Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
2.2.3 The evolution of biotechnological knowledge
In analysing the evolution of biotechnological knowledge it is helpful to think in terms of a development cycle. During the earliest stages of this cycle individuals began to realize that the underlying scientific knowledge has possible commercial applications. Accordingly, steps were taken to appropriate financial returns from this knowledge. Examples are the patents taken out by scientists and universities and the founding of the first generation of relatively small new biotechnology firms. This stage of the development cycle is typically characterized by a high degree of uncertainty-regarding markets, desirable product characteristics, production processes, forms of organization, sources of funding, and, particularly important in the case of biotechnology, the features of state regulation. In the Schumpeterian sense these individuals were acting as 'entrepreneurs', seizing on the commercial potential of new scientific knowledge (see Kenney, 1986, for further details on the evolution of the new biotechnology firms). In view of the high degree of uncertainty, the prevailing macroeconomic conditions may be particularly important during this early stage. For example, it is likely, and somewhat ironic, that the deepening world economic recession of the mid- and late-1970s, together with the relative, and in some cases absolute, decline of some of the older mature industries, created the climate of optimism which greeted the first stock market flotations of equity in the new biotechnology firms.
In some cases biotechnology offered new ways to produce either existing or similar products, or substitutes, for established markets. In these instances, uncertainty related less to the existence and size and more to the ability of biotechnologically produced products to compete efficiently.
An example of an identical or similar product is insulin, which is used to treat diabetics. The gene for insulin was first cloned and expressed in bacteria by Genentech in 1977. Previously, insulin was extracted from the pancreas of pigs and cattle, and about 80% of the world market was controlled by Eli Lilly of the United States and Novo Industri of Denmark.
Two examples of substitutes, one successful the second so far unsuccessful, are starch-based sweeteners and single-cell proteins. In the case of starch-based sweeteners, immobilized enzymes are used as catalysts to transform starch from sources such as corn, potatoes, wheat, or cassava into high-fructose syrup. In many areas, fructose-based sweeteners have competed successfully with sugar and sugar beets. In 1980, for example, Coca-Cola switched half of its sugar purchases to high-fructose corn syrup (HFCS) and 7-Up is now sweetened entirely by HFCS (Ruivenkamp, 1986). The story of single-cell protein (SCP), however, has been less successful. SCP was hailed in the 1970s as an important new industry and received large investments from big firms in sectors such as chemicals and oil. SCP was produced by microorganisms, such as bacteria and yeast; from feed-stock, like North Sea gas, ammonia, and air; and from sugar cane derivatives, such as molasses and bagasse. Used for animal consumption, SCP was seen as a substitute for soya meal. However, despite low prices of oil and oil by-products, SCP has not yet proved to be clearly economically preferable and some of the large industrial SCP projects have been discontinued.
These examples illustrate the extent to which competition between technologies influences technological development. As Rosenberg (1976) has shown, this competition can also stimulate change in old technologies. For example, SCP may provide a replacement for soya as an animal feed. On the other hand, by introducing nitrogen fixation systems to nonleguminous plants, biotechnologies may also increase the productivity of soya plants, which is the old competing product.
In other cases biotechnology opened up the possibility of entirely new products and markets. For example, the production of monoclonal antibodies made possible the development of diagnostic techniques in humans, animals, and plants. One instance is in vivo diagnosis using injectable radiolabelled antibodies to facilitate tumour imaging. In addition, monoclonal antibodies have potential therapeutic uses (e.g., a way to target attack accurately on a particular kind of cancer).
In these cases the uncertainty relates more to potential markets and desired product characteristics. In the early stages of the development cycle, profitability and competition are often based on product characteristics rather than cost (although, as shown above in the case of soya, when there is competition with preexisting products relative cost can be important). Furthermore, there is a relatively high degree of flexibility and variation in process technology as scaling-up proceeds and the search takes place for new methods to overcome constraints and bottlenecks and to make improvements.
In this respect there are important similarities between biotechnology and other industries whose innovation process over time has been closely studied. The relationship between product innovation, including design and process innovation, has been examined for a number of industries (e.g., Abernathy and Utterback, 1975; Utterback, 1979; Abernathy et al., 1983; and Clark, 1985). These studies show that in the early stages of the development cycle, before the emergence of hierarchically structured dominant design concepts, process technology remains flexible, and competition is based largely on product innovation. Clark (1985) elaborates further on the relationship between the emerging dominant concepts that underlie market demand and the development of dominant design concepts.
During later stages of the development cycle, however, the dominant market concepts and dominant design concepts tend to converge. It is during these later stages that Nelson and Winter (1982) hypothesize that further technical change occurs only within the confines of the prevailing 'technological regime'. Such change may result from alterations in relative costs; from shifts in demand within the limits of the existing dominant market and design concepts; or from 'compulsive sequences' (Rosenberg, 1976), 'technological trajectories' (Nelson and Winter, 1977), or 'technological momenta' (Hughes, 1983), which are relatively impervious to shifts in economic variables. Abernathy and Utterback (1975) and Utterback (1979) argue that during these later stages, process technology becomes relatively rigid; process innovation becomes incremental rather than radical; and considerations of economies of scale tend to dominate as the basis of competition shifts increasingly to cost-competition. During these stages market structure also changes, with oligopolistic markets becoming increasingly prevalent.
The main difference between the industries studied by these authors and biotechnology lies in the relationship between the final product (including its dominant design concepts) and the production process. For products such as automobiles, the relationship is close: production is partly structured by the characteristics and stability of design. In bioprocessing, on the other hand, systems are employed in which complete living cells or their components (such as enzymes) are used to effect desired physical or chemical changes. The output of a bioprocessing system (for example, a packed-bed or fluidized-bed reactor) can be used for any number of final products. Accordingly, the relationships among market demand, product design characteristics, and process technology differ from those in industries like automobiles or semiconductors. In this sense, biotechnology is more like process industries, such as chemicals, petrochemicals, or steel, in which the output can be incorporated into a wide range of final products.
Despite these differences, there are important similarities between biotechnology and the innovation cycle studied by the authors cited above. This can be seen clearly by the flexibility and variability of process technologies being employed during the current early stages of the cycle. One example of this flexibility is the current choice between the alternative techniques of batch processing and continuous steady-state processing. Both processes are used in the conventional chemical industry (this overlap also illustrates how knowledge in the 'new' biotechnology industry draws and elaborates on the inherited stock of knowledge). In batch processing the bioreactor is filled with the medium containing the substrate and the nutrients and the biocatalyst are added. After the conversion is completed, the bioreactor is emptied and separation and purification take place. In continuous steady-state processing, raw materials are added and spent medium withdrawn continuously from the bioreactor.
Although most biotechnology production currently employs batch-processing methods, they do have a number of drawbacks. These include the costly turnover time between batches; the greater difficulty of product recovery due to the presence of contaminating biocatalyst; and the greater cost resulting from the difficulty of reusing the biocatalyst. In principle these difficulties can be overcome by using continuous processing methods, which have resulted from the development of techniques to immobilize biocatalysts. This allows the catalyst to be reused, which reduces cost and simplifies product recovery. However, continuous processing methods have their own drawbacks. These drawbacks include the difficulty of optimizing reaction conditions in a single-stage process; of maintaining the stability of biocatalysts over long time periods; and of maintaining sterile conditions over time.
Alternative techniques, and therefore flexibility, also exist at the product-recovery stage. The alternatives include ultrafiltration, which employs membranes and other filters to separate and purify the product; electrophoresis, in which separation is achieved using the different ionic charges of the products; and the use of immobilized monoclonal antibodies as purification agents (U.S. Congress, Office of Technology Assessment, 1984).
Careful analysis of the determinants of process innovation in biotechnology is an important area for future research. To what extent is the search (always an uncertain process) for new and improved biotechnology processes a response to economic conditions, such as cost, availability, and demand? And to what extent is it shaped by 'technological trajectories and momenta' that are relatively uninfluenced by economic considerations? Do competitive processes play a role in bringing about a convergence in processing techniques in areas where one or some techniques begin to establish their superiority to other alternatives? Questions such as these are not purely academic and answers to them would improve our understanding of the forces shaping biotechnological innovation.
One tendency noted in virtually all other industries and which has begun to assert itself in biotechnology is the attempt to realize economies of scale (i.e., reduction in unit costs as volume increases). Indeed Nelson and Winter (1977) go so far as to refer to the tendency towards increasing economies of scale as a 'natural trajectory'. One example in the field of biotechnology is the preparation of monoclonal antibodies (MAbs). The standard technique, pioneered by Millstein and Kohler, involves injecting a purified antigen into a mouse and then, after the mouse has produced the antibodies, removing its spleen and extracting the antibody-producing B lymphocytes. These cells are then fused with mouse myeloma (tumour cells). These tumour cells result in a fused cell, or hybridoma, with the ability to multiply continuously. The hybridomas are then cloned and screened for their ability to produce the desired antibodies (additional details are given in U.S. Congress, Office of Technology Assessment, 1984).
There are two ways to produce the antibody. When relatively small quantities are desired and purity is not at a premium, a hybridoma clone may be injected into mice where it will grow in the abdominal cavity fluid (ascites) from which the antibodies can be collected. When larger quantities are required, or when greater purity is desired (for example MAbs used for human therapeutic purposes must be free of mouse-derived contaminants), the hybridoma clones may be established in an in vitro culture system.
Large-scale cell culture systems may employ techniques of cell immobilization, which allow the MAbs secreted from the cells to be recovered. Damon Biotech Corporation of the United States has patented a microencapsulation technique. In this technique, the hybridoma is surrounded by a porous capsule, which allows nutrients and metabolic wastes to be circulated while retaining the antibodies. The company claims that this technique significantly reduces unit costs in comparison to the ascites method (U.S. Congress, Office of Technology Assessment, 1984).
The importance of economies of scale emerges from data released by Celltech (UK). As can be seen in Figure 2.2, as batch yield increases, the cost of labour per unit of output falls. The cost of materials per unit of output rises, but somewhat less than proportionally at batch yields greater than 100 g. The cost of depreciation per unit of output also rises, but begins to fall slightly after the same yield. Reduced unit costs with increased output thus appears to result primarily from a fall in unit labour costs. This leads the authors to comment that 'The development of small, highly productive fermenters is therefore less critical in terms of production costs than has been supposed' (Birch et al., 1985). Conversely, however, larger reactors do not appear to offer significant capital savings.
Figure 2.2 Effects of scale on components of unit cost
Whatever cost components are responsible, to the extent that economies of scale are realized and become important, they imply (1) increasing barriers to entry and (2) increasing tendencies towards concentration of capital and oligopoly on the processing side of the biotechnology industry. These implications are important for the future of small firms and Third World countries in biotechnology, and they will be examined in more detail below.
These consequences of scale economies may be accentuated by other advantages that large firms or groups of firms might have in bringing together diverse technologies to improve bioprocessing. One example is the use of computers to analyse data from sensors and other monitoring instruments and to make optimal adjustments in nutrients and other variables during bioprocessing. Computer-aided design of proteins can facilitate the production of new enzymes. Another example is special instrumentation: liquid chromatography is used to identify chemical compounds and flow cytometry is used to measure factors, such as cell size, and to indicate the adequacy of nutrient flows.
Opportunities like these have induced many electronics firms to become increasingly interested in bioprocessing, as can be seen by the recent joint venture signed between Genentech and Hewlett-Packard (U.S. Congress, Office of Technology Assessment, 1984, p. 53). However, as we shall see in more detail later, potential contract problems, such as opportunistic behaviour on the part of a research partner, may be an obstacle in the way of joint research. For example, the electronics company might subsequently sell the equipment to other biotechnology firms, thus undermining its initial research partner's competitive advantage derived from the research. (See Williamson, 1975, for a discussion of the possible costs of transacting across markets.) As an alternative to subcontracting or jointly developing desired equipment, a firm has the option of in-house production. However, the viability of this option will be limited by the firm's existing technological capabilities and by the attendant risks and uncertainties.
When limitations such as these prevent in-house production through vertical integration (the 'hierarchy' approach), and when high transaction costs prohibit market-contracted joint research (the 'market' approach), then alternative forms of organizing the search for new biotechnologies become potentially important. One example of such an alternative is the Japanese group, or Keiretsu, in which noncompeting firms are loosely structured around the group's trading company and one or more financial institution. These firms are linked by obligational, rather than arm's-length, market relationships, and these relationships are reinforced by (1) regular meetings between the Presidents and Vice-Presidents of the most important firms in the group; (2) common dependence on credit from the group's financial institutions; and (3) a common relationship with the group's trading company (see Fransman, 1986d, for further details). The large Japanese groups, such as Mitsubishi, Mitsui, and Sumitomo, contain both biotechnology-using firms (in areas such as pharmaceuticals, chemicals, and brewing) as well as electronics and instrumentation firms. As a result of the obligational relations these large groups are well placed to reap the benefits of technological synergies.
Forms of cooperation also exist among firms belonging to different groups. The biotechnology project established by the Japanese Ministry of International Trade and Industry (MITI) will be examined in greater detail below. Other biotechnology projects have been established privately. For example, the Biotechnology Product Research Development Association was established in 1983 to develop chemical products biotechnologically. This association includes both chemical and electronics firms-Kao Soap, Mitsui Petrochemical, Dainippon, Sanyo Chemical, Ajinomoto, Hitachi Electric, and Mitsubishi Electric (Tanaka, 1985, p. 26).
These alternative forms of organization to 'markets' and 'hierarchies' tend to favour large firms. As is made clear in a report on Japanese biotechnological capabilities undertaken under contract from the U.S. Department of Commerce (1985a, b), and published in June 1985, it is from such firms that the United States 'sees the strongest future competition coming'.
Japan will rapidly become more competitive with the US and Europe [in the field of biotechnology] because much of the commercialisation of biotechnology in Japan is being carried out by large established companies. These companies have extensive experience in necessary process control and the financial backing so necessary for bringing products to market.
(U.S. Department of Commerce, 1985a, p. xviii)
In reporting on a recent visit to the United States, a team from the European Community (EC) pointed to a contradiction in the conventional wisdom there: on the one hand, a large part of the vitality and competitive strength of the U.S. biotechnology industry is argued to be derived from the efforts of relatively small biotechnology firms, while on the other hand the greatest fear of future competition arises not from small new biotechnology firms, but from the established Japanese giants.
Two points emerge from the present discussion. (1) As the development cycle evolves for biotechnology, large firms and concomitant oligopolistic market structures are likely to become more important for the reasons outlined above. This will imply a tendency towards increasing barriers to entry with important implications for small firms in industrialized countries and for the Third World. (2) Understanding the evolution of biotechnological knowledge requires analysis of forms of organization.