Understanding Learning in Technological Trajectories: Combining Organizational Integration and Industrial Dynamics

Transcription

1 Understanding Learning in Technological Trajectories: Combining Organizational Integration and Industrial Dynamics by Frank Skov Kristensen IKE Group/ DRUID Department of Business Studies Fibigerstræde 4, Aalborg University 4, 9220 Aalborg Ø. Abstract This paper argues that a cross fertilisation exercise combining the insights from neo- Schumpeterian and evolutionary economists work on industrial dynamics, with Lazonick s organisational integration hypothesis, may offer the organisational integration hypothesis some valuable insights. Namely insights on firstly what a firm or an organisation must be able to integrate to become and stay competitive and secondly how this differs across sectors. By employing Pavitt s sectoral innovation taxonomy as a starting point, the paper demonstrates the sectoral differences related to the underlying technological trajectories and the learning regimes within sectors. These sectoral differences relate to the locus of learning within sectors that are essential in Lazonick s organisational integration hypothesis. It is also argued that the cross fertilisation exercise offers insights that can be useful for the neo- Schumpeterian and evolutionary scholars in the elaboration on both the theory of the firm, and on industrial dynamics. Where neo-scumpeterian and evolutionary scholars mainly have focussed on technology; the underlying technological knowledge base and the path dependence related to technology, the exercise in this paper leads to suggest a broadening of the focus. A broadening that includes a focus on the cumulativeness in developing along a specific learning regime. Applying such a broadened focus is argued to be useful in discussing firm integration, disintegration and firms product diversification. On the industry level it is useful when discussing the emergence of industries and the spin off of new industries from existing industries. Acknowledgment The Author would like to thank Ed Steinmüller and Nick von Tunzelmann for comments on a previous version of the paper. The usual disclaimer applies.

2 I. Introduction This paper elaborates on the organisational integration hypothesis put forward by William Lazonick, see Lazonick (1990), Lazonick and West (1995), Lazonick and O`Sullivan (1996). The aim of this discussion is to contribute to theories of industrial dynamics and theories of the firm. Lazonick`s organisational integration hypothesis is empirically oriented towards explaining the rise and decline of competitiveness of countries or regions. Within this empirical setting Lazonick develops a theory of competitiveness that focuses on organisational integration. Lazonick`s understanding of an organisation is broad in scope. It includes firms, groups of firms, suppliers, costumers, financial institutions, other institutions, social organisation, policy institutions, Lazonick and West(1995), Lazonick and O`Sullivan (1996). In this respect, Lazonick s ideas is somewhat similar to the ideas in Porter (1990), since both argue that a firm s or an industry s strength cannot be fully understood if it is treated as a selfcontained entity. Like Porter, Porter (1990), Lazonick also emphasises that a theory of competitiveness, be it regional or national, must spring from a theory of the firm. The organisational integration perspective, however, suggests that the research agenda must begin at the level of the strategy and structure of business enterprises as a prelude to a consideration of the roles government policy might play. Lazonick and West (1995, p. 268) While both Porter and Lazonick suggest to focus on the firm, and also agree that a firm s competitiveness cannot be fully understood if one abstracts from the firm s environment, Lazonick s strong focus on collective learning and intra organisational integration, Lazonick and O`Sullivan ( 1996) cannot be found in Porter`s work. We will suggest that this difference stems from a fundamental difference in their approaches. Whereas Porter largely ascribes to the classic structure-conduct- performance (henceforth SCP ) methodology as presented by Shepherd, Shepherd ( 1985 ), we will 2

3 argue that Lazonick methodology relates to a dynamic (SCP) methodology, as outlined by Steinmueller (1998). We shall later on return to the dynamic SCP methodology, and to how it differs from the classic SCP approach. In Lazonick s organisational integration hypothesis the broad social and cultural norms are emphasised as important. This line of thinking is shared in the concepts of national systems of innovation put forward in Lundvall (1992), Nelson (1993), and in the ideas of technological systems and technological paradigms, Freeman (1982), Perez (1983). Lazonick stems from those political economists arguing for industrial democracy, this might also hold for some ideas within the neo-schumpeterian and the evolutionary 1 schools. Still we will argue that the similarities in ideas to a higher extent stems from the fact that there is a common methodological starting point, namely a dynamic SCP methodology, rather than from the fact that they might share a common political ideology. Those who are familiar with the concepts of national systems of innovation, technological systems and technological paradigms, will know that the interactions between the firm and its surroundings are numerous and complex. Lazonick makes a similar observation of scope and scale in relation to the organisational integration hypothesis: Such an analysis is obviously the work of a multi-year, multi-person research project that has yet to be done (Lazonick and West 1995, p.235). 1 One critique that can be raised towards Lazonick s work is that the ideas have quite clear political motives and undertones. Industrial democrats in general propose that especially the US should be a more egalitarian society, both from a social and democratic point of view. But others as Lazonick claim that a more egalitarian and democratic approach involving the workforce will lead to productivity increases. See Bowles et al (1990) for further elaborations of thoughts among the political economists, proposing industrial democracy. Still as an answer to this critique, we will claim that Lazonick and his colleagues are not alone in pursuing worker integration and empowerment as an explanation for the Japanese miracle in the Automobile industry. In Womack et al (1990), workers integration, empowerment and the focus on employee skills is also put forward as a vital reason for the success of the Japanese or the Japanised automobile manufactures. 3

4 Acknowledging this point, this paper focusses on the firms organisation and structure, hereby also acknowledging Lazonicks proposal for the point of departure when developing the organisational integration hypothesis. In Lazonick and West (1995) the focus of the analysis is mainly on the integration of the shopfloor workers in the US and Japanese automobile industry. It is argued that it was the strong integration of shopfloor workers in the industry in Japan that rendered the Japanese a competitive advantage. This is because the shopfloor workers widely and effectively participated in the continuos improvements of products and processes. While the integration of shopfloor workers was important in the automobile industry Lazonick and West also notes that this does not seem to be the case in the chemical and pharmaceutical industries, Lazonick and West (1995). This implies that what is important for a firm to integrate, varies according to the sector in which the firm is situated. Lazonick emphasises this when he argues. Only through careful case studies can we discover just what are the key relationships within business organizations that influence the choice competitive strategy and that generate competitive advantage, and how these relationships vary across industries. Is it in relation with production workers that are the key? Or with suppliers? Or with other enterprises engaged in cooperative research or other types of collective investments? Lazonick and West (1995, p. 268) This paper elaborates on what is to be integrated in which sectors. Put differently we ask the questions : C Are there any general key relationships that a firm in eg. the pharmaceutical industry integrate, and are this different for a firm in computer industry? C What causes these differences in what must be integrated? Lazonick suggest that these differences and causal relations can only be found by 4

5 employing careful case studies. While the author largely agrees to this we will claim that by employing the ideas of Pavitt, Pavitt(1984) and Pavitt (1997), and the ideas developed within the neo-schumpeterian and evolutionary schools, will bring some useful insights to address Lazonick s and our question of industry differences in relation to what a firm has to integrate. We further argue that the neo-schumpeterian and the evolutionary schools can benefit from Lazonick s ideas related to the high costs of creating structures that enhance learning. We argue that these insights have some useful implications for both our understanding of the firm and also on our thinking of industrial dynamics. In what follows Lazonick s organisational integration hypothesis will in some depth be presented in section two. This is followed by a discussion of Lazonicks views on the firm level integration and proposed industry differences in section three. In section four the Pavitt taxonomy, Pavitt (1984) is applied in an adapted form to discuss the neo- Schumpeterian and evolutionary insights on industry differences. In section five we conclude by discussing the implications of our analysis for theories of the firm and on industrial dynamics. 5

6 II. Lazonick s organisational integration hypothesis Lazonick s starting point is to develop a sound understanding of why the American industry was not able to sustain its world leadership, which it had upheld for almost a century. Where American industry and American firms dominated almost all industries up until the 1960`s the competitive advantage of the Americans in the 1960`s and 1970`s rapidly shrunk, and in the 1980`s the leadership was lost. The former American advantage and leadership was first and foremost taken over by Japanese, and to some extent, European firms and industries. In the 1980`s and 1990`s several studies of the decline of the US industries and firms, and the rise of their Japanese and later East Asian, counterparts were conducted, Baumol (1989), Porter (1992), Dertouzas (1989), McKinsey Global Institute (1993). The studies aimed to explain the rapid decline of the competitiveness of US firms and industries. Insights into the reason for the decline should enable business and policy advise as to what should be done to regain US competitiveness. It is within this research area Lazonick and his colleagues place the organisational integration hypothesis which they argue can, when elaborated, provide insights into why firms in specific industries, regions and nations gain competitive advantages. They further, not surprisingly, argue that only with these insights will it be possible to give the right advice to policy makers and firm managers. As mentioned in the introduction the organisational integration hypothesis is broad in scope and it does entail many aspects related to integration, Lazonick and his colleagues defines organisational integration as follows. What then is organisational integration? It is a set of ongoing relationships that socialises participants in a complex division of labour to apply their skills and efforts to the achievements of common goals. The foundation of the socialising process that achieves organisational integration is membership ; the inclusion of the individual or groups into the 6

7 organisation with all the rights and responsibilities that membership entails. Lazonick and West (1995, p. 231 ). From this paragraph one can get the impression that we are within the boundaries of a firm. This is not the case as Lazonick and West continues. Our conception of organisational integration may include individuals and groups who are employed by legally distinct firms that pursue common goals. Conversely, individuals and groups employed by the same firm may not cooperate to achieve common goals (Lazonick and West 1995, p. 231). The common goal that is referred to is success, or perhaps more precisely the sustained well being of the members and of the entity in which they are members. In the latter paragraph it becomes evident that we are way beyond the boundary of the firm in the perception of its possible members. This definition of an entity or organisation is much in line with what Granovetter calls a business group, Granovetter (1995). Granovetter excludes many US conglomerates form his notion of business groups since these only are consolidated legally into one firm, without having any other formal or informal bindings. Granovetter (1995, p. 95). As is it can be seen from the two quotations above, Lazonick and West s definition also excludes conglomerates of this kind. These conglomerates are excluded if the business units within them do not pursue common goals. Or if there are no ongoing relationships that socialise the participants in a complex division of labour. Further, and as we shall see later, of crucial importance, employees in a business unit can be excluded form the organisation, either as a consequence of their own choice or the choice of others within the organisation. In this broad definition of the organisation the members can be numerous. Employees, managers, board members, suppliers, costumers, banks, shareholders, competing firms with whom the firm cooperates and institutions are only some of the possible members that can be mentioned. The analytical focus in this broad entity is directed upon how well the integration of the entity is. And it is the level or the strength of the integration 7

8 in the organisation that matters when explaining the competitive ability of an organisation. The reason why integration is of such significant importance, Lazonick relates to the need for innovativeness. Lazonick stresses innovativeness as the mean for a firm to gain and sustain competitiveness. This, since innovativeness is what ensures the firm, or the group of firms, its ability to constantly improve or change their processes or products which in turn renders the firm improved competitive abilities. Innovation is the outcome of learning which by Lazonick is seen as a concentrated, continuous, cumulative, collective and above all a costly process, Lazonick and West (1995) and Lazonick (1993). That the learning process is concentrated, continuous, cumulative and collective requires that the entity engaged in learning is organisationally integrated and the fact that it is costly requires what Lazonick calls financial commitment. Financial commitment represents the willingness of those who control financial resources to commit these resources to financing the high fixed costs of developmental investments that, because they entail innovation, 2 promise uncertain returns. Lazonick and West (1995, p.236). The financial commitment is central to Lazonick`s works and the idea is thoroughly elaborated and analysed, especially in Lazonick and O`Sullivan (1996). We will argue that in the Lazonick framework financial commitment is the core condition for organisational integration, learning, innovativeness and competitive ability, since Lazonick and West argues. Financial commitment plays a critical role in the innovation process because those who control money get to choose what type of strategy a firm will 2 It is evident that Lazonick does not see learning as a costless externality that stems from using as Arrow do in his learning by using concept, Arrow (1962). 8

9 pursue. Lazonick and West (1995, p. 236). What Lazonick and his colleagues basically argue is that financial commitment is a precondition for pursuing a innovative strategy. Further they argue that an innovative strategy is a necessary condition, if a firm is to create and sustain its competitiveness. Borrowing the vocabulary from Nelson and Winter, Nelson and Winter ( 1977), it is the tradeoff between exploration and exploitation. A strategy based on exploitation will yield high incomes now as the knowledge is cannibalised, but it will also leave the future insecure. On the other hand a strategy to emphasise exploration yields less income now but will increase the potential of future incomes. While financial commitment will ensure a firm the possibility to pursue an innovative strategy, this alone will not ensure competitiveness. As Lazonick and his colleagues demonstrate, some firms are better than others in creating structures to reap the benefits of such an investment into learning which leads to innovativeness, Lazonick and West (1995), Lazonick and O`Sullivan (1996), Lazonick (1990). The claim made in Lazonick and West (1995) is that it was the higher level of organisational integration and stronger financial commitment to learning and innovation that enabled the Japanese firms to overtake the leadership form the US firms in some industries. The much closer integration of the shopfloor workers in the Japanese firms in the automobile industry is argued to be the key factor that was better integrated. Through case studies, the closer integration of shopfloor workers in Japanese firms are elaborated. One of the core factors in this closer integration is believed to be the lifelong employment which is common in some Japanese firms. Another core factor is the heavy investments made by Japanese firms to continuously upgrade the shopfloor workers skill. In return this investment in worker skills will enable the workers to participate in the continuous learning process within the firm that enables both innovations in processes and products. In the US firms on the other hand, the relation between the shopfloor workers and the management was more characterised by an us versus them relationship. This relationship lead to investments 9

10 in displacing shopfloor workers, and ensuring that the remaining workers was a common and unspecialised good. An unspecialised good that could be hired and fired at whish and will. In the Lazonick framework this means that while the Japanese shopfloor workers were viewed upon as members of the group. The US shopfloor workers were externalised from the group which only included managers. While Lazonick and his colleagues were able to show that in fact this closer integration was present, this is of cause not enough to prove that this was the causal factor. This fact is acknowledged and dealt with by showing that the US firm that actually took up the Japanese challenge and Japanised (integrated the organisation better, and pursued innovative strategies by committing more financial resources to learning) was more successful than the US firms that continued the US way. This is done by developing a taxonomy of US firm responses or strategies; first movers, fast movers, slower movers, no movers and removers. Lazonick and West (1995). Among the examples used in the analysis are Ford who is a slow mover and GM who is a no mover. Ford was, when it eventually responded, much more successful than the other big US company, GM, that continued the US way. While Ford committed financial resources to improve learning and started to integrate shopfloor workers, GM had not changed and were in a sense cannibalising previously established strengths. This is argued to be why GM were losing ground to as well the Japanese and to Ford, Lazonick and West (1995). A curiosity in this story which we shall return to later, is that GM actually have Japanised a part of their production. In the Saturn project, GM employ Japanese production and management techniques. But in spite of the high quality and efficiency improvements GM have achieved in connection with the Saturn model and also in spite of the Saturn models success in the market place. The insights gained within the Saturn project was not infused into main stream GM. GM, as it is convincingly phrased, largely continues its... automation around the shop-floor workers and fighting suppliers, rather than integrating either group into the company s investment plan Lazonick and West (1995, p. 260). Whereas the above section of the present paper briefly has presented the core ideas in 10

11 the work made by Lazonick and his colleagues on the organisational integration hypothesis. The following section will focus more narrowly on the main topic in this paper. The following section therefore aims to extract from Lazonicks work what can be said on sector specific differences; differences in what a firm within a given sector must manage to integrate to gain and sustain competitiveness. 11

12 III. Lazonick on differences in sectorial dynamics and organizational integration While the former section concentrated on what is meant by integration of shopfloor workers and financial commitment which ensures that an innovative strategy can be pursued. We shall in this section return to the broad framework of the organisational integration hypothesis. But now we direct the focus on the sector specific differences in what must be integrated. The idea that there are sectoral differences in the interface or interfaces the firm must manage to integrate, is explicitly found in Lazonick s works. This can be seen in the quotation in the introduction of this paper, and also in the quotation below in which Lazonick notes. In industries in which, from the 1960's, an organisationally integrated managerial structure alone continued to suffice in global competition - industries such as pharmaceuticals and chemicals - the Americans continued to be leading innovators, and the Japanese companies were unable to mount an effective competitive challenge. Lazonick and West (1995, p. 232). By making these statements Lazonick gives us two implications. Firstly he explicitly states that competition vary among sectors and this will affect the strategy and structure of the firm. Secondly it is the structure in which a firm is situated that determines, or give imperatives to, its behaviour. This means that we can understand Lazonick s ideas as belonging to a structure - conduct - performance (SCP) line of 3 thought. SCP is also the methodological starting point for the neo- Schumpeterian and 4 evolutionary schools, in which sector specific structures for many years have been a 3 Here the author will subscribe to the more dynamic and broad notion of structure - conduct - performance, suggested by Steinmüller (1998). 4 Though, some evolutionary economists or lines of thinking might be excluded, we will argue that those following the idea in Nelson and Winter (1982), of natural trajectories do subscribe to the SCP methodology. 12

13 central agenda. Pavitt (1984), Malerba and Orsenigo(1995) develop and employ specific sector taxonomies. Abernathy and Utterback (1978), Barras (1986) are examples of studies theorising on the dynamics of industries, Klepper and Grady (1990) give empirical accounts for the different dynamics and structures in industries. Dosi (1984), Nelson and Winther (1982), theorise on the reasons for different trajectories. Malerba (1992) addresses the relationship between different trajectories and different types of learning. Among the studies that more explicitly address the organisational and strategical relationship with sector specific structures is (Pavitt, 1997), who addresses the relationship of sector difference in technological imperatives and sector difference in the governance structure of the firm. Besides the few studies mentioned here there are a wealth of studies of which some are reviewed in Dosi (1988) Malerba and Orsenigo (1996). Following the idea in Lazonick, we basically argue that the actions of the firm, that is, the integration and management of different interfaces within the organisation, will be determined or guided by the constraint and the rules of the industry structure in which it is situated. Therefore, knowledge of structures within a industry is a precondition to understand the firms actions. It is these insights related to the knowledge of the different industry structures, we argue is well studied and elaborated in the above mentioned studies within the neo- Schumpeterian and evolutionary schools. Therefore we further argue that building on these insights will enable us to elaborate on the organisational integration hypothesis. In the following section we will address the issue of what industry structures is, how they evolve and how they differ. But before turning to that we will shortly elaborate on what we have called the dynamic SCP approach. The dynamic SCP approach have not to our knowledge been fully elaborated in any 5 writings. This elaboration rest solely on a series of seminars given at SPRU, by Ed Steinmüller in 1997/ All the above mentioned works on industrial dynamics 5 Science Policy Research Unit, at Sussex University, UK. 13

14 emphasise structures and changes in structures to be crucial. Especially the technological paths or trajectories are seen as to be more or less given or natural, Nelson and Winter (1982). Industries evolve over time and as they do the firms in the industries will adapt. Of cause the firms take an important part in creating this evolution why they cannot be understood as to be just price takers. The structures in an industry are developed by the nature of the technology within the industry, but it is also affected by innovation made outside the industry. The firms within the industry do develop their conduct which affects the performance of the firms and also the industry as a whole. The fundamental difference between the classic SCP methodology and following Ed Steinmüller a dynamic SCP methodology is related to equilibrium and innovation. In the dynamic SCP methodology, equilibrium is rejected and innovation and technological development is set at the heart of industrial development. Whereas the classic SCP disregards technology and builds on equilibrium. 14

15 IV. Combining neo-schumpeterian and evolutionary ideas of industry structure, evolution and differences with the organisational integration hypothesis. It is beyond the scope of this paper, to discuss or review the field of industrial dynamics and we will restrict our discussion to concentrate on five concepts These five concepts are; Natural trajectories, Nelson and Winter (1982); Technological trajectories, Dosi (1984); Technological opportunities and appropriability, Malerba and Orsenigo(1993); Learning regimes, Malerba (1992); Sectoral patterns of innovations, Pavitt (1984). For reasons of simplicity the following discussion will be structured around the latter, Pavitt s taxonomy of sectors. The Pavitt taxonomy, Pavitt (1984), was made on the basis of a descriptive analysis of the innovation process in different sectors. The empirical data were a database of nearly 2000 innovations made in the UK in the period 1945 to 1979, Pavitt (1984, p.344). Pavitt descriptive analysis addressed some of the major issues which had emerged from various indebt industry studies of innovation and innovation processes. The outcome of Pavitt s analysis was two fold. Firstly it posed a taxonomy of patterns of innovation. Secondly it presented a theory of innovation flows among four different sector types. Here we will mainly turn to the first part of the outcome; the taxonomy of sectoral patterns of innovations. Figure 1 presents our adapted use of Pavitt s sector taxonomy. In the 1984 taxonomy there were only four sector types; the fifth type the information intensive sector type was added later and can be found in, Tidd, Bessant, Pavitt (1997). 15

16 Figure 1. Adapted from the Pavitt taxonomy of innovation patterns. Sector type/ Science based. Scale intensive. Supplier Specialised Information variables. dominated. suppliers. intensive. Firm size. Large firms. Large firms. Small firms. Small firms. Large firms. Type of Mixed Processes. Processes. Products. Mixed products innovation. products and and processes. Locus of processes. Production. External. Decentralised. R&D innovation. R&D departments. Sources of departments. Production and Specialised Science based Universities, innovation. Universities, specialised suppliers. firms and research research suppliers. costumers. centres. Means of centres. Tacit knowledge Tacit Tacit knowledge, appropriability Patents and and entry knowledge. reputation. Tacit. entry barriers. barriers. knowledge and entry barriers. Competitive Price, quality. Price. Quality, form. Performance, performance. Performance, quality and quality, price. Sectors. price. Automobiles. Stone, Glass. Mechanical Pharmaceutical engeniering. Banks,, insurance, microelectronic retailers. 6 Trajectories. s. Improvements in Improvemen Improvements in process yields ts in process the performance Improvements Performance and increases in yields. and reliability of in processes and and physical the scale of products. in products. properties of production products, and processes and improvements automation. Learning in process yields. Learning by learning by Learning by 7 regime. doing/ learning using. interacting/ Learning by Learning by by using. learning by doing. searching/ searching/ learning by learning by Doing. doing. 6 Taken from Malerba and Orsenigo (1996, p.58) the suggestions to the logic of the trajectory in the information intensive sector, is though the author s own speculation. 7 Based on the results from Malerba (1992). 16

17 Figure 1 above presents the sector types and the stylised facts of these sectors. From these we will begin our elaborate on differences in industry structures. At the bottom of the figure we have added the nature of the trajectories in the different sector types and the learning regime related to the development of the trajectories. While the sectoral taxonomy focus on the level of industry sectors, and as can be seen from the figure provide insights to the firm size, the innovation process and the competitive parameters within the sectors. The notion of technological trajectories is related to the path of development within a given technological field, Dosi (1984). Trajectories differ in the direction and also in their advances. These differences in the advance, or the rate of development of a given technology are following, Nelson and Winter (1982), determined by the underlying technological regime related to it. Malerba and Orsenigo (1993) have proposed that the technological regime is determined by the opportunity conditions ( the ease in which the technology can be developed ), by the appropriability conditions ( the possibility to protect the innovation against innovation), by the cumulativeness ( which means that today s innovation is the starting point for the innovations tomorrow ) and by the nature of the underlying knowledge base ( which may be tacit and locally embedded or codified and universally available ). In our context technological trajectories and technological regimes are of importance since, Nelson and Winter (1982) propose that the nature of the regime will effect the firms actions a proposal that finds empirical support in the results in Malerba and Orsenigo (1993). The learning regime relates to the dominant way in which learning takes place within a given trajectory. Malerba (1992) found that there were correlations between the logic of a trajectory and the dominant form of learning within the trajectory. ( In trajectories where the logic for example is yield improvements, the dominant way of learning is learning by doing. ) Figure 1, displays that the structures within the sectors differ, and also the locus of learning differ from sector to sector. We will argue that exactly the locus of learning is what Lazonick is addressing, when he implies that there might be sector differences in what is essential for a firm to integrate. By applying the trajectories and the learning 17

18 regimes related to the trajectories we are also able to identify different development paths and mechanisms of learning among the five sectors. Taking the example from the automobile industry, which is a scale intensive sector, the locus of learning lies in the production or is closely related to the production. The trajectory in this sector is focussed upon improvements in process yields, increases in the scale of production, processes and automation. The dominant mechanism of learning is assumed to be learning by doing. In arguing that the locus of learning is situated in the production, we do not imply that other sources may not be important but we do imply the learning in production is of essential importance. Exemplifying this in the Lazonick automobile case, we can say that the Japanese companies by integrating and investing in the shopfloor workers enabled the workers to engage directly in the locus of learning. In a cumulative learning process aimed at improving process yields. The exercise above give some emerging stylised facts of what a firm have to integrate 8 given the structure, and the dynamics of the sector in which the firm is situated. We have already discussed the scale intensive sectors and now turn to the Science based sectors. In the science based sector the locus of learning is within the firm s R&D department their relations to external institutions such as universities and research centres. Here the key integration will be among the R&D department and the institutions in the external environment. Whereas integration of customers and suppliers and shopfloor workers in relation to pharmaceuticals, are of less importance. One explanation as to why the Japanese could not take over the US in pharmaceuticals, is that the regime of learning by doing in which the Japanese have invested heavily, is of no vital importance within the pharmaceutical industry. In pharmaceuticals investments in learning by searching and integration of external knowledge are far more important than learning by doing in the factory. 8 We deliberately use the phrase stylised facts because it is not the intention to pose a rigid idea of any sectors entailing only one learning regime. Or an idea that a firm only have one interface to integrate. In most sectors and trajectories there are more than one learning regime that a firm must handle. Still we argue that one seems to be the dominant. 18

19 Within the specialised supplier sector strong integration of consumers if of central importance since the locus of learning is nested within the user-producer relations, Lundvall (1988), Von Hippel (1988). The users knowledge related to eg. processes is of crucial importance for the supplier firm in order to deliver. From these interactions the firm learns to solve problems and to make equipment what also entails a focus on learning by doing. In relation to supplier dominated sectors one can say that there is a reversal userproducer relation, where the user must be able to explain its needs to the supplier. The ability to do so the firm build up by learning by using the equipment. In the following section we will conclude on our discussion in this paper. The form of these conclusions will be to point towards some implications both in relation to theories of the firm and in relation theories on industrial dynamics. 19

20 V. Conclusion and implications. We first turn to the implications for Lazonick organisational integration hypothesis. One puzzle was, if and how there were sector specific differences in what was a key factor to integrate within the organisation. Drawing on neo-schumpeterian and evolutionary contributions discussed on the basis of Pavitt s sector taxonomy of innovation, we have put forward a framework of the sector specific factors that a firm must integrate given its sector relation. We have proposed that these different interfaces embed the locus of learning, and that the mechanisms or ways of learning differ. We argue that these locus s of learning or dominant learning regimes which are, if not given, then closely related to the technological trajectories within the industry basically are what the firm must develop along to innovate and stay competitive. Some forms of organisation and some firm structures enable this learning process better than others like we saw in the automobile case. While different organisational forms may coincide in a period of time we argue that eventually the potential in one or the other ( in the automobile case the Japanese ) will create advantages that force the whole industry to change towards this form. Where as this cross-fertilisation exercise offers insight to the organisational integration hypothesis, in relation to sector specific differences in the interfaces that are crucial to integrate. We also argue that there are insights to be gained for a neo-schumpeterian an evolutionary firm theory and for neo-schumpeterian and evolutionary theory on industrial dynamics. In a recent article Pavitt claims that large firms rarely fail as a result of their lack of technological competencies but rather because of a lack in managerial competencies to link the underlying knowledge base, to the possibilities in the market, Pavitt (1997). Whereas the author agrees to this point, we would suggest that managerial failure to understand the logic of the learning regime is equally important. This can be illustrated In the case of GM presented above in section 2. We saw that GM actually have had some success in the Saturn project. Still GM is not applying the Japanised ideas into the 20

21 mainstream GM. For this there can be many reasons, eg lack of financial resources, strength and inertia of managerial groups. Hence whatever the reason we in this case can talk of a management failure in understanding the underlying learning mechanism or the logic of the learning regime within the trajectory. Still even if we here talk of a managerial failure, it is worth to bare in mind Lazonick s point that creating innovation is cumulative, collective and costly. This, we will argue implies that shifting from one way of learning to another is both a costly and time consuming process. There are so to say sunk costs in the old learning regime. If we pursue this thought, of the difficulties in changing from one learning regime to another due to cost and time, and add that a firm in eg pharmaceuticals also have invested large sums in the technological fields related to pharmaceuticals. We will propose that a firm may have an advantage when diversifying into markets or industries where it can use either its investments in learning regime or its investments in technological fields. But in markets where the firm can not use any of these accumulated competencies, the firm has no specific advantages in relation to other firms. This might help to explain why there as Pavitt mentions are quite clear patterns in firms diversification, Pavitt (1997) even if, as some claim, we in the contemporary economy see an erosion of the traditional industry boundaries. One illustrative example of this can be the Pharmaceutical company Smith- Kline & Beecham who now define them selves as a health care company rather than a pharmaceutical company. While Smith-Kline & Beecham do move into diagnostics where they can use their technological base, they do not move into making diagnostic instruments which entails quite different technological competences and as we argue has a different learning regime. Further our exercise might also give some implications when trying to understand the spinoff of new industries. One case can be the spinoff of soft ware from the computer industry. Though there is a clear interdependence between these two industries neither the underlying technological base nor the learning regime is similar. We might say that 21

22 hereby the hardware companies did not have any real advantages to other intruders. In other words the dynamic structure of the software industry was entirely different from the structure in the hardware industry. Therefore the conduct of the firms within the industry, has to differ as well. We do not held to have presented anything close to a full blown theory of neither the firm or of the industrial dynamics. Our claim is somewhat more humble and will be limited to argue the following. Including the processes, the rigidities and the costs of different forms of learning in our theorising of the firm and of industrial dynamics, do bring these theories closer to what firms actually do. They produce goods and services. We argue that a theory of the firm and a theory of industrial organisation must place the production and the ongoing development production factors at the heart of its theory, rather than disregarding production by keeping it constant. This argument is not new nor is it our own invention; it is the core in A. Smiths understanding of the specialisation of labour which in his view drove the economic growth and development. 22

23 Bibliography Abernathy, W.J. and J.M. Utterback (1978), 'Patterns of Industrial Innovation', Technology Review, Vol. 50, pp Arrow, K.J. (1962), 'The Economic Implications of Learning by Doing', Review of Economic Studies, Vol. 29, pp Barras, R. (1986), 'Towards a Theory of Innovation in Services', Research Policy, Vol. 15. Baumol, W.J., S.A.B. Blackmann and E.N. Wolf (1989), Productivity and American Leadership: The Long View, MIT Press. Bowles, S., D.M. Gordon and T.E. Weisskopf (1990), After the Waste Land: A Democratic Economics for the Year 2000., New York, M. E. Sharpe Inc. Dertouzas, M.L., R.K. Lester and R.M. Solow (1989), Made in America: Regaining the Productive Edge., MIT Press. Dosi, G. (1984), Technical Change and Industrial Transformation: The Theory and an Application to the Semiconductor Industry, London and Basingstoke, Macmillan. Dosi, G. (1988), 'Sources, Procedures and Microeconomic Effects of Innovation', Journal of Economic Literature, Vol. 26, pp Freeman, C. (1982), The Economics of Industrial Innovation, London, Frances Pinter. Granovetter, M. (1995), 'Coase Revisited: Business Groups in the Modern Economy', Industrial and Corporate Change, Vol. 4, pp Hippel, E.v. (1988), The Sources of Innovation, New York and Oxford, Oxford University Press. McKinsey Global Institute, (1993), Manufacturing Productivity, McKinsey and Company. Klepper, S. and E. Graddy (1990), 'The Evolution of New Industries and the Determinants of Market Structures', Rand Journal of Economics, Vol. 21, pp

24 Lazonick, W. (1990), Competitive Advantage on the Shop Floor, Cambridge, MA, Harvard University Press. Lazonick, W. (1993), 'Learning and the Dynamics International Competitive Advantage', in R.e. Thomson (ed.), Learning and Technological Change., Macmillan. Lazonick, W. and M. O'Sullivan (1996), 'Organization, Finance and International Competition', Industrial and Corporate Change, Vol. 5, pp Lazonick, W. and J. West (1995), 'Organizational Integration and Competitive Advantage: Explaining Strategy and Performance in American Industry', Industrial and Corporate Change, Vol. 4, pp Lundvall, B.-Å. (1988), 'Innovation as an Interactive Process - from user-producer Interaction to National Systems of Innovation', in G. Dosi, C. Freeman, R. Nelson, G. Silverberg andl. Soete (eds.), Technical Change and Economic Theory, London, Pinter Publishers. Lundvall, B.-Å., (ed.) (1992), National Systems of Innovation. Towards a Theory of Innovation and Interactive Learning, London, Pinter Publishers. Malerba, F. (1992), 'Learning by Firms and Incremental Technical Change', Economic Journal, Vol. 94, pp Malerba, F. and L. Orsenigo (1995), 'Scumpeterian Patterns of Innovation', Cambridge Journal of Economics, Vol. 19, pp Malerba, F. and L. Orsenigo (1993), 'Technological Regimes and Firm Behaviour', Industrial and Corporate Change, Vol. 2. Malerba, F. and L. Orsenigo (1996), 'The Dynamics and Evolution of Industries', Industrial and Corporate Change, Vol. 5, pp Nelson, R.R., (ed.) (1993), National Innovation Systems: A Comparative Analysis, New York, Oxford University Press. Nelson R.R. and S. Winter (1977 ), In Search of a Useful Theory of Innovation, Research Policy, vol 7. Nelson, R.R. and S. Winter (1982), An Evolutionary Theory of Economic Change, Cambridge, Massachusetts, Harvard University Press. 24

25 Pavitt, K. (1984), 'Sectoral Patterns of Technical Change: Towards a Taxonomy and a Theory', Research Policy, Vol. 13, pp Pavitt, K. (1997), 'Technologies, Products and Organisation in the Innovating Firm: What Adam Smith tells Us and Joseph Schumpeter Doesn't.', SPRU mimeo. Perez, C. (1983), 'Structural Change and the Assimilation of New Technologies in the Economic and Social System', Futures, Vol. 15, pp Porter, M.E. (1990), The Competitive Advantage of Nations, New York, Free Press. Porter, M.E. (1992), Capital Choises: Changing the Way America Invest in Industry., Council on Competitiveness. Shepherd, W.G. (1985), The Economics of Industrial Organisation, 2.ed, Englewood Clifs, Prentice Hall Steinmüller, W. (1998), Seminar at SPRU, Sussex University. Tidd, J., J. Bessant and K. Pavitt (1997), Integrating Technological, Market and Organisational Change, Chichester, John Wiley and Sons Ltd. Womack, J. P. and D. T. Jones and D. Ross (1990), The Machine that Changed the World, New York, Rawson Associates. 25