpeterian innovations (A1, A2, and A3 of Figure 1). These have the effect of offering superior technologies to firms that are prepared to make the necessary investment in equipment embodying the latest designs and modifications of earlier major innovations that have experienced this subsequent improvement process. Many of these improvements are the outcome of what is essentially a "learning by using" process. That is to say, there are many ways of improving the design and operation of new equipment that become apparent only by observing difficulties or opportunities that emerge during the actual operation of the new equipment.3 Obviously, these small, continuous improvements in design and components become possible only after major, Schumpeterian inno- FIGURE 1. Learning curves in innovation Legend A = Present plant and technology a,b,c = Movements down the learning curve of present plant A1,A2,A3 = Minor, continuous improvements embodied in new plants B = Learning curve associated with major innovation

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering vations have occurred. Such improvements do not—cannot—take place in a vacuum. They are, rather, improvements on a prior innovation that provides a new framework of opportunities; they do not occur independently of such innovations. The essential point is that major innovations set the stage and provide the specific context and opportunities for the smaller, subsequent improvements process. There is much evidence that the cumulative importance of these individually small improvements is immensely important to productivity growth. Unfortunately, it is an aspect of the innovation process that has been badly neglected. The overwhelming emphasis that has been placed, in recent years, on moving down the learning curve of an existing, unchanging plant [category (2)], fails to take account of the steady flow of incremental improvements in plant design [category (3)] that, at some point, makes it economically attractive to introduce new facilities incorporating these later improvements. Thus, a more complete depiction of the competitive process in this industry is that there is a simultaneous movement on two fronts: (a) The technological frontier, originating with a major innovation, is being continually pushed out, as design and component improvements become available and offer competitive advantages to adopters of this latest technology. This is represented in Figure 1 as learning curves shift inward toward the origin—A1, A2, A3 etc.; (b) Firms have the opportunity of moving down the traditional learning curve established by their existing plant and equipment. But it should now be apparent that it is a serious mistake to visualize the competitive process as if it were entirely a matter of squeezing out, as rapidly as possible, the cost reductions offered by such existing learning curves. This is because the ongoing changes in designs and components mean that the well-known learning curve improvements take place on technologies that are, themselves, quickly becoming at least slightly obsolete. In this industry the rapid rate of technological change means that the economic life of a technology is commonly rendered obsolete long before its useful life is exhausted and, perhaps, also long before the firm has been able to approach the lower asymptote of its existing learning curve. Thus, a critical decision is to determine when it becomes worthwhile to commit to an investment that will replace the existing technology with the newest technology. There is an easy formal answer that is provided by economic analysis, which states that firms ought to continue to operate existing technology so long as it covers its marginal costs by doing so. This is, however, only a very inadequate short-run answer in the context of an industry undergoing rapid—and uncertain—technological change. Thus, the fundamental tension in chemical processing plant is this:

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering Adopting the newest technology requires a huge financial commitment in physical and intangible assets of a long-lived nature. Once such an asset is acquired, learning curve improvements make it possible to raise productivity, reduce costs, and perhaps also raise product quality from this equipment. At the same time, however, the steady forward movement of the technology frontier means that it is often possible for a later entrant to start with an equipment base which begins at a cost level that may be lower than that of the earlier entrant. Nevertheless, partially for the reason already mentioned, if the earlier entrant has had the opportunity to move rapidly down his learning curve and gain a commanding market share, the new entrant may not be able to dislodge him. But even this statement understates the inevitable uncertainties and surprises that characterize the innovation process in chemical processing. On the one hand, as already suggested, technical improvements commonly occur before an innovative process has moved very far down its potential learning curve. This is particularly poignant since many promising new technologies are promising precisely because they offer the prospect of sharply declining learning cost curves, but they must nevertheless begin their productive lives at cost levels that may be even higher than the present costs of technologies already in existence (In Figure 1, the upper portion of learning curve B). Finally, sudden shocks, such as a sharp rise in energy costs, or the availability of a cheaper feedstock, can lead to a rapid redefinition of what constitutes an optimal technology. In an industry of long-lived and expensive assets, these uncertainties render the investment decision an especially painful process—one need only recall the years immediately following the oil boycott by the Arab members of OPEC in 1973. As the chemical industry has grown and matured, it has given rise to an entirely new specialization: The discipline of chemical engineering, which simply did not exist a hundred years ago. The chemical engineer has become the critical factor in taking the products of the research process and developing feasible techniques for producing them on a commercial basis. It must be emphasized that the findings of laboratory research do not provide the information necessary for commercial production. Such production is not a matter of simply scaling up the tubes and retorts in which a new product was originally developed. That is often physically impossible and hardly ever economically sensible. Nor is chemical engineering reducible to applied chemistry. It could be better described as the application of mechanical engineering to production activities involving chemical processing. The essence

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering of chemical engineering, then, is a cluster of integrative skills that are applied to the design of chemical processing equipment. But there is much more to it than that. The chemical engineer has, at the center of his activities, the examination and synthesis of different technologies from the point of view of their comparative cost. The work and decision making of the chemical engineer is inherently economic as much as it is engineering, since it involves the explicit consideration of innumerable tradeoffs in determining optimal design. Moreover, it is clear from what has already been said that success in the commercialization of chemical processing innovations has depended critically upon the productivity gains realized through an improvement process that takes place after an innovation is introduced in the market. An integral part of this process of cumulative improvement, which deserves separate recognition and treatment, has involved the exploitation of economies of large scale production and therefore a movement toward larger scale plants. Historically, success in the commercialization of new technologies in this sector has turned upon the ability to make the transition from small scale, batch production to large scale, continuous processing plants. The benefits of larger scale have been so pervasive in this sector that chemical engineers have developed and employed a ''six-tenths rule,'' which is regularly invoked, that is, capital costs increase by only 60 percent of the increase in rated capacity. A distinctive characteristic of the American chemical processing scene even in its earliest years was the continuous pressure toward the exploitation of larger size, and the alacrity with which American firms moved in that direction. One authoritative study, discussing the American situation shortly before its entry into the First World War, has referred to"... the American attitude to the size of chemical works, which was, in short, to build a large plant and then find a market for the products."4 It would seem plausible to infer that such an attitude developed at the time because the relevant markets were, as a matter of fact, both large and growing rapidly. As the industry shifted to petroleum feedstocks in the interwar years and mastered the problems of large-scale, continuous process operations, the optimal size of plant often grew to exceed the market requirements of even the largest of western European countries. Since the European industry had relied much more heavily in its earlier years upon coal as the basic raw material, the transition to larger scale was impeded by skills, attitudes and educational preparation that had been developed under that coal-based industrial regime. European developments were also influenced by the determination of each country to maintain a capability for satisfying the require-

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering ments of its own domestic market. "Even in countries with a relatively large population, such as France and Great Britain, chemical firms planning new projects in the postwar period found it difficult to build a large enough plant that would have reasonably attractive economics. Substantial exports were needed to build such plants, but the products in question would not necessarily be saleable in adjacent European countries, since potential purchasers were still averse to being dependent on supply from across the border."5 Building larger chemical processing plants is, however, much more than merely having assurance of access to sufficiently large markets. Such larger plants are necessarily also a product of technological innovations that make them feasible. In this respect it is much more common than it ought to be to assume that the exploitation of the benefits of large scale production is a separate phenomenon independent of technological change. In fact, larger plants typically incorporate a number of technological improvements, based upon the wealth of experience and insight into better plant design, that could be accumulated only through prolonged exposure to the problems involved in the operation of somewhat smaller plants. The building of larger plants must, as a result, often await advances in the technological capabilities in plant design, equipment manufacture, and process operation. Thus, the benefits of scale cannot be attained until certain facilitating technological conditions have been fulfilled. Both as a conceptual matter and as a practical matter, it is not easy to disentangle the benefits of larger scale production from those achieved through introduction of improved equipment, improved design, or better "know-how," that is, better understanding of the technological relationships that are eventually embodied in the larger plant. Such later plants typically incorporate a large number of cumulative improvements and conceptual insights. The discussion of scale raises a final set of considerations. Scale factors have been important not just at the level of the individual plant and its optimal output, compared to the size of the available market. A central additional question is whether the market is large enough to support specialist plant contractors and designers who will eventually be responsible for delivering the plant and the equipment. This is a critical and badly neglected consideration, because the chemical sector has developed a unique set of specialist firms and organizations which have, in turn, played a major role in the innovation process. These specialists now operate on a world scale for a world market, and commercial success and failure must inevitably be addressed

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering in terms of that world market and the ability of various specialist firms to prevail against competition in that market. Specialized engineering firms (SEFs) came to play a critical role in the chemicals sector during the years following the Second World War. Chemical firms had subcontracted functions like procurement and installation to the SEFs even before the War, when design and process development was essentially carried out in-house. Chemical companies typically carried out their own process design, and used external contractors to handle construction, piping and mechanical work, electrical work, and other separate facets of the project. Petroleum companies typically farmed out most of the detailed design as well to SEFs. After the war, the chemical firms increasingly relied upon SEFs to design, engineer, and develop their manufacturing installations. In the 1960s, nearly three-quarters of the major new plants were engineered, procured, and constructed by specialist plant contractors." 6 There were various specific advantages accruing to SEFs in designing and developing chemical production processes. First, during the 1920s and 1930s, while large chemical companies had concentrated mainly upon product innovation and development, SEFs had acquired an ability to handle sophisticated process design and development work. In this, they had benefited greatly from their experience in the petroleum sector, which had faced, earlier than the chemical industry, problems of large scale processing and refining. The unique capabilities derived from this earlier involvement in design and development work for the petroleum sector constitutes a critical instance of the role of path-dependent phenomena, referred to earlier. As the world moved into the petrochemical age, some countries were better situated by their own past for dealing with the new design and production problems of the new chemical industry. History indeed matters. A further important source of advantage to SEFs came from their opportunity for exploiting economies of specialization and certain forms of learning by doing. Once a major new process technology was developed, or the scaling up of a given production process was carried out, SEFs could reproduce that new technology, or larger scale production process, for many clients. Such economies could not be accumulated by the chemical manufacturers themselves, precisely because they could produce that technology only for their own, limited internal needs, whereas SEFs had a much more extensive experience with designing that particular plant many times for different clients. Moreover, as they worked for many different clients, they

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering accumulated useful information related to the operation of plants under a variety of conditions. This represented an opportunity for accumulating knowledge and specialized skills which were not available to the chemical producers. SEFs thus acquired the capability to design better plants for other potential customers. The role of SEFs had important consequences with respect to competition among chemical manufacturers on a global scale. The most significant was their development of the complete technology and plant designs for the basic building blocks of the chemical industry, for example, olefins and aromatics. American-designed ethylene cracking plants appeared all over the world, and these in turn required technologies for the manufacture of key intermediates for the chemical industries of many nations. Such technologies were supplied by SEFs and manufacturers in other countries who sought to generate additional revenues outside their own domestic markets. Latecomers to a particular chemical technology could benefit from their relations with SEFs, which were able to provide them with the process know-how that they had accumulated, at least in part, through their previous relations with earlier entrants (Thus, in terms of Figure 1, latecomers were likely to be supplied with plant that incorporated the design improvements designated by broken lines Al, A2, and A3). Moreover, the availability of such technology from SEFs also encouraged many new entrants into the industry from related sectors such as petroleum, paper, food, metals, and the like. A result was intensified competition, including periods of overbuilding and excess capacity. In the postwar period, then, the world chemical industry was powerfully shaped by successive waves of diffusion of new technologies, including both product and process technologies. Although the sources of chemical innovation were diverse, a major factor was the role played by American specialized engineering contractors. More specifically, the division of labor between SEFs and the chemical manufacturers had important consequences for the diffusion of new technology, both at domestic and international levels. SEFs licensed extensively to chemical firms all over the world. As a result, they served as major carriers of technological capabilities, including highly elusive but significant "know-how," that is, essential knowledge of a noncodified sort that was, nevertheless, vital to successful plant operation and performance. The vital role played by SEFs in designing and diffusing new technologies in the chemicals sector underlines a point that it is useful to make in closing. That is, the competitive process, even in high tech industries, needs to be examined in terms of a range of activities

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering located "downstream" from the scientific research process. Economists have not, so far, done a very thorough job of this. They have, on the whole, treated technological innovation in a highly abstract way as a collection of activities going on inside a black box, the contents of which are never subjected to systematic examination. When the inputs into that black box are unpacked, it turns out that R&D expenditures are, in fact, not primarily spent on scientific research, but on development which, in the United States has, for many years, constituted more than two-thirds of all R&D spending. Alternatively, even where scientists dominate the initial stages of new product development, the later stages, and eventual commercial success, are likely to be dominated by engineering, design, and technological capabilities. If these "downstream" activities seem to be lacking in glamour and to be, in fact, rather pedestrian, no doubt they are, at least from certain perspectives. But that perspective is likely to belong to the academic or the intellectual, who is interested in "the big picture" or in large conceptual breakthroughs. It is essential to understand that the marketplace renders judgments that are based on modest improvements and the cumulative effect of individually small, pedantic modifications in product or process design. Small, incremental improvements have brought the semiconductor industry from a handful of transistors on a chip to more than a million such transistors; in telecommunications, it has brought the channel capacity of a 3/8-inch coaxial cable to more than an order of magnitude increase over an earlier level; and in the computer industry the speed of computational capability has been increased, by Individually small increments, by many orders of magnitude. In high-tech as well as in low-tech industries, an unkind Providence seems to have ordained that commercial success is likely to favor particularly the possessors of a varied assortment of grubby skills. NOTES 1. See David Mowery and Nathan Rosenberg, Technology and the Pursuit of Economic Growth, Cambridge University Press, New York, 1989, pp. 64-71. 2. See Peter Spitz, Petrochemicals, John Wiley and Sons, New York, 1988, pp. xiii, 26-29, and 57-60. 3. These design and component improvements can sometimes be installed or retrofitted into existing equipment, but usually at a higher cost than when they are introduced at the stage of the actual manufacture of new equipment. In other cases new components can sometimes be installed during normal maintenance and replacement activities. See Ralph Landau (ed.), The Chemical Plant, Reinhold Publishing Corporation, New York, 1966. For further discussion of learning by using, see Nathan

OCR for page 107
Technology & Economics: Papers Commemorating Ralph Landau's Service to the National Academy of Engineering Rosenberg, Inside the Black Box, Cambridge University Press, New York, 1982, chapter 6. 4. L. F. Haber, The Chemical Industry 1900-1930, Oxford University Press, Oxford, 1971, p. 176. 5. Spitz, Petrochemicals, op. cit. p. 348. See also Ralph Landau, "Chemical Engineering in West Germany," Chemical Engineering Progress, July 1958. 6. C. Freeman, "Chemical Process Plant: Innovation and the World Market," National Institute Economic Review, #45, August 1968.   RALPH LANDAU owns Listowel, Inc., and is a consulting professor of economics at Stanford University and a faculty fellow at the Kennedy School of Government at Harvard University. He codirects programs in technology and economic growth and policy at both institutions. The holder of an Sc.D. degree in chemical engineering from M.I.T., Dr. Landau in 1946 cofounded Halcon International, a chemical engineering firm that he headed for 36 years; 20 years later he cofounded the Oxirane Group with ARCO. He is a past vice president of the National Academy of Engineering, and in 1985 was among the first recipients of the National Medal of Technology. His other awards include the Perkin and Chemical Industry Medals, AIChE's Founders Award, and the John Fritz Medal.   NATHAN ROSENSERG is a Fairleigh S. Dickinson, Jr. Professor of Public Policy in the Department of Economics, Stanford University. He is past chairman of the Department of Economics and is the director of the Technology and Economic Growth Program in Stanford's Center for Economic Policy Research. He is the author of numerous articles and several books focusing primarily on the economics of technological change. He has also served on the faculty at the University of Wisconsin, Harvard University, Purdue University, and the University of Pennsylvania. Dr. Rosenberg earned his B.A. degree from Rutgers University and his M.A. and Ph.D. degrees from the University of Wisconsin.