The Dynamics of Industry Structure: The Chemical Industry in the US, Western Europe, and Japan in the 1980s

Ashish Arora

Alfonso Gambardella

Center for German and European Studies, University of California at Berkeley

March 1997

Abstract

This paper analyzes the evolution of the structure of the chemical industry in the US, Europe, and Japan. Differences in institutions, historical conditions, and resource endowments across the three regions reinforce differences in initial conditions. However, technological innovation, the internationalization of the industry, and the development and operation of markets, especially markets for technology, capital, raw materials, and corporate control, are powerful forces encouraging convergence. Convergence is less marked at the level of the firm than at the level of the industry, and is more marked between the industries of Western Europe and the United States.

Ashish Arora, Heinz School of Public Policy and Management, Carnegie Mellon University

Alfonso Gambardella, Dept. Of Economics and Management, University of Urbino

1. INTRODUCTION

This paper differs from some of the other papers prepared for this conference in some respects. This paper is not explicitly about the German chemical industry. Rather than an in-depth study of a particular firm, or even a handful of firms, we describe here the evolution of industry structure in Western Europe, US, and Japan. However, even though the reader will not find specific and detailed information about the German chemical companies, much about the German industry can be understood and appreciated by studying how the chemical industry in the developed world has evolved. Our interest is as much in explaining what we describe as in describing it. We have used a number of historical studies dealing with various aspects of the industry, and we have used them to inform ourselves and our analysis. In this paper, however, we do not much more than point to some of the historical forces that powered the evolution of industry structure.

To paraphrase Karl Marx, while firms may well shape their own destiny, they do not shape it in circumstances of their own choosing. Industry structure provides the context within which firms make their choices. The analytical lens we use is the dynamic tension between firm-level economies of scale and scope (e.g. Chandler, 1990), and industry-level economies of specialization (e.g. Young, 1928, Stigler, 1951). Chandler's historical account of the evolution of capitalism in the US, UK, and Germany emphasizes the importance of scale and scope at the level of the firm. While the chemical industry is a canonical example of firm level scale economies, it also provides evidence of advantages of bigness at the level of the industry which arise from an increase in specialization and division of labor.

A historical look at industry structure reveals reinforcing mechanism that tend to preserve and accentuate initial differences across countries. But there are forces that make for convergence as well. Technological innovations, the growth of international trade, and the growth of new markets are powerful forces that encourage greater convergence across regions. In fact, our analysis leads us to the conclusion that persistence and path dependence are likely to be more significant at the level of the firm than at the level of the industry.

We begin, in section 2, by describing the structure of the US, European, and Japanese chemical industries around the end of the 1980s. Section 3 discusses the relationship of these inter-country differences to historical factors, and to differences in size and other characteristics of the markets. Sections 4 and 5 look at dynamics factors. Section 4 examines the effects of the two world wars, anti-trust, and related institutions. Section 5 discusses the effects of technological advances on industry structure - notably polymer chemistry, chemical engineering, and the convergence between oil and chemicals. Section 6 examines how these and other forces are changing the industry structure in the recent past.

2. INDUSTRY STRUCTURE: A SNAP SHOT

2.1 All around vs. focused chemical companies

As a first cut, the chemical industry can be thought of as consisting of two main types of firms. There are "all around" companies, which are involved in different value chains, operate in several of those with large scales, and often occupy a number of stages of these value chains. In addition, there are many other companies which are more tightly "focused", and smaller. The focused companies are typically found downstream, where product specific features, such as brand name, and variety are more important.

To investigate the distribution of these different types of firms, we constructed a sample of leading US, Western European and Japanese chemical companies. These are the top five producers, ranked by production capacity, in each of the three regions in a sample of 123 chemical products (not including pharmaceuticals). ₁ , ₂ We divided these companies in three categories, which we labeled L2, L35, and L6. The L2 category consists of firms that are leaders (among the top five producers) in no more than two of the 123 products in our sample in their home region. The L35 companies are firms that are leaders in three to five products in their region, and L6 companies are leaders in six or more. The L6 category includes the largest, all around, chemical corporations (e.g., Dow, Du Pont, ICI, BASF, Bayer, and Hoechst).

Table 1 shows that although the distribution of firm types is similar across the three regions, Europe and Japan have a relatively higher number of large firms than the US. Table 2 shows the degree of internationalization of these companies. Two features of this table are worth noting. First, the vast majority of the L2 companies are leaders only in their home region. The focused companies are thus largely "domestic". Second, most of the US and European all around companies are multinational - they are market leaders in a number of products outside their home region. By contrast, 23 of the 25 Japanese L6 companies are leaders only in Japan.

Table 3 further explores the characteristics of the largest chemical firms. It lists the largest all around firms from Europe, US, and Japan, ranked by chemical sales in 1988. The last column shows the number of product markets (among the 123 in our sample) in which they are among the five leaders in their region. The relatively large number of European companies in this group reflects the fragmentation of the European market, with each country featuring one or more national chemical producers. Moreover, although the three leading all around Japanese producers are smaller than the European or US firms, they operate in roughly the same number of products in the chemical industry. Put differently, Japanese chemical producers may be excessively diversified, when we restrict attention to the chemical industry.

While table 3 looks at the largest all around companies, table 4 lists the chemical companies among the top 50 in the world (ranked by chemical sales) that are typically associated with a major core business. These can be thought of as the largest focused chemical suppliers. Note that pharmaceuticals is the dominant industry among the largest specialized US and European firms. By contrast, the large specialized Japanese manufacturers have more varied backgrounds including textiles, and fertilizers. Moreover, unlike the Japanese "all around" companies, the Japanese "focused" companies are not much smaller than their European or US counterparts. In sum, the Japanese chemical industry appears to perform relatively better in the more specialized segments of the chemical industry vis-à-vis the more homogeneous commodity products. The same pattern is evident if one goes beyond the top 50 companies in the world.

Table 4 also shows that, unlike the all around companies, the number (and size) of the large specialized US producers is similar to that of the European firms. Moreover, all the large focused European companies come from Germany, Switzerland and the UK (see also Aftalion, 1989, Appendix). The evidence suggests that, compared to Western Europe, the US is more conducive to the growth of more specialized firms. Thus, for instance, while many all around European chemical companies also operate in pharmaceuticals (e.g., Bayer, Hoechst, and till recently, ICI), in the US the chemical and pharmaceutical businesses tend to be the domain of distinct firms. Tables 5 & 6 reveal that in a number of basic intermediates and large volume products, Western Europe as a whole tends to have a larger number of producers than the US. However, both the average capacity per producer, and average capacity per plant, are larger in the US than in Western Europe. Once again, the largest contrast is between the West as whole and Japan. Japanese firms tend to produce smaller volumes using smaller plants than in the West.

Industry structures are remarkably similar across the US and Western Europe, and only somewhat different from Japan. The large all around Japanese firms are smaller than their counterparts in the US and Western Europe. They are less international in orientation, but display the same degree of diversification. The picture is different when we look at the more specialized firms. All regions have a number of specialized firms of comparable size and internationalization. While in the US and Western Europe, these large specialized firms are predominately life sciences based companies (along with a few other specialized sectors, like rubber or consumer products), in Japan they have a varied background. Europe tends to have fewer large specialized firms, most of them from Germany and the UK. Western Europe and Japan have a greater number of producers, and a larger number of plants with smaller average production capacity. The problem appears to be especially severe in Japan, wherein the manufacture of a sample of products that we examined is fragmented across several companies, each of them producing relatively small volumes (Arora and Gambardella, 1998).

In the last decade or so, developments that we describe at the end of the paper have changed many of the firm boundaries, so that lists analogous to those in tables 1-4 constructed for 1997 would look somewhat different, especially for Western Europe. They would include some new names reflecting some large mergers and demergers, such as Montell (a Montedison-Shell joint venture in polyolefins), Novartis (the merged life-science businesses of Sandoz and Ciba Geigy), Zeneca (the life sciences businesses of ICI), Borealis (polyolefins business of Statoil and Neste Oy), and Marlene (polyolefins businesses of BASF and Shell). Virtually all of these large companies were fashioned from existing businesses in the last couple of years. And if one were to construct such lists, the net result will be to make the European landscape appear even more similar to that of the US.

3. INTER-COUNTRY DIFFERENCES:

3.1 Differences in historical origins

The similarity in industrial structures that we show is perhaps surprising. The historical conditions under which the chemical industry arose have differed across countries. Moreover, it is now a commonplace that path dependence and similar phenomena reinforcing initial differences are important in industry evolution. Indeed, a more qualitative look reveals some interesting patterns of persistence. One also sees some interesting and subtle differences between Europe and the US, but the most prominent differences are with Japan.

3.1.1 Japan: Late Industrialization

The Japanese chemical industry developed later, when the world chemical industry was already dominated by Europe and the US. As one might expect, Japanese technological capabilities in chemicals were far less advanced than in the West during the first half of the century. The late industrialization had several important aspects. First, Japanese firms have extensively imported and used technology developed elsewhere. Second, Japanese chemical firms have had to face competition from European and US firms that had far greater technological capabilities. Finally, the Japanese economy as a whole developed behind trade barriers. All three features left their mark on structure of the Japanese industry.

• The role of users

The availability of imported technology, and the general backward state of the chemical industry itself enabled users - firms in downstream sectors - to play a more prominent role in the chemical industry. While firms from downstream sectors are not unknown in the chemical industries in other regions (e.g. General Electric in plastics, Eastman Kodak) the relative technological weakness of the chemical firms in Japan, and the business group structure (sometimes referred to as keiretsu, although this usage may not always be appropriate) increased the relative importance of companies coming from user industries. For instance, synthetic fiber production, particularly rayon, was dominated by textile producers, such as Kanebo, Teijin, and Toray. By contrast, the major rayon producers in the West included chemical firms like Du Pont, ICI, and Hoechst. During World War I, many textile companies started producing organic dyestuffs after imports of German dyestuffs were disrupted due to the war. Although most failed, some such as Yura Seiko survived. Similarly, nitrogenous fertilizer companies in Japan integrated backwards into the production of chemicals such as ammonia (Molony, 1990: 35-36).

After World War II, many of the downstream user firms further integrated into petrochemicals, largely for internal use. Thus, Showa Denko, originally into fertilizers, diversified into various polyolefins; Daicel produced acetic acid and other derivatives to be used in its cellulose business; and textile companies like Toray, Teijin, and Kanebo made similar moves to integrate backwards (Spitz, 1988: 375-384; Aftalion, 1989: 205). Other Japanese companies, like Kao (detergents), or Takeda Pharmaceuticals, integrated backward to acquire the in-house chemical capabilities that were increasingly necessary to run their businesses.

Many large petrochemical producers were themselves part of larger group of companies, or industrial complex, which included firms operating in downstream sectors such as electronics, and automobiles. Chemical companies belonging to a group sold a substantial fraction of their output to downstream companies in the same group, rather than in the open market. The influence exercised by the users is likely to have exaggerated the search for detailed specifications of chemical products to the somewhat idiosyncratic requirements of special applications. This contributed to the high degree of product customization that exists in Japan. For instance, the number of grades of plastic resins is said to be an order of magnitude larger than in Europe or the US, and Japanese automobile companies are credited (or blamed) for inducing such high degree of product differentiation.

• Protection, technology imports, and market fragmentation

Import protection provided incentives for domestic firms to diversity into the production of a wide range of products. During the 1950s and the 1960s Japan made an astonishingly rapid entry into petrochemicals. Apart from the three main keiretsu - Mitsui, Mitsubishi, and Sumitomo - several other companies, such as Asahi Chemical, Maruzen Oil, Idemitsu, made considerable investments in petrochemical plants (see Hikino et al., 1998). Each keiretsu attempted some measure of self-sufficiency in the basic chemical inputs. The keiretsu structure therefore exacerbated the tendency, caused by import protection, towards production at scales that were too small to be economic. Thus, while international technology trade and other factors had made some features of the Japanese industry more similar to other countries (e.g. the wide use of petrochemical technologies after World War II), as noted earlier, important differences remain.

3.1.2 Historical origins in the US, Britain, and Germany

Like Japan, the US was also a "follower" in the chemical industry in the sense that till well into the first part of the 20th century, it relied upon technology imports and had periods of tariff protection. One important difference however, lies in the size of market. The large size of the US market implied that, with the growth of the US economy during the 1920s, firms could specialize without sacrificing size. As Arora and Rosenberg (1998) show, a characteristic of the US industry was the application of technology to exploit the natural resource base, and produce on a large scale. This is reflected in the structure of the US chemical industry - in the large size of firms, in their willingness to invest in the commercial application of technology, in their attention to manufacturing costs, and in the extensive division of labor in the chemical industry.

The industrial structure also reflects other aspects of the US growth experience. For instance, unlike Japan, technology was often transferred through subsidiaries of foreign firms, as well as through scientists and engineers who immigrated. Sectors such as fine chemicals, pharmaceuticals, and dyestuffs relied heavily upon European technology, and the subsidiaries of German and Swiss dyestuff and pharmaceutical producers played an important role. Over time, and particularly after World War I, many of these subsidiaries disassociated themselves from their parent companies and grew into full fledged chemical companies (e.g. Schering Plough, Merck, and Rohm and Haas). One result of this trend, which is evident even today, is the fairly clear separation between chemical and pharmaceutical firms in the United States.

The origins of the first chemical industry in the world, Britain, were largely inorganic. The production of inorganic chemicals such as soda ash, soda, and bleach grew because of the opportunities created by the growing demand for such products by the textile industry, and by the supplies of natural materials from its dominions. Inorganic materials are simple, and even production processes are far less complicated than those for producing organic chemicals. With limited opportunities for product innovation, and more generally, for the application of science to chemical production, British chemical firms (with some notable exceptions in later periods) de-emphasized investments in science and technology, as well as in complementary activities such as commercialization. Instead, firms focused on the volume of output, sales and distribution, because profits, and growth came, from new markets, including overseas markets, cheaper sources for inputs, and control over pricing. Many of these strategies proved to be poorly suited to managing in the organic chemicals industry, especially in periods of rapid technological advance.

Dyestuff represented the engine of growth of the German chemical industry till World War I. The German chemical industry grew through the systematic applications of scientific discoveries to chemical manufacture. Advances in organic chemistry clarified how carbon atoms are linked to hydrogen and other atoms to form more complex molecules. These discoveries proved to be a general purpose "method" for developing new products, such as alizarin and indigo. In addition, it was soon realized that the method provided by organic chemistry could be extended to other products, like pharmaceuticals, explosives, and photographic materials.

The dyestuff "model" displayed several features that have characterized the German chemical industry for many years. In the first place, companies had to invest in long and systematic research and development to find new products that had desired characteristics. Whether a given molecule worked or not in practice could only be established by careful and methodical tests, and patient modifications of basic compounds. Moreover, the dyestuff companies undertook large investments in marketing. These were necessary to find the special and distinct grades of products that were needed by the users. ₃ Thus, the origin of the German industry laid the basis for the creation of a sector that had considerable skills in applied chemistry, and was skilled in applying chemistry to commerce. As a result, German companies accumulated competencies that represented an important source of competitive advantage.

Since the new method research and development had large fixed costs, German chemical companies expanded abroad soon after being established. This early globalization started with the end of the 19th century, and it has persisted for over a century. Vertical integration and horizontal mergers also took place early in the history of the German industry. Among other things, extensive vertical integration up to the most basic materials for chemical production was necessary to maintain proprietary control of the know-how about the production process that each dyestuffs manufacturer used. As the early producers started by integrating backward, there was little room left for independent suppliers. Newcomers, in order to survive, had to vertically integrate as well. In turn, this meant that when pressures arose to achieve greater economies of production, and to moderate competition, the solution took the form of horizontal integration, of which the IG Farben is the best known example.

3.2 Differences in the size and nature of demand

The importance of cartels and other combines in pre world war II Europe has been well documented. But even without cartels, the smaller and segmented European markets exaggerated the influence of the large firms, effectively focusing the other firms on niche markets. The rise of the German chemical industry till World War I, enabled Bayer, BASF, and Hoechst to grow and dominate the chemical landscape. This dominance was reinforced by the formation of IG Farben in 1926. The new trust became the dominant German firm, with activities in a wide number of organic and inorganic products. The second largest German chemical firm in that period, Deutsche Solvay Werke, was ten times smaller. While IG Farben dominated a variety of organic chemicals and metals, explosives were produced by Dyanamit AG, and soda and other inorganics by Deutsche Solvay. This separation was mirrored within IG Farben itself, and the lines of business of the companies that had formed the conglomerate remained fairly distinct.

In Britain, ICI acquired a dominant position, and it was a domestic monopolist in a large number of products. Other companies like Albright & Wilson, Laporte, and British Oxygen focused on narrower businesses, such as phosphorous and metals, hydrogen peroxide, and industrial gases. ICI, however, at least till the end of the World War II, was careful not to enter into more downstream markets, out of a reluctance to compete with its customers. This enabled the companies in those areas, most notably Courtaulds in fibers and the Anglo-Dutch company Unilever in consumer products, to develop their chemical operations. As in the US, the industry also gave rise from very early on to a specialized pharmaceutical sector, which was operated by companies like Boots, British Drug Beecham, and Glaxo.

The US market also allowed the formation and growth of many large chemical firms. An integrated domestic market however implied that firms could not find protected niches where they found shelter from competition. As well, the large size of the US market enabled firms specialize more narrowly because product specific investments could be amortized over a large volume of output. This gave the US industry a greater variety in firm types, their competencies, and strategies. The failure of a single firm to take advantage of emerging opportunities, even one as large as Allied Chemicals, did not affect the US industry in the same way as the weaknesses of ICI affected the British industry. ₄ This was an important advantage that tends to be overlooked in discussions of the advantages of large markets that focus on bigness at the level of the firm alone. It is difficult to assess the extent to which this was important in Germany because the period of IG Farben dominance overlaps heavily with the period where IG Farben energies were closely guided by the dictates of German rearmament.

4. DYNAMICS OF INDUSTRY STRUCTURE:

4.1 World Wars

World War I spurred nationalistic and autarkic concems, which in turn led many countries to promote the development of domestic capabilities in areas that were previously served by international trade. The war also meant massive government intervention in the form of a large demand for chemical products and technologies. Finally, the war effort encouraged the dissemination of technological knowledge, as governments encouraged domestic chemical companies to pool their efforts for military purposes. At the international level, the release of German patents and technologies into the public domain created a similar inter-country diffusion of technology. In 1914 Germany was a virtual international monopoly in synthetic dyes, controlling 90% of the world production. The US and Britain expropriated German industrial property, including extremely valuable patents on ammonia and dyestuffs, thereby denying the Germans an important means of deterring entry. Greater technological diffusion had a "leveling" effect among major firms within and across countries, with greater international competition in sectors such as inorganics, explosives, dyestuffs, and even ammonia and nitrogenous fertilizers.

The Second World War had effects similar to those of World War I. In Germany the Nazi government pushed IG Farben into intensive programs like production of synthetic rubber, and synthetic gasoline from coal. The US government contributed to radical changes in the chemical industry by launching a large program for research and production of synthetic rubber, and by creating massive demand for oil needed as aviation fuel. ₅ Apart from the impetus to innovation, these and other war time programs (such as the Manhattan Project) explicitly promoted co-operation among many domestic companies. The rubber program involved information sharing, co-ordination in research efforts, as well as wide downstream applications of generalized technological principles and capabilities. In addition, at the end of the war, plants and other facilities managed under the Rubber Program were sold out to different companies, whether oil (e.g. Exxon) or chemical producers (e.g. Dow, Monsanto, Union Carbide).

4.2 Switch to oil after World War II

While wars have been very successful in shaking the industry free from the grip of history, so has the development of new markets. The switch from coal to oil based feedstocks was rapid, even by countries such as Germany where firms had made large, irretrievable (sunk) investments in coal related technologies.

The switch to oil began in the US, which had abundant oil and natural gas reserves. By 1950 half of the total US production of organic chemicals was based on natural gas and oil, by 1960 the proportion was 88%. The switch came later, but as rapidly, in Western Europe. In the United Kingdom, in 1949, only 9% of total organic chemical production was based on oil and natural gas but rose to 63% by 1962 (Chapman, 1991: 82). In Germany, the first petrochemical plant was set up in the mid-1950s, and by 1973 German companies derived 90% of their chemical feedstocks from oil. Given its limited endowment of oil and gas, and its accumulated technological expertise in coal, this appears to be a significant reverse of initial conditions. (See also Stokes, 1994.)

The discovery of large oil reserves in the middle East, which lowered both prices and fears of an impending oil shortage, was critical in inducing firms (and countries) to switch. Moreover, technology for refining crude oil, and for producing a variety of chemical intermediates, became widely available, as a market for such process technology developed (see below). One must note, however, that absent the development of a world market in crude oil, many countries would have hesitated to switch to a commodity whose supply was controlled by a few countries. The US guarantee of unhindered oil supply to Western Europe was also very important in this respect.

5. TECHNOLOGY

Technological advances have had a profound impact on the evolution of the industry. This section discusses two classes of technological innovations which have been crucial for understanding the evolution of industry structure in chemicals: polymer chemistry and chemical engineering. ₆

5.1 Polymer chemistry: "Materials by design"

Polymer chemistry was really the start of "materials by design". It provided the industry with a method of developing new products by modifying the type of molecules and the way in which they were connected. Producing commercially successful new materials was not simple. Long and systematic experimentation was still needed for commercial innovations. However, the new theoretical knowledge of the relationship between molecular structure and physical properties helped in making this search more productive.

Two characteristics of polymer chemistry are worth noting here. First, polymer science facilitated the development of many different applications of given substances. This produced knowledge based economies of scope, thereby creating linkages among hitherto distinct and unconnected markets. Second, because applications had to be adapted for specific uses, the new science shifted the problem of innovation from "how" to produce different products to "what" to produce. Companies had to decide which applications were to be developed among the many that could be produced. This increased the relative importance of marketing and downstream links with users to find out how to tailor products for their needs (e.g., Hounshell, 1995).

These features of these new technological opportunities were not qualitatively different from those opened up in dyestuffs, photographic materials, and pharmaceuticals by developments in organic chemistry. The knowledge-based economies of scope prompted expansion of the three leading German dye firms, from dyes to pharmaceuticals, and photographic materials. The "dyestuff model" also required extensive investments in marketing to find out "what" users wanted, and, more importantly, to train them in how to use the new products. Both technologies exemplify the economic potential of generalized knowledge bases, and the economies of scope that can be created by abstract conceptualizations (Arora and Gambardella, 1994b). The striking aspect of such generalized knowledge bases is that the underlying technology generates opportunities that link markets which are disconnected when viewed from the point of the user industry. Ultimately, polymer chemistry supplied a common technological basis in three quite different areas - plastics, fibers, and rubber. Fibers required extensive laboratory experimentation to define the organic chemical building blocks to obtain synthetic products that could mimic natural fibers. Fibers and plastics are used very differently, require different types of processing equipment and machinery, and therefore, must be sold differently. In other words, downstream development and commercialization capabilities had far more limited economies of scope than the basic polymer technology itself.

The main difference was that German companies moved first to take advantage of the opportunities created by the developments in organic chemistry. By contrast, when polymer chemistry arose, there were many large, technologically advanced companies in the world. The German success had led to a much greater appreciation of the importance of scientific capabilities for commercial objectives. The events of World War I, and the rationalizations of the 1920s had created large sized firms such as ICI in Britain, Montecatini in Italy, Solvay in Belgium, and Akzo in the Netherlands, in addition to the well known chemical companies, such as Monsanto and Dow, in the US. Competition led to the rapid development of polymer science and technology. Competition also reinforced the search for product differentiation as a way of relaxing price competition, encouraging additional investments in R & D to develop new products or variants of existing ones.

As with synthetic dyestuffs, the rise of polymer chemistry opened up new opportunities in the industry. But unlike dyestuffs, these opportunities were now exploited by a substantial number of large firms worldwide, with comparable commercial and technological capabilities. As Freeman (1982) points out, the presence of a large number of firms with comparable capabilities in polymers implied that even "small" information leaks allowed very rapid imitation. Thus, many chemical companies and some oil producers found themselves operating and competing in very similar markets. For instance, Union Carbide, Goodrich, General Electric, IG Farben and ICI performed research on PVC and produced the polymer since the very beginning of this business. Similarly, Dow, IG Farben and Monsanto were all involved in polystyrene from very early on. Du Pont, ICI, UCC, Monsanto, Kodak, and many others invested in various kinds of polyamides, acrylics, and polyesters. (See Spitz, 1988, and Aftalion, 1989.) The net result was an increase in competition in virtually every market segment.

5.2 Chemical engineering and the specialized engineering firms

If polymer chemistry is the science of chemical products, chemical engineering is the science of the chemical process. More appropriately, chemical engineering was the science and economics of the chemical process. By separating the task of process design from the details of the particular product being produced, chemical engineering freed process engineers to the think about chemical processes in general. By providing a unified framework, chemical engineering allowed for the experience gained in one petrochemical process to be applied to others, further enhancing the benefits to specializing. As a result, chemical engineering made possible the rise of specialized firms that focused on engineering and process design services for chemical plants - the so-called specialized engineering firms (SEFs).

The development of an independent engineering design sector is an example, par excellence, of what we mean by economies of specialization at the level of the industry. A large market for basic petrochemicals plus the relative independence of process design from products, implied that SEFs could design many more plants for a variety of closely related processes than any single chemical company could. The first SEFs were formed in the US, early in this century (Arora and Gambardella, 1997). The large oil companies, which concentrated their energies upon ".. searching for crude oil and establishing retail market facilities...", were amongst the first set of clients for SEFs (Landau and Brown, 1965: 7). In the years before World War II, the chemical companies did not rely much upon SEFs for the design and engineering of entire production processes. SEFs were mostly employed as suppliers of specialized equipment and the like. In part, this was due to the long standing traditions of secrecy that characterized chemical companies. Further, most chemical operations tended to be batch processes, with relatively low volumes, and with strong elements of "art", embodying a great deal of company specific know-how.

The situation changed markedly after World War II. The growing importance of petrochemicals, and the increase in the scale of production, raised the payoff to improvements in plant design. In turn, the growth in the size and complexity of plants, as well as the concomitant development of chemical engineering laid the foundations for division of labor and vertical specialization in the chemical processing sector. By the 1960s, SEFs had come to occupy an important place in the industry. In a pioneering study, Freeman noted that for the period 1960-66, "... nearly three quarters of the major new plants were `engineered', procured and constructed by specialist plant contractors" (Freeman, 1968:30). Moreover, Freeman finds that SEFs were an important source of process technologies. During 1960-66, they accounted as a group for about 30% of all licenses of chemical processes. Freeman's findings are confirmed by more recent data which show that, for the period 1980-1990, almost three-fourths of the total number of plants in the world were engineered by SEFs.. Although SEFs are not sources of major innovations, they account for about 35% of the licenses (Arora and Gambardella, 1998).

The role of SEFs in diffusing basic petrochemical and refining technology the world over has been insufficiently appreciated. The importance is two fold. First, they enabled European countries to rapidly catch up with the US in these technologies. Companies such as Scientific Design, UOP, Kellogg and others transferred technology and design know-how to European chemical companies, facilitating their shift from small volume, batch based production techniques, to high volume, continuous flow production systems. This was critical in enabling European chemical companies to shift from coal to naphtha based feedstocks.

Second, SEFs facilitated a great deal of entry in the post war decades, initially from the developed countries themselves, but since the `70s, from developing countries enter as well. In addition, by acting as independent licensers, SEFs also induced chemical firms themselves to license their technology. In essence, SEFs helped create a market for technology, making process technology into a "commodity" which could be bought and sold. The implications for industrial structure have been profound. Spitz (1988: 313) notes that in most major products, the main producers were between five and fifteen. By contrast, in the pre- World War II era, it was unusual to have more than three manufacturers (See also Backman, 1964: 47-50). Similarly, in a study of 39 commodity chemicals in the US in a period from the mid `50s to the mid `70s, Lieberman (1989) entry into concentrated markets, which were also marked by low rates of patenting by non-producers (both foreign firms and SEFs), usually required that the entrant develop its own technology. By contrast, less concentrated markets were associated with high rates of patenting by non-producers and high rates of licensing to entrants.

6. THE INDUSTRY IN THE 1980S AND AFTER

6.1 Maturation, competition, and restructuring

An important trend encouraged by the developments discussed in the previous sections was a substantial increase in the number of firms in most markets. During the 1950s and the 1960s, the industry could accommodate such increases because demand grew rapidly - typically two or three times faster than GDP. Even so, observers suggest that profitability in the chemical sector already began to decline in the early 1960s (Chapman, 1991: 234-6). The problem became severe when demand growth slowed in the 1970s and the 1980s, growing at the same rate as GDP. The problems were exacerbated by the growing convergence between oil refining and basic petrochemicals, as a result of which, several oil companies had installed substantial capacity in basic petrochemicals. Entry and competition from firms in developing countries, including the Eastern European countries, aggravated the problem.

The slowdown in the growth of demand, and the rise of new sources of supply, prompted a move toward a new long-run equilibrium. But the adjustment to the new equilibrium was slow and painful. Economies of large scale production meant that existing producers had sunk large investments in capacity, especially in the basic intermediates. The problem was magnified because many chemical and petrochemical operations were highly integrated (both vertically and horizontally). To make matters worse, not all firms accurately foresaw the slow growth of demand. Despite the oil shocks, many companies continued to invest as if they were expecting to return to the high growth rates of the 1960s. A comprehensive re-alignment of expectations was completed only during the 1981-1982 recession. (See Aftalion, 1989; Chapman, 1991; Lane, 1993.)

The pressures for capacity rationalization were not identical across the US, Western Europe, and Japan. The problem of large installed capacities and rosy expectations about future demand growth were common to all three areas but the threat posed by new producers was mainly a European and, to a lesser extent, a Japanese problem, because of greater export dependence (Grant, 1991: 255). In addition, Europe had to cope with the very special problem of its many large and diversified national companies which were much more vulnerable to pressures from stakeholders and employees. Capacity reduction and rationalization proceeded in different ways in the US, Europe, and Japan. In the US the reductions were market driven and without explicit co-ordination. In Europe both market forces and government intervention (especially in Italy and France where the state had large ownership stakes) played a role (Martinelli, 1991). In Japan, MITI played a major role by coordinating capacity rationalizations. Its Industrial Structure Council formed a committee aimed at establishing consensus behind main policy objectives (Bower, 1986: 202). In all cases, the prescription was clear. Many of the firms had to move away from petrochemicals to focus on high value added products, particularly new materials.

• Capacity reductions and plant shut downs

Capacity reductions through plant shut downs started in the early 1980s, and the bulk of it was finished before 1984-5. Lane (1993) points out that in the US, restructuring via unilateral capacity reductions preceded changes in ownership of plants. This suggests that US chemical companies responded via internal rationalization before attempting consolidation at the level of the industry. Capacity reductions and plant shut-downs took place in Europe as well. Both in Britain and Germany a number of plants were closed during 1981-2, and the exit of many US companies from the continent helped as well (Chapman, 1991: 250).

In Europe restructuring also involved interfirm agreements. Such agreements played a critical role in the restructuring of the PVC market. ₇ In addition, in Europe, industry associations fulfilled an important function. In 1985, thirty-four large petrochemical firms formed the Association of Petrochemical Producers in Europe (APPE). APPE did not have any formal role in encouraging agreements. However, by collecting and providing data on capacities in different product markets, it diffused information about the industry among its members, helping them take informed decision about capacity expansions and reductions. The capacity rationalization in the chemical industry is said to have been far more successful than in other sectors, such as steel (Grant, 1991: 265). In some cases, European firms tried to form cartel, but with limited success. The number of producers had increased considerably over time, and made enforcing agreements difficult. In addition, the pro-competitive attitude of the EC restrained such deals. ₈ Through a variety of ways, the problems of overcapacity were largely resolved by the second half of the 1980s.

Table 7 documents the effects of the ongoing restructuring process in selected petrochemical products for the US and Europe. As the table shows, the number of producers fell more sharply in Europe, while concentration of capacity appears to be more substantial in the US. This is consistent with the view that restructuring in the US took the form of reduction of capacity by the largest producers. Moreover, the entry of specialized commodity manufacturers (discussed below) both reduced concentration and kept the number of producers from falling too far. In Europe, instead, the figures reflect partly the exit of US firms, and partly the consolidations carried out through swaps and interfirm agreements. While we lack systematic evidence for Japan, the available evidence suggests that restructuring has not proceeded much beyond capacity reduction and some MITI brokered swaps to promote rationalisation (see MIT, 1989: 37, and Hikino, 1998).

6.2 Restructuring of the firm and the industry

There are two important aspects of the restructuring over and beyond capacity reduction. First, firms are increasingly looking outside national boundaries. The globalization of the industry involves moving production closer to the emerging markets in Asia and East Europe. Second, and perhaps more interesting, firms have reversed a long trend towards diversification and have narrowed their business portfolios. The winnowing of portfolios is often guided by a strategic intent to specialize. While some firms are focusing on commodity chemicals, others are focusing upon the so called "specialty" sectors - sectors marked by greater possibilities of product differentiation, less acute price competition, greater service content, higher margins, and often though not always, lower volumes. The focusing also reflects the sharply reduced perceived synergies between life-science and chemical businesses.

• Restructuring: Mergers and acquisitions

To examine these issues systematically, we analyzed data on acquisitions and divestitures in the chemical sector in last ten years or so. The data we obtained cover deals in the industry (defined as SIC 28) with value exceeding 1 million dollars that were announced in the US between 1984-1992, and in Europe between 1987-1992. ₉ Table 8 displays the overall trends in the number of acquisitions in the three regions. While differences in coverage period do not allow us to draw firm conclusions about when restructuring starts in the different regions, the trends are consistent with the idea that restructuring began earlier in the US. Further, European and Japanese firms are more likely to be involved in cross-border deals than are US firms.

Financial deals have been particularly important in creating new companies which focused on commodity chemicals. In the US many such companies were founded in the 1980s, and most of them were created from acquisitions of plants sold by larger producers that were moving downstream. Financial investors, such as George Huntsman and Gordon Cain, played an important role in the formation of the new companies. Britain too has its "commodity specialists" in the form of the Hanson Group, which at one point also attempted a takeover of ICI. In the rest of Europe, however, the formation of new companies played a much smaller role, due to the very different role of stockmarkets in corporate governance. The difficulties were most marked when it came to LBOs and MBOs. Instead, the refocusing took place via re-organization and splits of existing chemical producers. The recent rounds of privatization of large state owned companies in Italy, France, and Finland have been important in this respect. Thus, for instance, in France, Rhone-Poulenc developed a very aggressive strategy of acquisitions and divestitures to enter into a number of new specialty business. Orkem, another French company, was split into Elf, which focused on petrochemicals and fertilizers, and Total, producing inks, adhesives, acrylics, glass, and paints. There have been some intra-group mergers in $$Word$$ and some overseas acquisitions - e.g. Dainippon Inks has acquired companies in the pigments business in US; Mitsubishi has acquired Aristech.

• Relaxing price competition: Move to specialty chemicals

While firms such as Huntsman, Quantum, and Borealis focus upon commodity chemicals, chemical firms are trying to move downstream into more differentiated product lines. ₁₀ After War II, rapid economic growth and economies of scale encouraged strategies aimed at large famous markets rather than entry in several market niches. But once growth slowed down, and $$Word$$ of scale at the level of the plant became less important, companies tried to segment the $$Word$$ market niches. Although both product differentiation and cost reduction can $$Word$$ profits, the former has the advantage of relaxing price competition (see for instance, Sutton, 1991). Product differentiation, and creation of brand image is therefore a more attractive strategy for firms like Bayer or Hoechst that can produce higher quality, or more sophisticated products.

The Japanese adjustment to the changing circumstances was much easier in this respect. As noted in section 3, the important role played by users in Japan had resulted in an early emphasis on product quality, and customization of product to user requirements, even at the expense of production efficiency. But in the 1980s the close links with users turned out to be an asset. For instance, the Japanese chemical industry linked very early on with its growing electronics sector, and as a result, acquired a leading position in this field. Although US and European firms are also active in chemical applications to electronics, they have focused largely on the life science business. Japanese chemical producers instead have diversified to a greater extent, moving into the life sciences as well as into electronics. Moreover, the market for electronics in Japan is far larger than the traditional chemical ones, which is likely to have important consequences on the opportunities of the Japanese chemical manufacturers.

We examined the extent to which chemical producers used acquisitions to move into specialties by studying the acquisitions of companies with SIC codes that could be clearly identified as specialty chemical productions. (Our definition of specialty sectors is then somewhat conservative.) Table 9 shows the number of specialty and non-specialty acquisitions by nationality of the acquiring company. The table shows that most specialty targets are US companies, and acquisitions of specialty producers by European and Japanese companies are more likely to occur when the target is a US company. In the case of European firms, only 4.5% of their total acquisitions are of specialty companies. This percentage raises to 9.6% in the case of US targets. ₁₁

6.3 Restructuring: Increasing focus

The separation underway between specialty and commodity chemicals is a particular instance of the prevailing wisdom in the industry, which says that firms ought to concentrate on their "core competencies". To see whether firms are actually focusing, we used a sample of the largest 250 chemical firms in the world. ₁₂ Of these, 153 are covered by our mergers and acquisition database. They are predominately companies from the developed world, and particularly from Europe and the US.

Are acquisitions being used to increase focus? There are two aspects to this question. We first examine the portfolio of acquisitions and sell offs for our sample of large firms to see whether firms make most of their acquisitions in a few sectors, or whether these are spread across a range of sectors. Table 10 reports the difference between US and European firms in terms of the herfindal index of acquisitions, and the index for sell offs, after controlling for the size and diversity of the firm.. ₁₃ The results show that US firms sell on average a slightly wider variety of businesses than European firms, but do not differ significantly from the latter in the spread of their acquisition portfolios.

We also analyzed the degree of focus in physical investment. We used data on all investments in new plants in the chemical sector during 1980-90 from the Chemical Age Profile data base. There are 781 US firms, 648 West European firms, and 267 Japanese firms in this data base. As before, we also separately analyzed data for the large firms from each region (US, Japan, and Western Europe). The results are shown in table 11. Columns B and C of table 11 confirm our findings in section 2, namely that Japanese firms (including the large ones) tend to be smaller, and more domestic in focus than US, and especially, European firms. However, column D of table 11 shows that once we control for size, there is no significant difference in the degree of focus. The similarity in the degree of concentration in physical investment is consistent with the similarity in acquisition portfolios. Moreover, in keeping with our earlier results, the US and especially the European firms have very geographically diversified investments, as compared with Japanese firms.

• The search for size

An important aspect of the consolidation process in the commodity chemicals is the search for larger size. This trend is found not just in commodity chemicals such as polyolefins and PVC. Indeed, consolidation has been widespread in specialty products. The merger of Akzo and Nobel has created the largest paint company, and the second largest producer of catalysts in the world (Chemical Week, 1994b: 25). Similarly, ICI acquired Gliddens, and the Grow group to become the leader in the paints and coatings business. In agrochemicals, Hoechst and Schering merged their operations into Hoechst Schering AgEvo, while Bayer and Monsanto have a joint venture for R & D, and commercialization in pesticides. In catalysts, Cytec (ex American Cynamide chemicals) has a joint venture with Shell - Criterion Catalyst (Chemical Week, 1994c: 27). In more recent years, mergers among Western European chemical companies have become even more eyecatching. These include Montell (a Montedison-Shell joint venture in polyolefins that is among the largest in the world), Novartis (the merged life-science businesses of Sandoz and Ciba Geigy), Zeneca (the life sciences businesses of ICI), Borealis (polyolefins business of Statoil and Neste Oy), and Marlene (polyolefins businesses of BASF and Shell). Virtually all of these large companies were fashioned from existing businesses in the last couple of years.

In sum, through a series of divestiture, acquisitions, mergers, and alliances, firms are attempting to increase both the absolute size, as well as market share, of their remaining businesses. At the moment it is not clear whether the search is for volume, or for market share. While the former would result in greater efficiencies, the latter is largely about controlling price competition and market leadership. Consolidation can also enable a firm to manage the inter-brand competition more effectively, especially when faced with a large "competitive fringe" of producers. This appears to be an important motive behind the division swaps.

• Globalization

As noted at the beginning of this section, the growing globalization of the industry played a role in initiating the restructuring that we have described. The key elements of firm restructuring, increasing focus on core businesses, and the search for size in individual markets, are complementary to a third element - global operations. Globalization is complementary to the search for size. It provides a way of increasing the size of a given business through geographical expansion. European firms, especially German ones, moved into the United States in the 1980s, and made a number of acquisitions as well. The economic liberalization in a number of countries in Asia and Latin America, and in Eastern Europe, and their emergence as major markets, attracted investment from the leading chemical companies in Japan and the West. ₁₄ Apart form market expansion, firms have also used their international investments as a part of global sourcing. India and China have recently emerged as the leading sources of fine chemicals, because of lower labor costs, and less stringent environmental regulations. For instance, Hoechst has decided to outsource much of its fine chemicals production, including chemicals such as anthraquinone, for South East Asia to its Indian affiliate.

There is no doubt that taken as a whole, the chemical industry is truly a global industry. As noted earlier in section 2, Western European firms were the first to move overseas. US firms have tended to be a bit slower, mainly on account of the large domestic market. Japanese firms have tended to lag both. Indeed, if we examine the median index of geographical concentration (table 11 column E), the same picture emerges. Large European firms are the most globalized (least geographically concentrated in their investments) followed by the US, and Japanese firms, in that order.

6.4 Industry maturity and the market for corporate control

The entry of new firms, such as Huntsman, and Aristech in the US during the 1980s is noteworthy because since the 1920s, entry has typically consisted of diversification by existing firms rather than entry by new ones. The phenomenon is especially intriguing because the new entrants were buying assets that had become unprofitable to incumbents, and for which market prospects looked dim. How could it be that they expected to do better than firms that had extensive experience in the business?

We do not have a complete answer. One likely explanation is related to the need for clearer separation between different types of products. Put simply, the managements of many of the existing chemical companies were paying inadequate attention to their commodity chemicals, in part because of the rapid growth and diversification activities of the 1950s and 1960s (see also Chandler, 1994). In diversified companies, less attention is paid to the management of individual product lines, and to managerial incentives for longer run product performance. One way, therefore, to interpret the drive for focus is that the management structures of existing chemical companies were better suited to the production of higher value added, technology intensive products. Leveraged, and management buyouts, and similar financial strategies, are a means for correcting the mismatch. These instruments require a liquid and broad based equities market, such as the one in the US. Indeed, as table 12 shows, such deals have been relatively more important in the restructuring process in the US. The fraction of such deals involving US acquirers is more than twice the figure for deals involving European acquirers, although the absolute percentages are small. (See also Hall, 1994.) At the same time, many of the deals involving European acquirers have US targets. Moreover, the influence that these deals have exercised is far greater than is suggested by their number. Two prominent cases come to mind, both unsuccessful, but both very influential: In the US, the abortive takeover of Union Carbide (UCC) by GAF, and in Britain, that of ICI by Hanson. In both instances, the companies under threat responded by selling off a variety of businesses, and becoming more focused. ₁₅

Will the more advanced US financial markets then stimulate US firms to become more efficient, and cause industry structure to differ from that in Europe or Japan? (See Richards, 1998, and Darin, 1998.) This appears unlikely for two reasons. First, as we have seen, European chemical industries are undergoing a restructuring similar to that in the US, even without the threat of hostile takeovers. Second, the world financial markets are becoming more integrated, and by the same token, large companies have to raise capital on world market. Even the large German chemical companies are making preparations to be listed on the New York exchange. While this does not guarantee that the management will be quite as responsive to short term movements in their stock price as their American counterparts, at the very least, they will become more responsive. ₁₆

This is not to argue that the greater attention to financial indicators (such as the stock price) will necessarily lead to greater long run efficiency. In some instances, the purchase and sales of businesses that such financial strategies dictate may not have any relationship with greater economic efficiency. But one thing is clear. As a technology intensive industry matures, the problems that management must solve change, and hence so must the organization of the firm and the industry. Whether the stock market is the most efficient way to bring about the necessary changes is a separate matter.

In Germany, and in Western Europe more generally, the restructuring has been slower, more deliberate, and under greater managerial control, than in the US. It has also been slower until now. But the more recent quickening of pace suggests that when industries mature, and innovation opportunities decline that cost reductions become more important for profitability and shareholders can more meaningfully monitor the performance of the managers in the short run. But this also means that firms may have to separate the mature segments from the newer ones, because the two require different management structures. The current restructuring in the industry still has some way to go in this respect.

References -- Books and articles

Aftalion, F., 1989, History of the International Chemical Industry, University of Pennsylvania Press, Philadelphia.

Arora, A., 1997, "Patent, Licensing and Market Structure in the Chemical Industry", forthcoming, Research Policy.

Arora, A. and Gambardella, A., 1994, "The changing technology of technological change: General and abstract knowledge and the division of innovative labour" Research Policy, 23, 523-532.

Arora, A. and Gambardella, A., 1997, "Domestic Markets and International Competitiveness", forthcoming, Strategic Management Journal

Arora, A. and Rosenberg, N., 1998, in Arora, A., Landau, R., and Rosenberg, N. (eds.) Strategies for Growth: Lessons from the Chemical Industry, forthcoming., John Wiley and Sons.

Backman, J., 1964, Competition in the Chemical Industry, Manufacturing Chemists' Association, Washington DC.

Bower, J.L., 1986, When Markets Quake: The Management Challenge of Restructuring Industry, Harvard Business School Press, Boston.

Chandler, A., 1990, Scale and Scope, Harvard University Press, Cambridge.

Chandler, A., 1994, "The Competitive Performance of US Industrial Enterprises since the Second World War", Business History Review, Vol. 68, 1-72.

Chandler, A., Mowery, D. and Hikino, T., 1998, in Arora, A., Landau, R., and Rosenberg, N. (eds.) Strategies for Growth: Lessons from the Chemical Industry, forthcoming.

Chapman, K. 1991 The International Petrochemical Industry, Basil Blackwell, Oxford.

Cohen, W. and Klepper, S., 1992, Anatomy of firm R & D, American Economic Review.

Da Rin, M., 1998, in Arora, A., Landau, R., and Rosenberg, N. (eds.) Strategies for Growth: Lessons from the Chemical Industry, forthcoming.

Freeman, C., 1968, "Chemical Process Plant: Innovation and the World Market", National Institute Economic Review, No. 45 (August), 29-51.

Grant, W., 1991, "The Overcapacity Crisis in the West European Petrochemicals Industry", in Martinelli, A. (ed.), International Markets and Global Firms, Sage Publications, London.

Hall, B., 1994, "Corporate Restructuring and Investment Horizons in the United States, 1976-1987", Business History Review, Vol. 68, 110-143.

Harada, T., 1995, "Institutionalization of Technical Change and Paradox of Technology Transfer: The Japanese Chemical Industry 1910-45", mimeo, Stanford University, Stanford.

Hikino, T, Tokuhisa, Y., and Harada, T., 1998, in Arora, A., Landau, R., and Rosenberg, N. (eds.) Strategies for Growth: Lessons from the Chemical Industry, forthcoming.

Hirooka, M. and Hagiwara, T., 19??, "Present Status of Japanese Chemical Industry: Competitiveness and Diversification", ????

Hounshell, D. 1995 "The Dialectics of New Materials: On the relationship between natural fibers, new knowledge, and synthetic fibers, C.1840-C.1960", in Geison, Gerald L. ed. The Dialectics of Materials, Princeton University Press.

Hounshell, D.A., and Smith, J.K., 1988, Science and Strategy: Du Pont R & D, 1902-1980, Cambridge University Press.

Landau, R., 1966, The Chemical Plant. From Process Selection to Commercial Operation, Reinhold Publishing Co., New York.

Landau, R. and Brown, D., 1965, "Making Research Pay", AIChE-I. Chem. E. Symposium Series No.7, London Institute of Chemical Engineers, London, 7:35-7:43.

Lane, S., 1993, "Corporate Restructuring in the Chemicals Industry", in Blair, M. (ed.) The Deal Decade, Brookings Institutions, Washington DC.

Lieberman, M., 1987, "Patents, Learning by Doing, and Market Structure in the Chemical Processing Industries", International Journal of Industrial Organization, 5, 257-276.

Lieberman, M., 1989, "The learning curve, technological barriers to entry, and competitive survival in the chemical processing industries" Strategic Management Journal, 10, 431-447.

Mansfield, E. et al., 1977, The Production and Application of New Industrial Technology, Norton & Company, New York.

Martinelli, A. (ed.), 1991, International Markets and Global Firms, Sage Publications, London.

Molony, B., 1990, Technology and Investment: The Prewar Japanese Chemical Industry, Harvard University Press, Cambridge.

Morris, Peter J.T., 1994, "Synthetic Rubber: Autarky and War" in Mossman, S.T.I. and Morris P.J.T. eds., The Development of Plastics, Royal Society of Chemistry, Cambridge UK.

Richards, A., 1998, in Arora, A., Landau, R., and Rosenberg, N. (eds.) Strategies for Growth: Lessons from the Chemical Industry, forthcoming.

Rosenberg, N., 1998, in Arora, A., Landau, R., and Rosenberg, N. (eds.) Strategies for Growth: Lessons from the Chemical Industry, forthcoming.

Spitz, P.H., 1988, Petrochemicals: The Rise of an Industry, John Wiley, New York.

Stigler, G., 1951, "The Division of Labor is Limited by the Extent of the Market", Journal of Political Economy, 59 (3), 190-201.

Stokes, R.G., 1994, Opting for Oil, Cambridge University Press, Cambridge.

Sutton, J., 1991, Sunk Costs and Market Structure, MIT Press, Cambridge.

Ward, M., 1992, "Japanese Chemicals: Past, Present, and Future", The Economist Intelligence Unit, June.

Wilson, G.D., 1994, "Polyethylene: The Early Years" in Mossman, S.T.I. and Morris P.J.T. eds., The Development of Plastics, Royal Society of Chemistry, Cambridge UK.

Wright, G., 1990, "The Origins of American Industrial Success", American Economic Review, 80(4), 651-668.

Young, A., 1928, "Increasing Returns and Economic Progress", Economic Journal, 38(152), 527-542.

References -- Trade Journals and Data Bases

Chemical Age Profile, 1991, Pergamon Financial Data Services, London.

Chemical Economics Handbook (CEH), 1991, SRI International, Menlo Park CA, June.

Chemical Week, 1994a, November 16.

Chemical Week, 1994b, July 6.

Chemical Week, 1994c, March 30.

Chemical Week, 1994d, April 27.

Chemical Week, 1994e, October 10.

Chemical Week, 1996, November 20

Hydrocarbon Processing Unit, 1991, HPI Data Base, Gulf Publishing, Texas.

IDD Information Data Base, on-line service from Lexus-Nexus.

Predicasts 1991, Company Thesaurus, Predicasts, Cleveland OH.

Table 1 Number of Chemical Companies by Type^*

Table 2 Internationalization of Chemical Companies^*

Table 3 Major European, US and Japanese All Around Chemical Companies

Table 4 Focused World Chemical Companies

Table 6: Number of Producers, Number of Plants, and Capacities by Region: LDPE, HDPE, LLDPE, Polypropylene (PP) PVC, Polystyrene (PS), Late 1980s

Table 7 Number of Producers and Capacities by Region: Major Fibers^*

Table 7 Producers, capacities, and concentrations for selected petrochemicals in Western Europe and the US.

Table 8 Trends in Acquisitions, 1985-1993, By Acquirer

Table 9 Specialty and non-specialty acquisitions in the US and Europe during the 1980s^*

Table 10 Focus in firm acquisition and divestment strategies of large chemical firms

Table 11 Concentration in investments in chemical plants, 1980-90, all areas

Table 12 Trends in financial acquisitions by country since the mid 80s to 1992

Notes

Note 1: The data are from Chemical Economics Handbook (CEH), SRI International, June 1991, and relate to the end of the 1980s. In some markets, typically involving differentiated products, capacities were not available, and we picked all the producers listed as leading producers in CEH. Our procedure can be best described as an informal stratified random sample of all the products covered in CEH. We selected our products to be as representative as possible of the chemical industry (excluding pharmaceuticals), and they cover all the major categories -- basic inorganic, basic organic, basic petrochemical, thermoset plastics, synthetic fibers, surfactants, resins, surface coatings, paints and pigments, elastomers, and fertilizers, and all the regions. Back.

Note 2: We used information from Predicasts (1991) and other company thesauruses to group the firms that were subsidiaries of other companies under the name of their parents, and to obtain their nationalities. Small firms are therefore likely to be under-represented in our sample since selecting only the firms that are covered in the commercial databases, one is more likely to exclude a greater fraction of the smaller firms. This procedure resulted in a sample of 473 companies -- 175 are from the United States, 132 are from Western European countries, and 166 from Japan. Back.

Note 3: In addition to dyes, the German chemical industry was strong in other fields, such as inorganics and high-pressure chemical processing from coal. BASF for instance had developed a contact process for sulfuric acid, and it was responsible for many process innovations, which culminated in the development of the Haber-Bosch process. BASF (within IG Farben) also pioneered research in the 1920s and 1930s on coal hydrogenation to produce synthetic gasoline. Back.

Note 4: Although it may seem difficult to believe today, Allied Chemicals in the 1920s was as large as Du Pont. However, unlike Du Pont it adopted a strategy of staying in the large volume inorganic chemicals, with the notable exception of ammonia and fertilizers. Alongside, the strategy was one of low cost, high volume production, with considerable divisional autonomy, and minimal attention to investments in in-house technological capability. Later events proved this to be a somewhat short-signed strategy for the prevailing conditions. What is important to note is that unlike the ICI (and its predecessor companies), which took a long time to come around to accepting the importance of in-house research capabilities, the effect of Allied's failure on US industry as a whole was very limited. By contrast, the reluctance of the leading British chemical companies, in the 1870s and 1880s, to invest in the new organic chemistry, proved to have long lasting effects on the British chemical industry. Back.

Note 5: Morris (1994) provides a detailed account of the synthetic rubber case. For aviation fuel, see for instance, Spitz (1988) and Aftalion (1989). Back.

Note 6: Another important class of technological innovations that is complementary to those in polymer chemistry and chemical engineering is in the area of catalysis. Indeed, many of the new products required advances in all three areas. For instance, a new polymer typically required both a new catalyst system, as well as substantial research in process design. Back.

Note 7: First, BP and ICI signed a deal in 1981 which led to the consolidation of their businesses. BP ceded its PVC operations to ICI, which pulled out of polyethylene by relinquishing its activities to BP. This concentrated PVC in ICI and polyethylene in BP. Then, in 1985, ICI formed a joint-venture, European Vinyl Corp., with the Italian company Enichem, which merged the PVC businesses of the two firms. The new company became the major European PVC producer. Similarly in polypropylene, Statoil and Neste have merged their petrochemical operations to form Borealis (sales $2.3 bn), to form Europe's largest, and the world's fifth largest polyolefin producer. Back.

Note 8: The most important intervention was the fine imposed by the European Court in 1986 against 15 leading polypropylene producers (among others, Montedison, ICI, Shell, Hoechst), which were found guilty of market-sharing and price-fixing actions. Back.

Note 9: In addition, selected deals from earlier years are also included. Notice also that acquiring companies include non-chemical firms. However, chemical companies account for about two-thirds of our sample of acquirers. Our data source also provided up to 10 four-digit SIC codes of the acquiring companies, as well as the four-digit SIC code in which the acquisition can be classified. We matched these data with other publicly available data sources to obtain nationalities of both the acquiring and acquired companies. Back.

Note 10: $$Word$$ was achieved in part through acquisitions. In the US for instance, Monsanto acquired G.D.Searle in $$Word$$ its entry into pharmaceuticals. In Europe too, many large companies acquired specialty producers $$Word$$ from commodity chemicals. (e.g. Rhone-Poulenc). Back.

Note 11: The low value of the US dollar, especially in 1987-1988, made US targets more attractive. Indeed, our data indicate that acquisitions of US companies by European firms grew significantly in that period. See also Lane (1993). Back.

Note 12: These are the 250 largest chemical firms ranked by world chemical sales in 1988. They are listed in the Appendix of Aftalion (1989). Back.

Note 13: In other words, if S_i is the share of acquisitions that the firm makes in a sector corresponding to the SIC code i, then the herfindal index is computed as the sum of the squares of the individual shares, ΣS_i ². This measure takes value one if all acquisitions are in a single sector, and becomes progressively smaller as a given number of acquisitions are spread over a larger number of sectors. Back.

Note 14: For instance, Dow has made significant investments in Leuna in Eastern Germany. DuPont has made a serious push into Asia, especially in its fibers business (Tyvek, lycra, nylon), in fibber intermediates such as adipic acid, adiponitrile, and in Titanium Dioxide (Chemical Week, 1994e: 24). The result is that 10% of its chemical sales are now accounted for by the Asia pacific region. Courtaulds which initially made investments in marine coatings in East Asia, and expanded on that base, now derives 14% of total sales from that region. Back.

Note 15: Hanson's failed attempt is credited with forcing ICI to restructure by dividing the company -- Zeneca, focusing on the life-science based and R & D intensive sectors such as pharmaceuticals and pesticides, and ICI, focused on the more traditional chemical business. GAFs raid in the aftermath of the Bhopal tragedy in 1984 forced Union Carbide to sell off its consumer goods businesses (such as eveready batteries, prestone anti-freeze), as well as spin off its industrial gasses Linde division as Praxair, and focus upon its polyolefins business. Back.

Note 16: As this paper was being written, there was ample evidence that even the large German chemical companies were becoming far more responsive to shareholder interests. For instance, Hoechst has recently announced the separation of its life sciences and chemicals business. Bayer is reportedly planning something along similar lines, and is said to be looking for a buyer for its photography business. Unlike Hoechst and Bayer, however, BASF does not plan to give up its strategy of vertical integration, leading the BASF chairman, Jurgen Strube, to note that "...For the first in the past 40-50 years the three German companies are developing in totally different ways." (CW, Nov 20, 1996: 20) Back.