Depret.fm

European Journal of Economic and Social Systems 14 N° 3 (2000) 229-270
Innovation networks and competitive coalitions
in the pharmaceutical industry: The emergence
and structures of a new industrial organization
MARC-HUBERT DEPRET* AND ABDELILLAH HAMDOUCH** Abstract. – Today, the pharmaceutical industry is subjected to wide-scale
technological, institutional and economic changes that bring about a radical recon-figuration of the structural parameters and strategic variables of competition.
While innovation processes are being increasingly structured around a new para-digm and markets are being both globalized and fragmented, firms are attemptingto adapt, above all by refocusing their businesses and by redefining all the links intheir value chain. However, the most significant changes in their strategy lie in theconception of their horizontal and vertical relations. The competitive game has atendency to being extended to phases upstream market competition, to be reconfig-ured in the form of intra and inter-coalitions and networks strategic games, and toquickly evolve along with technological innovations, the opening of new market"frontiers" and successive concentration operations. The pharmaceutical industrythus constitutes a sort of life-size "laboratory" for an in-depth analysis of theeffects of radical institutional and technological changes on the new shapes ofcompetitive structuring within markets undergoing rapid globalization, and of the"spontaneous" emergence of a new form of industrial organization. In the image of many industrial or service sectors, the pharmaceutical industry is todaysubjected to wide-scale technological, institutional, and economic changes that bringabout a radical reconfiguration of the structural parameters and strategic variables ofcompetition: soaring "new biotechnology" (see Depret and Hamdouch, 2000b), in-depthredefinition of innovation processes, and continuous growth of R&D costs; changes * IFRESI-CLERSE, Université de Lille I, Faculté des Sciences Économiques et Sociales, 2 rue des Canon-niers, 59800 Lille, France. E-mail: [email protected]** CLERSE, Faculté des Sciences Économiques et Sociales, Université de Lille I, et MATISSE-CRIFES,Université de Paris I, Maison des Sciences Économiques, 106-112 boulevard de l'Hôpital, 75647 ParisCedex 13, France. E-mail: [email protected] Keywords: New biotechnology applied to health, institutional and technological changes, competitive coali- tions, globalization, pharmaceutical industry, innovation, interfirm networks.
EDP Sciences 2000 M.-H. DEPRET AND A. HAMDOUCH in legislation, market deregulation, and extension of the geographic area of competitiveinteraction for firms; rationalization of health expenditures; emergence of new thera-peutic needs, and structural changes as regards drug demand, etc. One of the most prevalent characteristics of this new context for competition lies in the concomitance of two apparently contradictory movements. On the one hand, competitionis becoming sharper at the technological, industrial and commercial levels as marketsopen and become globalized; on the other hand, it seems to be becoming relativized orattenuated through the multiplication and diversification at a completely new level ofcooperation and/or integration between firms that are current or potential rivals.
How can this "competitive paradox" be explained? If one is to refer to the usual approaches of industrial organization, there is no paradox at all: to escape head-on com-petition by prices, quantities or horizontal differentiation, rival firms in a market candecide to either "cooperate" by adopting collusive strategies, or purely and simplymerge, within the more or less strict limits of national or international antitrust legisla-tions.
This explanation, of course, holds a share of truth: the waves of mergers and acquisi- tions (M&A) at the international level over the past twenty years in many sectors consti-tute a direct consequence, both structural (concentration via the elimination of part of thefirms) and strategic (M&A deals), of the globalization of competition. This explanation, however, seems to be partial or limited, at least for three reasons.
Firstly, the proliferation of cooperation agreements persists while the sectors being glo-balized have not ceased to concentrate. Concentration and cooperation therefore seem tobe at least partly compatible.
Secondly, the traditional view of interfirm cooperation with an exclusively collusive goal hides all the cooperative strategies that are taking hold outside the markets (R&Dalliances, compatibility of norms and standards, etc.) and systematically assimilatescooperation with restriction in the competitive game. On the contrary, as this has alreadybeen demonstrated in many works, cooperation is extremely diversified in its forms andmotivations and often contributes to preserving and stimulating a non-destructive com-petitive game (Teece, 1992; Jacquemin, 1994; Hamdouch, 1998). Thirdly, and finally, the traditional approach to competition and cooperation only con- siders the relations between individual firms, assumed to be independent from each otheroutside the market, while interdependency within industrial groups and between compa-nies are today a structural feature of a competitive game that is increasingly "collective". In fact, this "competitive paradox", brought about by the different changes in progress, is without any real substance. It indeed appears through a new light when a more openand more dynamic approach is adopted as regards interfirm relations, as shown in thecase of the drug industry, studied in this article. Thus, pharmaceutical companies con-fronted with these deep changes in institutions and technologies and with the quickreconfiguration of their competitive space, are now obliged to take on more "global"strategies, which are both highly costly and risky. Yet, paradoxically enough, globaliza-tion is demonstrated by a wider fragmentation of markets that requires capabilities andcompetences on the part of firms, suited to each type of market (traditional, new orpotential markets, in terms of products and/or geographic areas).
NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY Within this context, no pharmaceutical company, no matter how powerful it may be, can hope to alone have an impact on the operation of markets and the determinants oftheir evolution. Thus, a firm's commitment in terms of relations of cooperation and/orintegration today appears to be all the more constrained since all pharmaceutical compa-nies are being subjected to this at the same time. This explains the gradual emergence ofmultilateral systems of interdependency. These interdependencies are reinforced by theglobalization-fragmentation of markets that make it necessary for each firm to enter intospecific partnerships on each targeted product or geographic market. However, the"stock" of potential partners in each market is all the more limited, on the one hand as allpartners do not present the same strategic interest for the firm and, on the other hand,because this stock precisely has a tendency to be reduced as the markets become concen-trated and alliances are made. As a result, neither rivalry, nor cooperation or concentration can be apprehended any longer in terms of interaction purely or even essentially inter-individual. The analysisproposed in this article attempts precisely to show how, in the pharmaceutical industry,the competitive game is leaning towards expanding to phases upstream industrial compe-tition, towards becoming reconfigured through struggles-cooperations among andbetween firm networks, and towards evolving quickly along with technological innova-tions, the opening of new markets, and successive concentration operations. More broadly, the current changes in the pharmaceutical industry form a sort of life- size "laboratory" for an in-depth analysis of the effects of radical institutional and tech-nological changes on new shapes of competitive structuring within markets undergoingquick globalization, and of the "spontaneous" emergence of a new industrial organiza-tion. In this respect, it appears to offer a relatively original example of industrial self-organization 1. The concept of self-organization indeed helps explain the institutional, technological and industrial dynamics from circular interactions between the firms studied and theirenvironment – which does not enable the framework of analysis based on Darwinianevolutionism (see Paulré, 1997). It thus becomes immediately possible to describe andunderstand (i) the emergence of a "global" (Paulré, 1997) or "collective" order (Stengers,1985), both selective and organized, based on "local" behaviour (Paulré, 1997), and (ii)changes (behavioural, strategic and/or organizational) in a system with respect to thenature of its interactions with its environment (Paulré, 1997). It is thus said that a system(company, organization, sector, etc.) possesses an ability for self-organization when it isable to, according to its experience and its environment (Yovits, 1962), "produce (in thepresence of a "noise" or disequilibrium), on its own, and not necessarily in a voluntary orconscious way, a structure, organization, new behaviour and/or its own rules of opera-tion" (Paulré, 1997, p. 134).
This is truly what seems to be happening today in the pharmaceutical industry, where we are witnessing at the same time the emergence of a new industry (that of new bio-technology; see Depret and Hamdouch, 2000b) and of a new model of technological 1 For a presentation and discussion on recent research contributions on the theme of self-organization, seePaulré, 1997; Lesourne and Orléan, 1999. M.-H. DEPRET AND A. HAMDOUCH innovation based on the exploration of living organisms, and a radical reconfiguration ofindustrial structures and competitive strategies of the drug industry on the whole. This article, as part of this theoretical perspective, attempts in this way to delimit the main explanatory factors of the current evolution dynamics of the pharmaceuticalindustry and to identify the prevalent traits of the new industrial organization that seemsto be coming into being. The article has been divided into five sections. The first onebriefly presents the general characteristics of the pharmaceutical industry and the struc-ture of drug markets. Section 2 analyses the new bases of the competitive game broughtabout by the changes in progress in this industry and the strategies of adaptation used bypharmaceutical companies. Section 3 then brings out the dynamic, multidimensionalnature and more and more interwoven nature of the modes of interaction amongst phar-maceutical companies, and between the latter and biotech companies. In fact, thisincreased phenomenon of interweaving illustrates not only the role, now extremelyimportant, held by interfirm cooperation but also, and above all, the emergence andquick distribution of more and more hybrid and collective configurations of interactionin the form of firm coalitions and networks (Sect. 4). Then, the modes of forming, opera-tion and the evolution of coalitions and networks indicate a broadened competitivedynamics in which the crucial determinants of market structures result from intra andinter-reticular strategic games (Sect. 5). 1. The pharmaceutical industry: Specificity and market
structures
1.1. The drug, a specific good
1.1.1. A supply under supervision and regulated all along the value chain A drug is not a good like others. Indeed, since it has an effect on the quality of life, onwell-being and on disease, death or recovery, it is assimilated with a semi-public good ora good under supervision (Buisson and Giorgi, 1997). It is, as a result, subjected to manycontrol agencies and to a complex network of national and international regulations(Richard and Senon, 1996; Walbrœck-Rocha, 1997) that concern all levels of the valuechain (Depret, 1999): R&D, through the "Good Laboratory Practices" and the "GoodClinical Practices", that determine "International Common Denomination"; classifica-tion and industrial exploitation of the drug, via the approval (A.M.M. in France, FDAApproval in USA); production, import, export and distribution, through the "Good Man-ufacturing Practices" and specific conditions of access to advertising. Innovation is thus closely structured by ethical considerations and the potential capa- bility of production and sale is limited by regulation policies regarding health expendi-ture, while the combined effect of approval procedures, rules for prescription andreimbursement policies prevent any direct confrontation between supply and demand.
Yet, however constraining they may be, statutory provisions produce, in parallel, a seriesof significant positive effects. In particular, sanitary safety encourages the quality ofsupply and the confidence of prescribers and patients, while rules of precaution are NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY inclined to be recognized by all and to be harmonized from above (see de Wolf, 1993;Buisson and Giorgi, 1997). 1.1.2. Drug demand, at the heart of health and reimbursement systems There is also a stake as regards demand for the entire sanitary channel as it relies closelyon a chain of varied intermediaries. Upstream the end consumption of a drug, whole-salers play an ever more pronounced role. At the heart of the system, the patient-generalpractitioner pair gives direction to demand, while remaining tributary to pressure exertedby financing agencies for health expenditure. Public agencies are indeed subjected to arationale to reduce the deficit of social expenditure and attempt to regulate demand by atighter control of prescriptions and by modulating the reimbursement rates. As for theprivate systems, they play on the price to be paid for services (calls for tender anddrawing up formularies). 1.2. A relatively atomised, fragmented and segmented market
1.2.1. An oligopolistic, but highly dispersed structure Worldwide pharmaceutical industry is organized around a loose oligopoly (an importantcompetitive fringe and high product differentiation), that is both stable and dynamic(Tarabusi, 1993), which means without really a new entry, but evolving internally alongwith firm strategies and the marketing of blockbusters 2 (see Tabs. 1 and 2). Indeed, even if successive M&A in the last two decades have demonstrated a gradual tightening of the industry and an increase in the market shares of the major pharmaceu-tical companies (see Tab. 3), the hard core of the oligopoly paradoxically represents onlya relatively restricted share of the world market (see OCDE, 1985; Tarabusi and Vickery,1996). 1.2.2. Market fragmentation In spite of globalization, the particular aspect of domestic pharmaceutical markets side-steps the competitive game and makes the industrial environment heterogeneous fromone country to the next. There is even greater local market specificity: national statutorypractices and health safety policies – that are either different or insufficiently harmonized–; cultural and human differences in clinical practice; heterogeneous modes of financingand consumption; etc. 1.2.3. Market segmentation Two different approaches to market segmentation can be envisaged as regards drugs. Thefirst one results from the combination of two criteria: the type of intellectual property 2 A blockbuster is generally considered as a drug with yearly sales that exceed one billion dollars. In 1999,thirty were counted, nine of which exceeded two billion dollars (see Tab. 2). M.-H. DEPRET AND A. HAMDOUCH TABLE 1. Evolution of the top 20 pharmaceutical companies.
GlaxoSmithKline (2) Bristol Myers Squibb (4) Pharmacia Corp (7) American Home Products Sanofi-Synthelabo (8) Boehringer Ingelheim (1) Sales, R&D expenses and R&D effort in billions of dollars; R&D effort as % of sales; 99 world marketshare; (2) GlaxoSmithKline = Glaxo + Wellcome + Smithkline Beckman + Beecham; (3) AstraZeneca = Astra +Zeneca; (4) Bristol Myers Squibb = Bristol Myers + Squibb; (5) Aventis = Rhône-Poulenc + Hoechst ; (6) Novar-tis = Ciba-Geigy + Sandoz ; (7) Pharmacia Corp. = Pharmacia & Upjohn + Monsanto ; (8) Sanofi-Synthélabo =Sanofi + Synthélabo; NC: Not Classified in the first twenty pharmaceutical companies before merger or acqui-sition; (e) Estimate by defect. Sources: Calculations by the Authors from Hamdouch and Depret (2000) and Pharmaceutical Executive(2000).
rights applicable (drug protected by a patent versus a patent that has fallen into the publicdomain), and the terms of sale (subject to prescription or not). It enables three types ofdrugs, for which the rules of operation on markets are highly heterogeneous (see Depret, NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY TABLE 2. The top 25 best-selling global products.
SmithKline Beecham SmithKline Beecham American Home Products Bristol Myers Squibb Monsanto (Searle) Bristol Myers Squibb Bristol Myers Squibb (1) Sales in billions of dollars.
Source: Pharmaceutical Executive (2000).
1999). On the patented ethical products market, competition is essentially upstream as itresults basically from an "innovation race" opposing R&D teams on several "therapeuticfronts". Once the product has been patented, competition on this market is therefore gen- M.-H. DEPRET AND A. HAMDOUCH TABLE 3. Evolution of the Concentration of World Pharmaceutical Industry (Lerner Indexes*)
* Lerner indexes Ci = market share cumulated of the first i companies, as % of global market. Sources: Calculations by the Authors from Hamdouch and Depret (2000) and Phar-maceutical Executive (2000).
erally not head-on; it becomes (a little more so) when the drug falls into the publicdomain. Indeed, manufacturers of generic drugs, who are not obliged to bear R&D costsand who benefit from a commercial effort of the innovative firm, only place their effortdownstream the value chain 3. They continue to be subjected, however, to rules regardingbans on advertising. Yet the latter is not imposed on manufacturers of "OTC" (over thecounter) products that can launch advertising campaigns in all media. In the latter case,competition is subjected to arbitration, no longer on the part of prescribers, but on thepart of consumers. The second approach of the market segmentation, by therapeutic category, enables the relevant markets to be detected more precisely on which technological and industrialcompetition between pharmaceutical companies takes hold. So, while in 1995 nearlytwenty thousand drugs were counted on the market (Colcombet and Aflalo, 1995), aclassification into fifteen consistent therapeutic categories (see Tab. 4) illustrates twoparticularly important observations (Depret, 1999). On the one hand, the market appearsto be highly segmented and concentrated at the same time, since the first three and fivetherapeutic categories cover respectively nearly 51% and a little over 70% of the market.
On the other hand, regardless of their size, pharmaceutical companies generally opt for arelative therapeutic specialization due to the high cost of pharmaceutical innovation(Depret, 1999). However, while large pharmaceutical companies are practically allpresent in the most important therapeutic categories, the smallest firms, on the contrary,practice very tight specialization by being positioned only on one or two promising orhighly advanced domains (Iung and Rupprecht, 1997). In fact, at the therapeutic level, 3 It is to be noted that to counter the inevitable expansion of the generic drug market (Alvarez, 1994; Martinand Rupprecht, 1998), innovative pharmaceutical companies have also entered the market by externalgrowth, by the creation of a dedicated division or by strategic alliances (Colcombet and Aflalo, 1995;Tarabusi and Vickery, 1996). NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY TABLE 4. The world pharmaceutical market by therapeutic category (1998).
Therapeutic Category Alimentary & Metabolism Central Nervous System Musculo-skeletal 5.0% Hospital Solutions Parasitology 0.2% Source: Depret (1999).
the market is characterized by a higher concentration than at the global level and by arelatively important share of the leading firm 4. Thus, the notion of a relevant market constitutes a notion of variable geometry: it depends closely on what is being considered (geographic area, nature of the drug, thera-peutic category, etc.). However, while this highly specific structuring of drug marketsdemonstrates a relative fractioning of the competitive game, the changes in progress inthis industry bring about an in-depth reconfiguration of markets and of their geographicand therapeutic boundaries. 2. New foundations of the competitive game and strategic
adaptation of pharmaceutical companies
The pharmaceutical industry is currently affected by five types of structural changes. Thesechanges, which oblige firms to make significant strategic and organizational adaptations,are creating the new foundations (or "rules") of the competitive game within this sector. 4 For example, the insulin market, controlled for over twenty years by Novo Nordisk and Eli Lilly (whichhold between two-thirds and three-quarters of the world market) is practically a duopoly (Hamdouch andDepret, 2000). Novo Nordisk and Eli Lilly indeed allow followers (such as Hoechst in Europe) only severalshares of domestic markets (Durand and Gonard, 1986).
M.-H. DEPRET AND A. HAMDOUCH 2.1. Radical changes in progress in the pharmaceutical industry
2.1.1. Rationalizing health expenditures and limiting the growth of firms The combined actions of a slowdown in growth in the 1980s, of the Engel's Law – withhealth expenditures increasing two to three times faster than income; see Colcombet andAflalo (1995) –, soaring unemployment and the "grandpa boom" are all at the origin ofthe crisis of the financing systems and of the rationalization policies of health expendi-tures begun in many countries. These policies have led to a slowdown in the growth ofpharmaceutical companies, while the pressure of the public authorities and new players(Health Management Organizations and Pharmaceutical Benefit Managers 5, patientassociations, generic drug manufacturers, etc.) limited the increase in drug prices,weighing down on the margins, on the ability for self-financing, and therefore, in fine, onthe pharmaceutical industry's ability for innovation.
2.1.2. Growing needs for innovation and increase in the critical size of R&D The institutional and financial pressure exerted on the industry is all the more strong asthe pharmaceutical companies must bear very high R&D investments (14.2% of turnoverin France; S.N.I.P., 1999) that are growing rapidly (increase in Europe of over 72%between 1986 and 1995; see Scrip Magazine, 15 January 1999), on the one hand,because the innovative processes are forever longer and more costly 6, and, on the otherhand, due to the medical needs that must be satisfied 7, from the maturity of "pipelines" 8(see Tab. 5) and the ageing of product portfolios 9, all in a context of market globaliza-tion and intensified competition. Pharmaceutical companies are thus obliged to "run" after their optimal size, i.e. the size 10 enabling them to reconcile on the one hand a relative specialization of the thera-peutic segments covered and a concentration of research products in some broad fields 5 Health Management Organizations are American private organizations in charge of managing the healthservices of companies. As for Pharmaceutical Benefit Management companies, they are management com-panies for medical prescriptions ensuring the handling of prescriptions and drug distribution (seeHamdouch and Perrochon, 1999b).
6 The average period between initial research and the sale of a product varies between eleven and fourteenyears, the average cost for a new molecule being between two and three billion French Francs, with anincreasing share devoted to development (51.5% of R&D expenses in 1997, as opposed to 35.5% in 1970),leaving 33% to applied research and 15.5% to fundamental research (S.N.I.P., 1997).
7 Also included here are cardiovascular disorders, diabetes, cancer, chronic respiratory disorders, degenera-tive diseases, psychiatric and neurological disorders, Aids, etc. 8 A pipeline can be defined as the portfolio of molecules developed by each laboratory (Spikler, 1993;Dumoulin, 1994). It is said that a pipeline is mature when it is made up of a great number of molecules in aphase of clinical development.
9 Indeed, half of the one hundred most widely sold drugs in the world must fall into the public domain bythe year 2005 (see Beynon, 2000). 10 As a reminder, size seems to present four advantages in terms of R&D (Schumpeter, 1961; Fisher andTemin, 1973; Panzar and Willig, 1981; Cohen and Levinthal, 1989; Mairesse and Mohen, 1990; Foray andLe Bas, 1991; Henderson and Cockburn, 1997): (i) economies of scale and scope; (ii) diversification ofresearch efforts; (iii) favoured access to financial markets; (iv) spillovers.
NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY TABLE 5. Top 20 R&D pipelines (Number of active products − May 1999).
No. And % of Drugs 126 (61.2%) 80 (38.8%) 79 (53.4%) 69 (46.6%) 81 (54.4%) 65 (45.4%) 67 (54.9%) 55 (45.1%) American Home Products 43 (46.2%) 50 (53.8%) 64 (71.9%) 25 (28.1%) 53 (59.6%) 36 (40.4%) 43 (53.7%) 37 (46.3%) 39 (52.7%) 35 (47.3%) 69 (94.5%) 30 (41.9%) 43 (58.9%) 36 (50.0%) 36 (50.0%) Boehringer Ingelheim 48 (70.6%) 20 (29.4%) Bristol Myers Squibb 32 (47.1%) 36 (52.9%) 44 (65.7%) 23 (34.3%) 43 (72.9%) 16 (27.1%) 37 (62.7%) 22 (37.3%) 21 (36.2%) 37 (63.8%) 36 (67.1%) 17 (32.1%) Source: Calculations by the authors from Pharmaprojects (2000).
(maximum exploitation of economies of scale and scope), and, on the other hand, theneed to hold more dispersed investments with a highly uncertain return (Cockburn andHenderson, 1998), but necessary in order to maintain a sufficient "absorptive capacity"(Cantner and Pyka, 1998; Cohen and Levinthal, 1989, 1990). 2.1.3. Market globalization and the opening of "new frontiers" While the critical size of R&D is increasing, this is also due to the fact that the relevantmarkets are henceforth continental or transcontinental, and will be tomorrow, in a thera-peutic category, ever increasingly global. The world market is estimated at 334 billiondollars (1999), and expected to reach 435 billion dollars in 2003 and 690 billion dollarsin 2010 (IMS, 1999). It is characterized by strong growth (6 to 8% per year) and the pre-ponderance of developed countries (85% of the market). However, new markets areemerging (Latin America, Asia, Central and Eastern Europe). The Asian market is M.-H. DEPRET AND A. HAMDOUCH TABLE 6. The world pharmaceutical market by regional market.
Regional Market of which Eastern Europe & CIS REST of the WORLD of which Latin America & Caribbean of which Middle East * Estimate; ns: not significant; n/a: not available. Sources: S.N.I.P. (1999); Eurasanté (1999a); Colcombet and Aflalo (1995).
expected to undergo the highest growth rate of the next century (11% per year), particu-larly the Chinese market (18–19% per year by the year 2004; see IMS, 1999). These paths represent real opportunities for firms seeking renewed sources of growth.
Indeed, as cost and the importance of experience in R&D limit the opportunities fortherapeutic diversification outside the specialization base, essentially (internal and/orexternal) growth on new markets may finance the future R&D of pharmaceutical compa-nies. The opening of borders, at the same time, demonstrates the arrival of outsiders andincreased market fragmentation. Therefore, pharmaceutical companies must manage thespecificity of the new markets, while remaining vigilant with respect to new competitorsin markets where they are already present. 2.1.4. Towards a new paradigm of pharmaceutical innovation This is where the changes in progress are undoubtedly the strongest (see Depret andHamdouch, 2000b). The irruption of new biotechnology (or third generation biotech-nology) indeed draws the contours of a new technological paradigm that is quicklyshaping the innovation processes and the competitive dynamics of the drug industry(based on the notion of technological paradigm, see Dosi, 1982). Today, if new biotechnology techniques applied to health are presented as an alterna- tive to the chemistry paradigm (Chesnais, 1981) and to methods that have however beenrationalized (Dumoulin, 1994), it is because they enable the development of more effec-tive drugs that are less expensive and launch on the market more quickly and, more par-ticularly, because they enable intervention no longer on the effects of a pathology but onits causes (Eurasanté, 1999a; Depret and Hamdouch, 2000b). Thus, the patient can betreated in two different but complementary ways: (i) by acting "directly" on the gene orprotein to be "repaired", thanks to what is known as DNA-based therapy; (ii) by NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY improving current treatments or by creating new drug solutions thanks to the toolsoffered by "technological platforms" 11.
New biotechnology applied to health in this way draw the new "technological fron- tiers" that superpose the new "geographical frontiers" brought about by globalization.
They are indeed at the origin of the introduction and generalization of new innovationroutines that impose the creation and gathering of new competences. In fact, pharma-ceutical innovation can today be considered as the result of ever more knowledge in dif-ferent disciplines and sectors induced by the convergence of the technologicaltrajectories of the pharmaceutical industry, biotechnology, information technology (bio-informatics) and electronics (biochips). 2.1.5. Towards broader coordination amongst the players The multidisciplinary nature of the knowledge and competences required obliges phar-maceutical companies to make the contributions and pressure exerted from new indus-trial players part of their strategy, particularly biotech companies. Their responsiveness,flexibility and critical "savoir-faire" therefore help pharmaceutical companies to makeever-closer ties.
At the same time, other institutional players are appearing, also changing the rules of the game. What can be seen is the growing role of retirement funds and big businessinvestors who limit room for decisions on the part of company leaders, sometimesobliging them to make radical strategic choices 12. "Worked" in an endogenous and exogenous manner by wide-scale technological, eco- nomic and institutional changes, the pharmaceutical industry is thus marked by a radicalredefinition of its structural determinants and its competitive dynamics. How do pharma-ceutical companies face the new "rules of the game"? Are they content with adapting tothem by limiting the less favourable consequences, or are they, on the contrary, seekingto channel them or even to redefine them to their best advantage? 2.2. New rules of the game or game on the rules?
Contrary to what may have been initially thought, the combination and the quasi-simul-taneity of the structural changes described above were not at the origin of a total and 11 In biotechnology, five technological platforms can be considered: (i) genomics, for a better understan-
ding of diseases so as to identify the pharmacological targets on which new drugs will be able to act;
(ii) Bioinfomatics which enables computer-aided modeling and the "in silico" identification of drug candi-
dates; (iii) biochips, to pave the way for innovation in the field of diagnostics and in the discovery of new
drugs, but above all for an accelerated understanding of their modes of action; (iv) combinatorial chemistry
and high throughput screening, which give researchers a method for drug creation at a speed ten thousand
times higher and at a cost ten times less than the traditional blind screening, by creating chemical diversity
that was not fathomable only a few years ago; (v) finally, "drug delivery", for better screening of drug
action, with greater therapeutic efficiency and/or reduced side effects (see Depret and Hamdouch, 2000b).
12 This is how Hoechst, for example, had to comply with the terms of its main shareholder (Kuwaiti KPC),
that called for (and obtained) both an acceleration of the merger process with Rhône-Poulenc, a double split
(Hoechst-Celanese and Rhône Poulenc-Rhodia), and the setting up of the new Group (Aventis) in Stras-
bourg. See L'Expansion 607, Oct. 1999, pp. 92-94.
M.-H. DEPRET AND A. HAMDOUCH erratic questioning of the structure of markets within the pharmaceutical industry. Thissituation, seemingly paradoxical, can be explained in fact, not by a limited impact ofthese changes or by a weak influence of the new industrial or institutional players, butrather by the remarkable effort in adapting pharmaceutical companies to the new techno-logical, industrial and competitive rules of the game in this sector. Indeed, as already demonstrated (see particularly Delapierre et al., 1998), it appears that while an insider manages to impose its rules, it temporarily keeps control and cancreate conditions for making the competitive regime it masters last, particularly withrespect to outsiders 13. Firms wishing to enter into business must then, either meet allthese rules (anterior position of biotech companies), or modify or create new ones, byattempting to, for example, have an innovation adopted that it has itself initiated (as it istoday the case for biotech companies). In the case of the pharmaceutical industry, this game on the rules between the insiders (pharmaceutical companies), and between the latter and the different outsiders (biotechcompanies and new institutional players) includes the commitment of pharmaceuticalcompanies in strategies of adaptation as well as the radical reconfiguration of the indus-trial and competitive industrial organization of the drug industry on the whole. 2.3. Imperative organizational and strategic reorientations
2.3.1. Radical strategic refocusing on core businesses After having attempted to develop through diversification and vertical integration –which explains the excessive dispersion of investment and research efforts –, pharma-ceutical companies have been developing over several years split strategies betweenchemical and pharmaceutical businesses, and more recently between pharmaceutical andagrochemical businesses, as well as strategies for the transfer of business activitiesdeemed as non-strategic. These strategies (see Perrochon, 2000) mark the end of the ver-tically integrated pharmaceutical company, as it seeks, develops, produces and sells onlyits own drugs. Today, this new direction appears all the more imperative as the arrival or increase in the power of the new industrial and institutional players exert additional pressure onpharmaceutical companies at three main levels: (i) exacerbated competition and a con-traction of margins; (ii) increased dependency of pharmaceutical companies with respectto public or private funding bodies through different systems of contractual regulation;(iii) and finally, consumption more and more guided by individual care costs and bypublic prevention policies. 13 In the pharmaceutical industry, barriers to entry are three-fold: statutory (see Sect. 1), technical and eco-nomic (necessary competence for the exploitation of pharmaceutical innovation), and commercial – whenwe know that only blockbusters are sufficiently profitable to be able to bear all the costs pertaining to thedevelopment and setting up of production capacities and the required sales resources. So, several studiesdemonstrate that investments only begin to be covered after the patent has expired (see for example:Grabowski and Vernon, 1990, 1994; Guillot and Simon, 1991).
NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY 2.3.2. Towards a rationalization of organizational structures While the global strategy must be reviewed in depth, the same applies for each link ofthe value chain that pharmaceutical companies are seeking to rationalize to the full.
Upstream, we are witnessing a double movement of therapeutic refocusing and decen-tralization – or even the outsourcing – of R&D activities. At the production level, phar-maceutical companies, have set up reduction programmes and site specializations, andhave accentuated resorting to custom-made production of specialties. Finally, down-stream, we are witnessing a double organizational and strategic evolution: on the onehand, marketing has been reinforced as the first budget item (24% of turnover since thebeginning of the 90s; see Tarabusi and Vickery, 1996), with the emergence of new pro-motional tools (co-marketing, co-promotion, etc.); on the other hand, the distributionchain has been rationalized as a result of the development of generic substitution, self-medication and concentration of the European market (Eurasanté, 1999b; Depret, 1999). Beyond these strategic and organizational changes, the defence and conquest of market shares means redefining the modes of interfirm interaction and adopting complex andchanging strategic combinations for these modes. 3. Nature and evolution of interfirm relations
in the pharmaceutical industry
The evolution in the relations between pharmaceutical companies, and between the latterand biotech companies, perfectly illustrates the new strategies of interfirm interaction inthe sectors undergoing wide-scale technological, institutional and economic changes.
Indeed, while in the drug industry these relations have always been marked by diversity,evolution and the combination of modes of interaction, today, they are becoming morecomplex as they are part of increasingly hybrid and dynamic configurations. 3.1. Diversity of the modes of interfirm interaction
To define the nature of interfirm relations in the pharmaceutical industry, according toHamdouch and Perrochon (1999b), we consider that a continuum of dynamic config-urations of interaction in which head-on rivalry, cooperation and integration/concentra-tion constitute only "generic attractors" of interfirm interaction trajectories.
3.1.1. "Relativized" competition on markets Within the pharmaceutical industry, direct competition exists only in a loose form,for two main reasons. Firstly, the existence of "industrial organization networks" 14 14 This notion must not be confused with the informal networks of top executives described by Froelicher(1996).
M.-H. DEPRET AND A. HAMDOUCH (national, transnational or international 15) ensures the close institutional coordinationbetween pharmaceutical companies that share, at least partially, "common interests" 16.
Secondly, the relative specialization of pharmaceutical companies in some therapeutic areas, and the highly uncertain nature of innovation in this field make it improbable fortwo pharmaceutical companies conducting research on the same pathology, with thesame approach to the solution, and with, in the end, two galenic forms with identical sideeffects, to launch the product on the market at the same time. Rivalry on pharmaceuticalmarkets thus appears to be essentially guided by the nature of the pathology and thetreatment, by the patient's case history, by the prescriber's and patient's habits, and moreand more by generic substitution. 3.1.2. Integration strategies vs. outsourcing and/or cooperation strategies In the pharmaceutical industry, taking the example of most industrial or service sectors,external growth strategies have been developing considerably over the past years, in suc-cessive waves 17 (see Tab. 7). In this sector, the motivations for such movements arenumerous (see Depret and Hamdouch, 2000a): an increase in the R&D critical size,research into technological, productive or commercial synergy, the stabilization orincrease in market shares in a globalized environment; and more tacitly: the eventualexpiry of a patent on a blockbuster while the pharmaceutical company has no alternativenew product, the need to create value for the shareholders, or yet the attenuation of thestrategic consequences resulting from a poor technological or commercial former choice.
Theoretically, the current M&A wave in the pharmaceutical industry can be explained in two ways. As for the context, it corresponds to a "natural" strategic reversal on thepart of firms. Thus, in an industry undergoing change, cooperative strategy is no longersuited to the quick technological, productive and commercial changes required (heavyinterfirm coordination, project blockage, difficulty in establishing alternative partner-ships which are strategically more advantageous, etc. (see Hamdouch and Perrochon,1999b).
More structurally, based on the new-institutional approach, cooperation is considered as the sole solution to a problem of uncertainty (Quéré and Ravix, 1996). It then appears 15 In France, pharmaceutical companies are organized within the National Pharmaceutical Industry Union(S.N.I.P.), member of the European Federation of Pharmaceutical Industry Associations (E.F.P.I.A.) and theInternational Federation of Pharmaceutical Manufactors Associations (I.F.P.M.A.).
16 These networks finance regular studies on the evolution of markets, legislation and technologies, orga-nize conferences and colloquia on a wide number of topics, and adopt common positions with participatingfirms so as to be heard by representative authorities and consumer associations. 17 While from 1988 to 1992, over two hundred cross-border M&A operations were carried out in the phar-maceutical industry for a total amount of 13.1 billion dollars (Tarabusi and Vickery, 1996), they are today ata much higher scale, above all due to the mechanical increase in the size of pharmaceutical companiesbrought about by successive concentration operations. It is also important to stress that current operationsbear essentially on mergers between firms of a comparable (significant) size: Ciba-Geigy and Sandoz(Novartis), Astra and Zeneca, Hoechst and Rhône Poulenc-Rorer (Aventis), Sanofi and Synthélabo, and oncurrent operations (see Tab. 7). NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY TABLE 7. Major mergers and acquisitions deals in pharmaceutical industry since 1988.
New identity Eastman Kodak (USA) Sterling&Winthrop (USA) Smithkline Beeckman (USA) Bio Science Labs (USA) Smithkline Beeckman Boehringer Ingelheim (Ger) Boehringer Ingelheim Nordisk (Danmark) Merrell Dow (USA) Marion Merrell Dow Smithkline Beeckman (USA) Smithkline Beecham Bristol Myers (USA) Bristol Myers Squibb Institut Merieux (France) Connaught (Canada) Pricordio AB (Sweden) Pharmacia AB (Sweden) Rhône-Poulenc (France) Rhône-Poulenc Rorer Merck&Co (Switzerland) du Pont de Nemours (USA) Sterling Drug Europe (USA) American Cyanamid (USA) American Cyanamid Pharmacia (Sweden) Montedison (Italy) (1) Warner Lambert OTC (USA) Wellcome OTC (GB) Warner Lambert OTC Hoechst (Germany) Roche (Switzerland) Smithkline Beecham (USA/ Diversified Phar.Serv. (USA) Smithkline Beecham American Cyanamid (USA) Sterling Winthrop (USA) (2) Smithkline Beecham (USA/ Sterling Winthrop (USA) (3) SmithKline Beecham Eastman Kodak Clinical Johnson&Johnson (USA) Ciba (Switzerland) PCS Health Systems (USA) Sandoz (Switzerland) Hoechst Marion Rous- Hoechst (Germany) Marion Merrel Dow (USA) Pharmacia AB (Sweden) Pharmacia& Upjohn Rhône-Poulenc Rorer (France) Fisons (GB) Rhône-Poulenc Rorer Continued on next page. M.-H. DEPRET AND A. HAMDOUCH TABLE 7. (continued).
New identity Ciba-Geigy (Switzerland) Sandoz (Switzerland) Roche (Switzerland) Boehringer Mannheim (Ger) Hoechst (Germany) Rhône-Poulenc (France) Synthélabo (France) Sanofi- Synthélabo Celltech Chiroscience (GB) Celltech Chiroscience Roberts Pharmaceutical Shire Pharmaceuticals (GB) Shire Pharmaceuticals Pharmacia&Upjohn (Swe/ Pharmacia Corporation GlaxoWellcome (GB) SmithKline Beecham (GB) Warner-Lambert (USA) (1) Pharmaceutical division only; (2) Prescription drugs; (3) OTC division only; (4) AHP: American Home Pro-ducts ; (5) In billions of US dollars; (6) In billions of French francs. Source: Hamdouch and Depret (2000).
as a hybrid, transitory and discontinuous mode for the discovery and coordination of newex post exploitable industrial opportunities by the partners. Thus, once the result of coop-eration is known (i.e. lifted uncertainty), economic coordination must be analysedthrough the standard market/organization dichotomy. In reality, generally, as in the pharmaceutical industry (Hamdouch, 1998; Depret, 1999), cooperation is a form of interfirm interaction of both a general – as it is practicedby all pharmaceutical companies and all mature biotech companies, for all the links inthe value chain – and a permanent nature, as all pharmaceutical companies have been orare, at present, involved in at least one alliance or partnership 18.
18 Based on available data (Windhover's Pharmaceutical Strategic Alliances 2000, in PhRmA, 2000), threemajor phases have been identified in the evolution of many strategic alliances in the pharmaceutical indus-try since 1986: a regular increase in 1986 (121 alliances) until 1990-91 (respectively 319 and 324 opera-tions); strong growth in 1992 (428 alliances) and stabilisation at a high level up to 1995 (over 445 allianceson average per year); and finally, a high increase in 1996 (577 operations) and 1997-98 (700 and 712 ope-rations respectively). An in-depth analysis of the content of the alliances shows that (i) at each period, thenumber of alliances increased convexly with respect to the number of M&A's, thus, invalidating the ideaaccording to which cooperation and integration-concentration are totally alternative modes; and (ii) thatalliances change through time. In this way, we can see a gradual change from upstream cooperation tointermediary and more and more downstream cooperation, which explains above all the growing ability ofbiotechnology to conduct co-development and co-marketing programmes with large pharmaceutical compa-nies as well as amongst themselves. NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY According to several recent studies, the thirty leading world pharmaceutical compa- nies are all involved on average in forty cooperation operations (Andersen Consulting,1997), representing from 10 to 20% of research budgets (PriceWaterhouse Coopers,1999) and 20 to 40% of income (Boston Consulting Group, 1999). Furthermore, it is tobe noted that while nearly seven research programmes out of ten were carried out inter-nally by pharmaceutical groups in 1995, research contracts and partnerships should rep-resent nearly 45% of total programmes in 2005 (Boston Consulting Group, 1999).
Finally, while the share of research conducted externally in the total R&D budget variesconsiderably according to the pharmaceutical company, this ratio is expected to quicklyexceed 25%, reaching about 30 to 40% in the years to come (see Eurasanté, 1999b, andsupra Tab. 5). This data thus moves away from the problematic of interfirm relations proposed by the transaction cost theory as, while R&D activity in the pharmaceutical industry is highlyspecific in essence, pharmaceutical companies do not hesitate to outsource or to carry outa rather substantial part of work in a cooperative manner (see Fig. 1 and Eurasanté,1999b).
In fact, the current increase in capitalistic mergers in the pharmaceutical industry can mainly be explained by technological change and market globalization and, more pre-cisely, by the persistent and amplified uncertainty with which pharmaceutical companieshave been confronted. Thus, the increase in size permitted by M&A seems to be less ofan objective in itself than a means for seeking new partners so as to master the competi-tive regime brought about by the structural changes in progress in this industry (Depretand Hamdouch, 2000a). 3.2. Changing modes of interaction
While the modes of interfirm interaction are diverse and multidimensional, they alsoevolve, partly due to the strong uncertainty and irreversibility that structurally make upthe drug industry all along its value chain 19, and, on the other hand, to the strategiesadopted by pharmaceutical companies so as to adapt to new rules within a context of"structural uncertainty" (Langlois, 1984). Firms must thus use all the strategic tools available, that they combine differently from their competitors and in a way that evolves in time, as generally speaking, there is no"one best way" when it comes to organization and strategy. As a result, four periods canbe observed as regards the evolution of relations between pharmaceutical and biotechcompanies (see Box 1). 19 In this industry, uncertainty relates to the nature of the request, legislation, techniques, know-how andmodes of R&D organization, as well as to the ex ante evaluation of promising projects and the future possi-bilities of strategic interaction. They are reinforced by the more or less strong situations of irreversibility,beyond some stages of the innovation process and sale (target discovery, patenting a molecule, authoriza-tion to launch a new product on the market, or the imposed or decided withdrawal (switch) from the mar-ket, etc. See Hamdouch and Perrochon (1999a).
M.-H. DEPRET AND A. HAMDOUCH Fig. 1. Organization of the value chain of the pharmaceutical industry.
This temporal approach of interfirm relations underlines two main points. On the one hand, the actual combination of modes of interaction differs according to a pharma-ceutical company's speciality and size, to the period and the partner(s) concerned, as wellas to the "maturity" of the firm's cooperation relations. Thus, cooperation is part of a lifecycle as it always occurs at a time when it either switches to an integrated structure, NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY BOX 1. The four stages of relations between the pharmaceutical industry and bio-
technology.
From the early 70's to the mid-80's, we have been witnessing a "wait and see" atti-tude that is all the same attentive to pharmaceutical companies with respect to thebiotechnology that are developing essentially in an isolated manner.
In the second half of the 80's, pharmaceutical companies became aware of the truecognitive potential of biotechnology (that advance "technically" and whose salesforecasts became more realistic) and then jumped into a frenzy of alliances, capital-istic mergers and research internalization.
In the 90's, products from biotechnology became more usual. Focus was then placedon sales, and property rights became a stake. We thus witnessed re-specializationand re-focussing strategies, and the formation of close ties with biotech companies.
Today, the new "rules" of the game are beginning to be felt. We then observe amovement, slow perhaps but clearly perceptible, of companies becoming graduallyautonomous with respect to pharmaceutical companies or, more precisely, of theinterdependence between partners that are able to "live" separately but prefer tointeract closely as this solution is quicker and more effective for the developmentand sale of new products. Sources: Raugel, 1993; Sharp et al., 1994; Maupertuis, 1997; Delapierre et al.,1998; Depret, 1999; Depret and Hamdouch, 2000b.
definitively stops, or is finally renegotiated (Hamdouch and Perrochon, 1999a). But, onthe other hand, this switch does not make cooperation "in general" a hybrid and tempo-rary form of interfirm relations. On the contrary, it constitutes a permanent and substan-tial form of interaction between rival and/or complementary firms (see Hamdouch, 1998).
It is thus distinguished from its particular modes of operation (alliances, agreements andpartnerships), which are necessarily transitory and discontinuous (Depret, 1999). 3.3. Interwoven modes of interfirm interaction
The exploration and exploitation, at the highest and with the lowest cost, of the complexprocess of pharmaceutical innovation, as well as management of the uncertainty/irrevers-ibility pair, are generally only possible by adopting "sequential strategies combiningmovements of commitment or, on the contrary, those of waiting, experimentation,consolidation or redeployment, or even withdrawal" (Hamdouch and Perrochon, 1999a).
This also above all includes the necessary complex combination of the different modesof interfirm interaction, as shown in the example of the agreement signed between theAmerican pharmaceutical company Abbott and the French biotech company Genset.
This agreement is indeed characterized by the combination of a "dose" of cooperation(by its complementary nature), of an equity participation (Abbott in Genset), and ofmaintaining a true rivalry between both partners (that basically remain competitors M.-H. DEPRET AND A. HAMDOUCH within the market 20). It, in this way, clearly demonstrates the consubstantial nature ofthe different modes of interfirm interaction (Hamdouch, 1998) and characterizes typi-cally what we call a coalition.
In fact, beyond their diversity, and their interweaving and basically changing nature, the modes of interfirm interaction in the pharmaceutical industry are today characterizedby the emergence of highly original competitive and cooperative combinations, particu-larly in the form of highly varied bilateral or multilateral coalitions and of firm networks(or coalition networks). 4. Coalitions and networks: towards a radical reconfiguration
of interfirm relations in the pharmaceutical industry
Faced with the problems presented by the development of new products and changes inmarket rules, the traditional shapes of competition and interfirm cooperation are losingmuch of their pertinence. Hybrid and more collective forms of interaction such as coali-tions and networks are becoming the favoured modes of interaction in this type ofindustry, as they make up the essential vectors of configuration of the dynamics of actualcompetition and of the future structuring of organizations and markets (Aoki, 1990;Hagedoorn and Schakenraad, 1994; Kotabe and Swan, 1995; Hamdouch, 1998, 2000).
Within a context of increasingly globalized competition, each firm must therefore buildand exercise (or simply belong to) coalitions and networks sufficiently powerful anddurable so as to stay – and possibly dominate – in the innovation race and the struggle formarket shares.
4.1. The nature, characteristics and shapes of coalitions and networks
in the pharmaceutical industry
The increasing number of uses of the notions of coalition and network gives them apolysemic nature that is particularly diversified, thus turning them into true "interdisci-plinary metaphors" (de Bresson and Amesse, 1991). They must be defined correctly soas to discern their main characteristics and shapes in the case of the pharmaceuticalindustry.
20 By way of this agreement, signed in 1997, Genset undertakes to identify the markers and genes associa-ted with the response to treatments proposed by Abbott (we speak of "pharmacogenomics"). The resultswill then be used by Abbott, which will develop, produce and sell the resulting diagnostics systems. Inreturn, Abbott undertakes to comply with the four following obligations: to finance company research byGenset for a period of eighteen months; to make payments according to the progress of programmes (miles-tones); to pay a fee on the sale of pharmacogenomic diagnostics; to enter into immediate shareholding of5% of the capital of Genset, as well as a future option in additional shareholding (Hamdouch and Depret,2000). It is to be noted that this principle has once again been selected by the two partners within the scopeof an agreement signed in March 2000 in the field of genetic diagnostics of manic-depressive psychosis andType II diabetes. NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY 4.1.1. Firm coalitions Based on the theory of games (see Von Neumann and Morgenstern, 1944), the notionof coalition has been covered in many works in areas such as sociology, politics andeconomy (Lemieux, 1998). In Industrial Economics, coalitions are often confused withthe notions of strategic alliance, agreement or even interfirm cooperation 21. Byborrowing from the fields of sociology 22 and economics, we offer a definition that isbroad enough to be able to take account of the variety of interfirm relations and themultiple combinations of forms of interaction observed in the pharmaceutical industry.
Coalition designates then here bilateral or multilateral relations for the commonachievement of a given activity, through the control, exchange or sharing of informa-tion, know-how, knowledge, as well as (intermediary or finished) products and/orcapital. Within this scope, three major types of coalitions can be observed (Hamdouch, 2000), as illustrated clearly in the current dynamics of the pharmaceutical industry: • Vertical coalitions (vertical partnerships in research, clinical development, production or sale, and, alternatively, vertical integration operations), initiated by one or severalfirms or structures that are complementary or potentially complementary; • Horizontal coalitions, the most commonly used, formed by competitive or potentially competitive firms, against rival firms and against suppliers, clients or producers ofcomplementary goods. We speak here of strategic alliances upstream (R&D alliances),intermediary (co-development, co-production) and/or downstream (co-promotion, co-marketing, product outsourcing), or, alternatively, of M&A's or equity participations(crossed or not); • Transversal coalitions, more rarely used, which bring together firms from different sec- tors of businesses that are (actually or potentially) involved in the preparation and saleof goods-systems 23.
21 By reminding of its contribution to the analysis of coalitions (Porter and Fuller, 1986), Porter (1993) thusspeaks almost indifferently of strategic alliance and coalition (pp. 72 and 591), of interfirm understanding(pp. 72) or of interfirm cooperation (pp. 648). Within this scope, coalition is defined as a long-term agree-ment (between two or several companies working in the same industry), going beyond simple commercialtransactions (without however being a merger), mainly in the form of joint ventures, licence exploitationagreements or supply agreements, but finally as way to take on a competitive advantage.
22 In sociology, the notion of coalition applies to relations between all types of organizations. It defines aconcerted and temporary group of individual or collective players that together exercise relations of coope-ration and conflict, and seek, by structuring the appropriate power, to dominate their opponents in order toobtain more significant advantages than if they were not part of the coalition (Lemieux, 1998).
23 For example, we think of coalitions between biotech companies and microelectronics or computer com-panies (Motorola, IBM, Hewlett-Packard, MIT, Perkin Elmer, Hitachi, etc.) or between pharmaceuticalfirms and companies specialized in the e-business (Healthon, PlanetRx.com, Drugstore.com, etc.). M.-H. DEPRET AND A. HAMDOUCH 4.1.2. Interfirm networks At the same time, the network is generally considered in economic literature as either:• a mode of coordination (Béjean and Gadreau, 1997) or organization of economic acti- vity (Joly and Mangematin, 1995; Dufourt, 1995) and interfirm relations (Massard,1997), intermediate between the market and hierarchy 24; • or a collective entity (Joly and Mangematin, 1995), specific and organized (Dufourt, 1995), whose behaviour cannot be reduced to that of some of its components (Jolyand Mangematin, 1995), and whose analysis cannot be reduced to that of the degrada-tion of the polar forms of organization, which are market and organization (Dufourt,1995) 25.
In reality, this distinction is not always evident, as shown in the case of interfirm pharma-ceutical networks. An interfirm network must thus be considered as a form of interfirminteraction enabling the coordination of activities and as an "institutional creation"(Dufourt, 1995; Lazega, 1996). More precisely, in the pharmaceutical industry, interfirm networks correspond to a form of organization of the productive process that is specific, dynamic, generalizedand relatively continuous. It brings together a group of heterogeneous players 26(Callon, 1991), with know-how and/or specific and/or complementary resources, inter-connected by informational, economic, financial and/or human links 27 so as to meetobjectives that are above all of an individual nature (Depret, 1999) 28. The players in anetwork are mutually interdependent within the scope of a generally long and costlyinnovation process. This interdependence then determines the stability and evolution 24 This conception of an industrial, technological or technical-economic network has been strongly inspiredby the theoretical new-institutional framework (see Hamdouch and Depret, 2000). For a discussion andmany applied approaches, see among others: Thorelli (1986); Powell (1990); Aoki (1991); Callon (1991,1994); Imai and Baba (1991); Maillat et al. (1991); Planque (1991); Axelsson and Easton (1992); Dufourt(1995); Freeman (1995); Joly and Mangematin (1995); Béjean and Gadreau (1997). 25 The network must thus be considered as a whole, as a "dual form of organization'' (Joly and Mange-matin, 1995) and as a producer of standards and regulations (Dufourt, 1995). 26 Here, we speak of pharmaceutical companies, biotech companies, service companies, universities, aca-demic or public research centres, foundations, etc. 27 Callon (1991) speaks of "intermediaries" that he defines as all that revolves around the players (codifiedknowledge, technical objects, currency, skilled individuals and know-how) and that constitute the form andcontent of relations that are set up between them (Callon, 1999). 28 Synthetically speaking, through the network, partners search (i) complementary activities (Richardson,1972), assets (Dosi et al., 1990; Garrette and Dussauge, 1991) and/or external (Hakanson and Johanson;1988) and specific resources (Bouabdallah and Dufourt, 1994); (ii) synergy between relatively specialized(Guilhon and Gianfaldoni, 1990) and operationally autonomous partners (Perrin, 1991), within a networkwhose management is codified (Maillat et al., 1991), collective and interactive (Perrin, 1991); (iii) the willto reduce uncertainty being linked to decision-making in a complex, transitory world that is constantlychanging (Cook, 1977; Pfeffer and Salancik, 1978); (iv) conveying information and fragmented knowledge(know-how) that can be at the origin of new competence and new knowledge. NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY of the network 29. In this way, on the one hand, each partner is relatively in need of theinformation of other partners. On the other hand, very few participants are able to breakthe bonds that have been established or they lose all credibility when seeking new part-ners (Depret, 1999). A network must therefore be understood over a period of time.
Indeed, while in the short term the network can be justified by a functional rationale ofactivity connexion, in the long term, the notion of relations between players comes intoplay (Filippi et al., 1996), based on technological and relational learning processes (DeBresson and Amesse, 1991; Bidault, 1993; Cohendet and Llerena, 1993), and on confi-dence and reputation (Callon, 1991; Granovetter, 1985; Koenig and van Wijk, 1992; Jolyand Mangematin, 1995; Massard, 1997; Depret, 1999). The nature of interfirm relations in the drug industry relies on four key dimensions. The first one is linked to the therapeutic nature of the network. Indeed, coalitions generallycover a wide therapeutic spectrum, ranging for example from the fight against cardiovas-cular diseases 30 to more rare connections specialized in "orphan diseases" (see Depret andHamdouch, 2000c). Networks thus tend to regroup firms or research centres conductingcomplementary research programmes, i.e. directed toward common or similar pathologies. The second dimension is related to the technology used. With the emergence of a new paradigm for pharmaceutical innovation, this dimension is destined to become increas-ingly important. In matters of biotechnology, DNA-based therapies can be distinguishedfrom technological platforms (see supra § 2.1.). While the first ones are substituted forthe final product, the second ones are complete so as to improve the innovation process.
Technological coalitions thus aim to complete therapeutic coalitions 31. The third dimension depends on the degree of market maturity. Indeed, the more mature the market is, the easier it will be for a firm to approach it individually. On thecontrary, collective strategies and coalitions tend to increase when the market in questionis insufficiently mature (biochips, proteomics, pharmacogenomics, gene therapy, etc.), or 29 It is said that a network is both stable and changing (Guilhon and Gianfaldoni, 1990; Maillat et al., 1991)when the players that form it cannot be changed, replaced, moved, without harming its uniqueness, flexibilityand malleability, or without questioning its ability for action and adaptation to outer stimuli (Rivière, 1997).
This is all the more true since, in the case of the pharmaceutical industry, the network is guided by the sameobjective, that of innovation. Yet the (pharmaceutical) innovation process is not linear in nature (Callon,1994). Consequently, a change in the network is closely linked to the project of innovation (Maillat et al.,1991). The network is built and transformed (Guillon and Gianfaldoni, 1990; Massard, 1997) according to theconstraints with which the players are confronted in their approach to innovation. It therefore takes on diffe-rent forms in time. Its main elements change and are organized differently according to a recurrent process(Guilhon and Gianfaldoni, 1990), but the network globally maintains its identity.
30 To cite an example, the Swiss company Roche has had, for several years now, a network of alliances cen-tred around American (Myriad Genetics, Centocor, EnzyMed, Focal, American Biogenetic Sciences, CVTherapeutics, Scios) and Canadian biotech companies (Nortran Pharmaceuticals), either alone or throughSyntex and Genentech, its autonomous biotech subsidiaries. 31 Here, we refer to the Gencell gene therapy network (see Box 2) as well as, to cite an example, the Ame-rican Merck network in combinatory chemistry, for several years centred on pharmaceutical (Abbott, Aven-tis, DuPont, Merck KGaA,, Astra and Sumitomo) and biotech companies (Argonaut Technologies, AriadPharmaceuticals, Genetics Institute, Dyax, 3-Dimensional Pharmaceuticals, EnzyMed, ISIS Pharmaceuti-cals and Axys Pharmaceuticals). M.-H. DEPRET AND A. HAMDOUCH hardly accessible (China, India, etc.). Forming coalitions and networks thus enablespharmaceutical companies to solve or attenuate the dilemma of tight/broad specialization(Depret, 1999) within the framework of strategies for renewing portfolios, going aroundbarriers as regards entry or strategic watch. New therapeutic and/or geographic marketsthus become more accessible to a greater number of firms. Finally, the fourth dimension is related to geographically locating partners. While net- works are generally bringing together partners that are relatively close geographically(Genet, 1997), more "complex" networks are also made up of partners seeking to benefitby the conditions offered as a result of the partners' differences in economic, scientificand statutory environments (see supra § 1.2). 4.1.3. The link between coalition and network in the pharmaceutical industry The notions of coalition and network come at least partially together. Indeed, coalitionsnaturally aim to take on a reticular form. The coalition-network link is thus brought tocome about in two complementary ways: reticular coalition (coordination in a networkof specific activities of individual firms) in which the notions of network and coalitionare confused, and the coalescent network (coordinated set of different coalitions 32whose members wish to benefit by the advantages of reticular organization).
Within this framework, it is possible to see today's pharmaceutical industry as a new industrial organization characterised by an interconnection of coalitions and networksin which each pharmaceutical company seeks to develop and defend the network(s) towhich it belongs. A pharmaceutical company forms coalitions with other pharmaceuticalcompanies, as well as with biotech companies, service providers and research centres,which in turn are forming coalitions with pharmaceutical companies, biotech companies,service providers, research centres, etc. 4.2. Outline of a taxonomy of networks in the pharmaceutical industry
In this industry, networks correspond to four basic configurations, whose varied formsand examples are shown in Figures 2 and 3. 4.2.1. Coordinated networks The coordinated network is made up of several bilateral coalitions between differentpartners working on common themes or issues. Three different types of coordinated net-works can be observed: vertical, when a pharmaceutical company forms a coalition withseveral biotech companies, service companies and/or academic research structures 33; 32 In fact, two firms can very well not be linked by a single coalition, while they are members of the samenetwork. For this, the two firms need only to have a common partner. 33 To cite an example, the American company Bristol Myers Squibb has had a network of alliances forseveral years now specialized in the fight against cancer centred around academic research centres andessential American biotech companies (Millennium Predictive Medicine, Somatix, Oxigene, Liposome,Seattle Genetics, Cydex, Neose, Research Triangle Institute, Cytoclonal Pharmaceutics, Genzyme Transge-nics Genetic Therapy, University of Texas, Progenics Pharmaceuticals, Supergen, EntreMed, Somatix The-rapy), as well as Canadian (University of Manitoba), Spanish (PharmaMar), Swedish (KarolinskaInstitute), British (Chiroscience) and Japanese (Taiho) biotech firms. NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY aphic modeling of networks in the pharmaceutical industry A gr
.
. 2
g
Fi
M.-H. DEPRET AND A. HAMDOUCH e 2 for le network of networks" xemple of An e
.
. 3
g
Fi
NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY inverted, when several pharmaceutical companies form a coalition with a biotechcompany 34, a service company (see Eurasanté, 1999b) or an academic research structure(see Cassier, 1995); and finally horizontal, when several pharmaceutical companies (orbiotech companies) form coalitions together 35.
4.2.2. Multilateral networks The multilateral network is made up of several partners, generally without a single orcentral coordinator. This includes the traditional multilateral network – binding three orfour partners to a new theme or to one that is not promising enough to be part of internalresearch 36 –, the collective consortium ( between several partners 37 that exchange infor-mation or coordinate their R&D activity together (such as SNP 38; see Fig. 2) –, and theindustry organization network (see supra § 3.1.). 4.2.3. Integrated networks The integrated network is centred round a core and satellites (see the "gene therapy pool"for example of Rhône-Poulenc-Rorer (RPR) represented in Fig. 2). Here, collaborationis not limited to only the complementary nature of competences. The central company ofa network in fact sets into motion a dynamics for mutual learning between the membersso that the product of interaction cannot be obtained (as quickly, in most cases) if thepartners are satisfied with only exchanging (vertically or bilaterally) their competences,know-how, techniques, etc. (Depret, 1999). 4.2.4. The network of networks Assembling this reticular puzzle into a cooperative network is ever more difficult todecode, as the networks are multiplying and becoming interwoven (for a simplified 34 In this way, the French biotech company Genset has set up a network of alliances (with Sanofi/Synthe-labo, Pharmacia&Upjohn, Janssen (Johnson&Johnson), Wyeth-Ayerst (American Home Products) andAbbott) so as to broaden competence, therapeutically – while working in different areas (Type II diabetes,manic-depressive psychosis, schizophrenia, etc. – and technologically, by developing programmes of phar-macogenomics, identification and genetic cloning, diagnostics and vaccination. 35 In the first case, we are thinking of the network set up by the American company Pfizer with Parke-Davis(Warner-Lambert), Searle (Monsanto), Eisei and Fujisawa, so as to co-promote Lipitor and Celebrex, aswell as to work with the Japanese in the areas of Alzheimer's disease, cancer, and osteoporosis (see Fig. 2).
In the second case, the network in gene therapy was made up of, in the 90's of Genetic Therapy, TargetedGenetics, GenVec, Canji, Introgen, Genzyme and Inclone Systems centred round the American NationalInstitutes of Health (Depret and Hamdouch, 2000c), as one example.
36 For example, we are speaking of the alliance formed in May 1998 by American Home Products, Molecu-lar Dynamics and Amersham Pharmacia Biotech in the field of DNA chips (biochips).
37 Generally speaking, the partners of the research consortium specialize within sub-programmes. Forexample, the European consortium on lipases was subdivided into five sub-programmes, thus forming asmany coalitions (see Cassier, 1998; Cassier and Foray, 1999).
38 The latter is made up of eleven bio-pharmaceutical companies, five American and British researchcentres and three computer and electronic companies (see Fig. 2) so as to prepare a detailed map of thegenome using natural markers known as "snips" (Rosier, 1999). M.-H. DEPRET AND A. HAMDOUCH example, see Fig. 3). We then speak of a cooperative network, i.e. a network of coordi-nated, multilateral and integrated networks that interpenetrate (Depret, 1999). Forexample, RPR today is involved in over one hundred and seventy coalitions, more thanone-quarter of which are considered as major (RPR 1999 Activity Report and diversedata taken from Recap Database). In addition to the coalitions forming the Gencell net-work (see Box 2), RPR has in fact signed agreements in different fields, on different linksof the value chain and in different reticular organizations. 5. The emergence of a new competitive dynamics
The quick changes in interfirm relations towards increased structuring into coalitions andnetworks are demonstrated by an in-depth reconfiguration of the competitive process.
The competitive game seems in this way to be extended to new "zones" of rivalry out-side the market, moving from a strict inter-individual level towards more and more "col-lective" interactions within a play of intra and inter coalitions and networks of firms. Theprocess for the setting up and evolution of these hybrid structures, which are both com-petitive and cooperative as a result, represent the major strategic stakes for bio-pharma-ceutical companies. Thus, analytically speaking, it is possible to observe at least four levels of competitive and cooperative games implied in the reticular rationale: (i) formation of coalitions andnetworks through a rationale of potential partner's pre-emption; (ii) consolidation of coa-litions and attraction of new partners; (iii) rivalry within the coalitions and networksformed, more or less consolidated; (iv) competition between coalitions and networks.
Here, we are just presenting an outline of the content of these games through the currentchanges in the pharmaceutical industry, a more general analysis of the rationale forforming coalitions being part of more formal studies, particularly in the case of standard-ization processes (see Katz and Shapiro, 1994; Foray, 1995; Hamdouch, 2000; etc.). Theexample of the network of networks in gene therapy illustrates the different types ofgames centred round the reticular approach (see Fig. 3 and Box 2).
5.1. The "race" towards forming coalitions and networks
The first game level, highly upstream the competition process in markets, involves dif-ferent firms in a sort of "race" towards forming coalitions, enabling each one to pre-emptthe partners that are necessary for its network. In this way, each new coalition aims tomake its network more consistent and effective, therapeutically, technologically and geo-graphically. The pre-emptive strategy, prevailing in the formation of coalitions and networks results from the fact that the "stock" of strategically decisive potential partners is rela-tively restricted. This strategy covers three major dimensions:• The first-mover is in a position to attract the best partners. By doing so, it reduces (in number and often also in "quality") the stock of potential partners for coalitions andnetworks that can form successive followers. Forming coalitions and networks has NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY therefore a formative and locking character, above all when the principle ofexclusivity 39 plays a major part (see Hamdouch, 2000); • The first coalitions and networks to be formed, and which are organized more rapidly at the operational level, become the most attractive ones for potential partners not yetunited. The successive members of a coalition or network thus bring about a cumula-tive process of an "increasing coalition returns" type, a bit in the image of the"increasing adoption returns" principle of a technology formalized by Arthur (1988); • Finally, the early formation of quickly structured coalitions or networks may discou- rage the formation of alternative coalitions or networks if the followers feel that thegame is bound to be lost. In a looser form, this configuration resembles the ε-pre-emp-tion strategy described in some models of patent race (see particularly Dasgupta andStiglitz, 1980). However, this can, inversely, incite followers to strengthen their effortsso as to join forces in an attempt to catch up and possibly exceed the leading coalition(leapfrogging strategy described by Fudenberg et al., 1983). This notion of competitive coalition can be illustrated by the formation of Gencell (see Fig. 3 and Box 2). This network is mainly centred on the central nervous system, cardio-vascular diseases and oncology. The partners are either specialized in one predeterminedtherapeutic direction, or integrated by their specific competence and know-how in genetherapy. Gencell thus enables RPR not only to have access to new competences andknow-how – which are sources of crossed fertilization and significant technological andindustrial synergy –, and to benefit by the conditions offered by the competitive distor-tions of the different economic, scientific, statutory environments of the partners, butalso to create a true mobility barrier, by reducing the number of potential partners for itscompetitors, and by gaining precious time in the innovation race. 5.2. The "race" for reticular consolidation
The second level of the game can be illustrated by strategies of attraction-confinement ofthe pre-empted partners or of those likely to reinforce the network. This is a question ofattracting companies still belonging to other networks (consolidation strategy) and, sym-metrically, defending oneself against the risk of having partners flee to other competitivenetworks (stabilization strategy). This power of attraction-confinement on the part of afirm will depend on the quality and variety of its network, of the partner's degree ofdependence with respect to the reticular core, and of the strategic nature of the partner. This game configuration can be illustrated by the recent confrontation between the American pharmaceutical companies Pfizer and American Home Products (AHP) overthe control of Warner-Lambert (WL). Thus, as AHP and WL revealed the terms of theirmerger, Pfizer announced its will to take over WL, for an exchange of shares (of a valueof 82.4 billion dollars). By overbidding by over 14% with respect AHP's bid, the manu- 39 When a firm enters into a coalition, it is generally prohibited for it, at least for a certain period, to takepart in an alternative coalition in the same area (existence of contractual and or irreversible financial short-medium-term commitments, difficulties as regards organization in terms of managing several programmesat the same time, etc.). M.-H. DEPRET AND A. HAMDOUCH BOX 2. The strategic games within and around the Gencell network.
Formed in 1994 by Rhône-Poulenc Rorer (RPR), Gencell is an integrated network of at least thirty-four
partners federated around gene therapy and directed towards therapeutic areas within the scope of its spe-
cialization (Staropoli, 1998; Rhône-Poulenc Rorer, 1998). Gencell plays a major role in the formation of
the RPR coalition network and is even able to illustrate the link between the different intra or inter-retic-
ular games.
(i) The formation of coalitions and networks and the pre-emption of the most strategic partners: before
1994, RPR took part in many French and European biotechnology programmes, such as BioAvenir, for
which it was the coordinator (Staropoli, 1998). This experience encouraged technological, therapeutic
and relational synergies, reduced any uncertainty linked to gene therapy and participated in making
RPR's reputation. All these factors naturally encouraged the creation of Gencell, as the main members of
BioAvenir joined in (CNRS, Pasteur Institute, Gustave Roussy Institute, etc.).
(ii) Competitive coalitions and the games of attraction-confinement of pre-empted partners or those
likely to strengthen the network: these games can be illustrated by the way the members of the network
interact with other networks. Therefore, of all the Gencell members, only eight have succeeded in
forming coalitions with other partners, within the "sub-network" (Vical, Transgène) and/or outside with
other "sub-networks" (Transgène, Duke University, Vical, Genetix Pharmaceuticals, Stanford, Introgen
Therapeutics, Baylor Medical Center, Johns Hopkins University). While the analysis has been extended
to areas not directly related to gene therapy, the partners of RPR-Gencell can be placed into at least three
categories: those who did not wish to form significant coalitions outside Gencell and gene therapy
(Genetix Pharmaceuticals, Genopoietic, Supratek, Transgène, University Louis Pasteur of Strasbourg,
Virogenetics, IntraImmune Therapies, NARA Institute of Technology, Xenon Bioresearch, University of
Cagliari); those who have succeeded in forming limited number of coalitions (CNRS, Darwin Molecular,
Endocyte, Enzon, Généthon, Curie Institute, Gustave Roussy Institute, Introgen Therapeutics, Lawrence
Berkeley Laboratory, Oxford BioMedica, Vical, St Elizabeth's Medical Center, Bichat Hospital of Paris,
Biosense); and finally, those who have succeeded in moving away from the "centripetal force" of the net-
work (Duke University, Pasteur Institute, LXR Biotechnology, Oncormed, Stanford, University of British
Columbia, University of Alabama, BASF, Sequus (Alza), Baylor Medical Center, John Hopkins Univer-
sity).
(iii) Intra-reticular competition is also as present in the scope of Gencell, even if, in this case, it does not
concern a fight for leadership, while that of RPR is in essence practically indisputable. The fact that the
partners of RPR-Gencell are either specialized in a predetermined therapeutic direction (cardiovascular
diseases, diseases of the central nervous system, oncology), or, on the contrary, integrated for their own
competence and specific know-how in gene therapy, in fact adds a dose of competition in the coordina-
tion of the network since the areas of research are relatively convergent.
(iv) When rivalry within a network is quite localized, the competition between networks appears to be
stronger. Indeed, the study of the different types of interaction taking place in gene therapy enables light
to be shed on four "sub-networks" that are relatively consistent and competitive, inasmuch as the mem-
bers of a "sub-network" tend to remain faithful to it (see Fig. 3):
• In the northwest quarter of the network of networks in gene therapy, the sub-network is dominated by
RPR and its partners; • In the northeast quarter, the sub-network is largely made up of biotech companies that interact around "pivot firms" (NIH, Genzyme, Introgen), and in parallel form, for the most part "buffer firms" betweendifferent sub-networks (Introgen, Targeted, New York University, Immune Response, Duke University,Genetic Therapy). This sub-network can then be considered as complementary to the other sub-net-works.
• In the southeast quarter, the sub-network is dominated by Novartis (and its subsidiaries) around a great number of biotech companies. • Finally, in the southwest quarter, the biotechnology company Transkaryotic succeeds in federating bio- tech companies but also pharmaceutical companies such as Roche, P&U, AHP and Hoechst. However,it is to be noted that the latter sub-network is the least homogeneous of the four. It can thus be consider-ed as an extension of the other sub-networks.
NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY facturer of Viagra launched the greatest hostile operation in the history of the pharma-ceutical industry 40.
The power of attraction-confinement thus constitutes a powerful lever for the consoli- dation and reinforcement of coalitions and networks. But, at the same time, it demon-strates a sort of mobility barrier 41 for the one that has it. The latter could indeed find itdifficult to move away from its partner, either because the contract does not provide forany real exits or because its reputation may risk being affected by it (see below).
5.3. Intra-reticular competition
While strategic interactions between pharmaceutical companies come into play upstreamin a dynamics of network pre-emption-attraction-consolidation-stabilization, they alsocome into play within each network. Indeed, consolidation-stabilization strategies areonly possible through the existence of relations of authority and confidence. A fight forthe leadership and for sharing gains from the coalitions then come into play within eachnetwork. The ability for a member of a network to enter into the reticular core, or even impose itself as leader, will depend mainly on its size, specificity and know-how, on its reputa-tion and cooperation experience, on the degree of maturity of the market concerned, onthe nature of interdependence between those who have formed coalitions, and on thedegree of specialization and autonomy of the partners. In this way, the model leader (ifthere is one) would be a pharmaceutical company, a member of the hard core of the oli-gopoly, dominating its therapeutic market, with true expertise in its field thanks to thecoalitions it was able to develop and pursue with the most advanced complementarycompanies, and knowing how to grant wider autonomy to the most powerful and/oruseful partners. In fact, the fight for leadership and the fight for sharing gains of the coalition appear to be linked. Indeed, sharing and the control of gains of the coalition 42 depend above all on 40 This operation can be explained for three main reasons: (i) Pfizer and WL were already in the midst ofnegotiations, even if the latter came up against the degree of parity of the merger; (ii) Pfizer, who had nottaken part in the recent waves of capitalistic mergers – preferring to reorganize its R&D activities "inhouse", to reinforce its sales teams and to develop diversified alliances and partnerships –, had the strengthand ambition to take on new markets; (iii) the creation of AmericanWarner by AHP and WL questioned theco-marketing agreement concerning Lipitor (a cholesterol drug) signed by Pfizer and WL in 1996. Howe-ver, this highly profitable drug (turnover of nearly 3.7 billion dollars in 1999; see Tab. 2) was headed for abright future, with a potential market estimated at nearly ten billion euros. The threat of seeing WL leavethe Pfizer network for that of AHP seemed unacceptable for Pfizer. The latter then brought its tacit pre-emptive right into play, despite WL management's preference for a merger between equals with AHP. 41 There are two reasons for this mobility barrier: (i) the importance of the size of the bodies and the R&Dcosts, which often prevent two alternative programmes from being run in the same field at the same time;(ii) confidence and experience, which very frequently cause partners to renew their coalition when theobjectives have been reached. 42 Gains from coalition can take on three main forms: (i) a patent, a licence or a principle agent; (ii) amethod or know-how; (iii) a scientific, technical, commercial, legal, financial, cultural and organizationalknowledge. M.-H. DEPRET AND A. HAMDOUCH the negotiation power of each member of the network and on the evolution of thepower struggles between coalitions. More precisely, the ability for a leader to reallyaffect the allocation of gains of the coalition between members of the network willdepend on many factors, particularly, the functionality of the network 43, the nature anddegree of industrial or scientific application of the results, its own ability to absorbresults, the dynamic convergence of objectives, the legal (i.e. contractual), competitiveand organizational environment, the nature of decisions taken and the methods for set-tling conflicts and, above all the climate of confidence existing within the network (seeSako, 1991; Koenig and Van Wijk, 1992; Mothe, 1996).
5.4. Inter-reticular competition
Finally, the fourth level of game represents the competition between coalitions orbetween networks in different markets (geographical and/or therapeutic). Thus, two rivalpharmaceutical companies are often seen entering into coalitions, each on its own, withbiotech companies, for example, to prepare for the next generation of products. At thesame time, a pharmaceutical company often enters into parallel coalitions with differentpartners, generally biotech companies or research centres, to bring them into competitionand to pre-empt them to the detriment of other pharmaceutical companies or coalitions.
Competition between networks takes on a more complex form. To illustrate this com- petition, it is important to observe, for an instant, the race for decoding the humangenome which opposes all the players in the sector 44. The stake in this race is indeedsignificant when one knows that, from the identification of new targets for future drugsto understanding the mechanism of action, including the diagnosis of genetic predisposi-tion to certain pathologies, all require gene identification (Depret and Hamdouch,2000b). In the hope of winning (or not being outdistanced in) this race, pharmaceuticalcompanies have two possible strategies. The first one consists in forming a coalition withone or several genomic companies (to form a coordinated network), but the extent ofthe task and the need to move quickly in decoding, make this strategy unsuitable. Bycontrast, the second strategy seems to be more suitable: generally, pharmaceuticalcompanies prefer belonging to consortia, and thus by moving competition to the inter-reticular level. Finally, it is necessary to stress that the existence and nature of inter-reticular competi- tion, the degree of market maturity and the degree of network complexity are connected.
Thus, the more mature the market concerned with the reticular rationale (traditionalmarket), the less the firms form coalitions together, and the more the competition tends 43 It is said that a network is "monofunctional" when its activities have been clearly determined a priori,"multifunctional" when it integrates the entire innovative process (Planque, 1991), or "hybrid", when itsfinal and intermediary functionalities cannot be defined a priori (Depret, 1999). 44 In this area, three major international consortia are being considered: (i) Human Gene Consortium (pri-vate consortium with Human Genome Sciences, Smithkline Beecham, Schering Plough, Takeda, MerckKGaA, Synthélabo); (ii) SNP (see Fig. 2); (iii) Human Genome Sequencing Consortium (public consor-tium). It is to be noted that these three consortia are involved in a fight that has opposed them to (structural)genomic companies such as Celera (run by Craig Venter, a renegade of the public sector) or Incyte.
NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY to be at between firms. On the contrary, the less the market is mature (emerging or poten-tial market), the more coalitions are natural and networks complex, and the more compe-tition moves towards the coalition and reticular level. Conclusion
Under the conjugated impact of radical and rapid technological, institutional and eco- nomic changes, the industrial and competitive organization of the pharmaceuticalindustry is being reconfigured deeply and durably. While the innovation processes arebeing structured more and more round the new biotechnological paradigm and the mar-kets are becoming globalized and fragmented, pharmaceutical companies tend to adapt atthe strategic and organizational levels, above all by refocusing their activity and by rede-fining all the links of their value chain. But pharmaceutical companies' strategies areundergoing the most significant changes with respect to the conception of their hori-zontal and vertical relations. As we have attempted to demonstrate, the innovationprocess and competitive dynamics are structured increasingly around firm coalitions andnetworks (complementary and/or rival) and around the varied institutional players, forwhich formation and consolidation according to a pre-emptive approach come into playupstream the markets. We are thus witnessing the emergence of new shapes of work division in this industry in terms of R&D and of the coordination of activities between players, of course, actu-ally or potentially rivals, but obliged to cooperate in order to innovate, survive and pos-sibly dominate in markets that are forever more globalized and fragmented, which arepromising but also highly uncertain. In fact, while cooperation now involves strategicnecessity and no longer tactical choice, this is because it constitutes the central directionfor competition between the different types of firms and organizations. The drug industryindeed seems to prefigure a new type of competitive principle in the sectors subjected toradical and quick changes, according to which "innovating to compete" is no longer whatcounts, but "competing to cooperate and innovate together… to stay in the innovationrace and the struggle for market shares" that constitutes the key to market competitionand domination.
Based on this hypothetical but probable foundation of the emergence of a new techno- logical and competitive dynamics, relatively robust and stable, it is possible to attempt tooutline the changes to come about in the pharmaceutical industry:• Re-focussing strategies are expected, both at the business and therapeutic levels, to continue. Within this scope, it is not at all certain that the combination of a commit-ment in the "Life Sciences" and a broader specialization strategy (the case of Aventis)will be a fruitful one, as demonstrated by the many recent examples of disinvestments(particularly that of Novartis and AHP in the agrochemical industry); • Globalization is expected to pick up speed. Then, the penetration of large pharmaceu- tical groups in new markets can only come about by forming collaborations with all thelocal partners (institutions, firms, etc.); • The increasingly sophisticated technology, requiring ever more mastery of different disciplines and sectors, will make it definitively utopian for a pharmaceutical company M.-H. DEPRET AND A. HAMDOUCH to alone mobilize all the competences, know-how and resources necessary to imple-ment the innovation process. Thus, even leaders with solid commercial ground will notonly have to systemize "cooperative deals", but also accentuate the formation of capi-talistic coalitions to strengthen and reinforce technological and industrial networks,which explains the probable continuation of M&A operations between large pharma-ceutical companies; • But the pharmaceutical company that is integrated vertically, seeking, developing, pro- ducing and selling its own drugs is destined to disappear. The pharmaceutical companyof tomorrow will most likely be a firm whose final aim will be above all to sell, underits own brands, drug solutions (developed in house, through collaborations, subcon-tracted, or purchased on the market) so as to respond to a great number of new or insuf-ficiently treated pathologies; • The pace of innovation, the progress in knowledge of living organisms and the purely "immaterial" nature of the assets of biotechnology make a wave or acquisitions orabsorptions of biotech firms by pharmaceutical companies quite unlikely (see Depretand Hamdouch, 2000a). We will therefore be witnessing more and more a confirmationof "free form of association" that will be "passionate", reasoned and well-balanced,between pharmacy and biotechnology, in what may tomorrow become more of a"healthcare industry" than a drug industry, particularly with the possibilities of cloningand cell, genetic or proteomic manipulation, as it already seems to be coming about; • Within this framework, the exploration and exploitation of this "new biotechnology economy" will only be optimal under three conditions (Depret and Hamdouch, 2000b).
Firstly, the different parties concerned must urgently rehabilitate the time factor ofR&D and remove it from a nearsighted view of the market. Secondly, the institutionalframework must be consistent, harmonized and revisable, but predictable. Finally, thepharmaceutical industry must become more involved in biotech research throughtighter, more lasting and more balanced collaboration, with biotech companies and uni-versity laboratories, above all by multiplying and consolidating the dense scientific andindustrial networks. Acknowledgments. The authors would like to thank Bernard Paulré and two anonymousreferees for their pertinent remarks and suggestions as well as Eurasanté, the non-profitagency dedicated to economic development for biotechnology & healthcare related activi-ties in the Nord-Pas-de-Calais region, for its financial support in this research. Of course,we are solely responsible for any possible error or omission that may remain in the text.
Alvarez F. (1994) Le marché mondial des médicaments génériques. Eurostaf, Paris.
Andersen Consulting (1997) Alliances: the velvet invasion & the dual challenge, Mimeo, 23 p.
Aoki M. (1991) Economie japonaise : information, motivations et marchandages. Economica, Arthur W.B. (1988) Competing Technologies: an overview. In: Dosi G. et al. (Eds.), Technical change and economic theory. Pinter Publishers, London and New York.
NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY Axelsson B., Easton G. (1992) Industrial networks: a new view of reality. Routledge, London.
Béjean S., Gadreau M. (1997) Concept de réseau et analyse des mutations récentes du système de santé, Revue d'Économie Industrielle 81, pp. 77-97.
Beynon K. (2000) Feeling confident in the countdown to patent expiry, Scrip Magazine 86,
Bidault F. (1993) Apprentissage et réseaux, Économies et Sociétés W-1, pp. 79-101.
Boston Consulting Group (1999) The pharmaceutical industry into its second century from
serendipity to strategy, Mimeo.
Bouabdallah K., Dufourt D. (1994) Le problème de l'évaluation de la performance des arran- gements institutionnels : l'exemple des réseaux. XLIIIème Congrès annuel de l'AssociationFrançaise de Science Économique, Paris.
Buisson J.P., Giorgi D. (1997) La politique du médicament. Montchrestien, Paris.
Callon M. (1991) Réseaux technico-économiques et irréversibilités. In: Boyer R. et al. (Eds.), Les figures de l'irréversibilité en économie. Presses de l'E.H.E.S.S., Paris.
Callon M. (1994) L'innovation technologique et ses mythes, Gérer et Comprendre, March, Callon M. (1999) Le réseau comme forme émergente et comme modalité de coordination : le cas des interactions stratégiques entre firmes industrielles et laboratoires académiques. In:Callon M. et al. (Eds.), Réseau et coordination. Economica, Paris, pp. 13-64. Cantner U., Pyka A. (1998) Performance économique et technologique comparée des stratégies d'innovation : un modèle de simulation dynamique de concurrence monopolistique. In:Bernard J., Catin M. (Eds.), Les conditions économiques du changement technologique.
L'Harmattan, Paris, pp. 63-94.
Cassier M. (1995) Les contrats de recherche entre l'université et l'industrie : l'émergence d'une nouvelle forme d'organisation industrielle, Socio-economics Thesis, École des Mines, Paris,605 p. Cassier M. (1998) Le partage des connaissances dans les réseaux scientifiques : l'invention de règles de « bonne conduite » par les chercheurs, Revue Française de Sociologie XXXIX,
pp. 701-720.
Cassier M., Foray D. (1999) La régulation de la propriété intellectuelle dans les consortiums de recherche : les types de solutions élaborées par les chercheurs, Économie Appliquée LII,
pp. 155-182.
Chesnais F. (1981) Biotechnologie et modification des structures de l'industrie chimique : quel- ques points de repère, Revue d'Économie Industrielle 18, pp. 218-230.
Cockburn I., Henderson R. (1998) Absorptive capacity, coauthoring behavior and the organiza- tion of research in drug discovery, Journal of Industrial Economics XVLI, pp. 157-182.
Cohen W.M., Levinthal D.A. (1989) Innovation and learning: the two faces of R&D, Econo- mics and Statistics 47, pp. 182-190.
Cohen W.M., Levinthal D.A. (1990) Absorptive capacity: a new perspective on learning and innovation, Administrative Science Quarterly 35, pp. 128-152.
Cohendet P., Llerena P. (1993) Apprentissage organisationnel et cohérence : l'importance éco- nomique du réseau. Colloque : Limitation de la rationalité et constitution du collectif, Cerisy,June.
Colcombet E., Aflalo C. (Mercer Management Consulting) (1995) Industrie pharmaceutique : enjeux stratégiques et perspectives 1995-1998. Les Echos Editions, Paris. Cook K.S (1977) Exchange and Power in Networks of Interorganizational Relations, Sociolo- gical Quarterly 18, pp. 62-82.
M.-H. DEPRET AND A. HAMDOUCH Dasgupta P., Stiglitz J.E. (1980) Uncertainty, industry structure and the speed of R&D, Bell Journal of Economics 3, pp. 1-28.
De Bresson C., Amesse F. (1991) Networks of innovators: a review and introduction of the issue, Research Policy 20, pp. 363-379.
De Wolf P. (1993) L'industrie pharmaceutique. In: Sachwald F. (Ed.), L'Europe et la globalisa- tion : acquisitions et accords dans l'industrie. Masson/IFRI, Paris, pp. 305-352.
Delapierre M. et al. (1998) Régimes de concurrence, stratégies de coopération et dynamique des structures industrielles. In : Vinokur A. (Ed.), Décisions Économiques. Economica, Paris,pp. 55-80.
Depret M.H. (1999) Nature des relations interfirmes, processus d'innovation et évolution des structures industrielles : le cas de l'industrie pharmaceutique, Mémoire de D.E.A, Facultédes Sciences Économiques et Sociales, Université de Lille I, 226 p.
Depret M.H., Hamdouch A. (2000a), F&A dans la pharmacie : les sociétés de biotechnologies sont-elles menacées ?, Biotechnologies & Finances 12, March, pp. 6-7.
Depret M.H., Hamdouch A. (2000b), L'économie des nouvelles biotechnologies, Biofutur 200,
Depret M.H., Hamdouch A. (2000c), Pharmacie et Biotech : l'ère des réseaux, Biofutur 203,
Dosi G. (1982) Technological paradigms and technological trajectory: a suggested interpreta- tion of the determinants and directions of technological change, Research Policy 11, 3,
pp. 147-62.
Dosi G., Teece D.J., Winter S. (1990) Les frontières des entreprises : vers une théorie de la cohérence de la grande entreprise, Revue d'Économie Industrielle 51, pp. 27-51.
Dufourt D. (1995) Introduction, Économies et Sociétés W-2, pp. 5-15.
Dumoulin J. (1994)) Innovations pharmaceutiques et réglementation : le paradigme de l'expli-
cation, Économie Appliquée XLVII, pp. 101-126.
Durand T., Gonard T. (1986) Stratégies et ruptures technologiques : le cas de l'industrie de l'insuline, Revue Française de Gestion, November-December, pp. 89-99. Eurasanté (1999a) Biotechnologie et santé : enjeux stratégiques et perspectives dans le Nord- Pas-de-Calais. Loos, 93 p.
Eurasanté (1999b) Les nouveaux partenariats de l'industrie pharmaceutique : l'état de l'indus- trie pharmaceutique dans le monde et l'externalisation croissante de ses métiers. Loos, 88 p.
Filippi M. et al. (1996) Quelles approches économiques pour la notion de réseau, Revue d'Éco- nomie Industrielle 77, pp. 87-98.
Fisher R.M., Termin P. (1973) Returns to Scale in Research and Development: What Does the Schumpeterian Hypothesis Imply ?, Journal of Political Economy 81, pp. 56-70.
Foray D. (1995) Coalitions and Committees: how users get involved in information technology (IT) standardization. In: Hawkins et al. (Eds.), Standards, Innovation and Competitiveness.
Edward Elgar, Aldershot, pp. 192-212.
Foray D., Le Bas C. (1991) Economie de la recherche industrielle. In: Arena R. et al., Traité d'Économie Industrielle, Economica, Paris, pp. 626-645.
Freeman C. (1995) Networks of innovations: a synthesis of research issues, Research Policy 25,
pp. 499-514.
Froelicher T. (1996) L'émergence des réseaux d'entreprises : du réseau de dirigeants à la coopé- ration interentreprises. In: Noel A., Koenig C., Koenig G. (Eds.), Perspectives en manage-ment stratégique, Tome 4, Economica, Paris, pp. 309-336. NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY Fudenberg D. et al. (1983) Pre-emption, leapfrogging and competition in patent races, Euro- pean Economic Review 1, pp. 3-31.
Garrette B., Dussauge P. (1991) Alliances stratégiques, mode d'emploi, Revue Française de Gestion 85, pp. 4-18.
Genet C. (1997) Quelles conditions pour la formation des biotechnopoles : une analyse dyna- mique, Revue d'Économie Régionale et Urbaine 3, pp. 405-424.
Gilbert R. (1989) Mobility barriers and the value of incumbency. In: Schmalensee R., Willig R.D. (Eds.), Handbook of Industrial Organization. North-Holland, Amsterdam, Vol. 1,pp. 476-535.
Grabowski H.G., Vernon J.M. (1990) A new look at the returns and risks to pharmaceutical R&D, Management Science 36, pp. 804-821.
Grabowski H.G., Vernon J.M. (1994) Returns to R&D on new drug introductions in the 1980s, Journal of Health Economics 13, pp. 383-406.
Granovetter M. (1985) Economic action and social structure: the problem of emdebdedness, American Journal of Sociology 91, pp. 481-510.
Guilhon B., Gianfaldoni P. (1990) Chaînes de compétences et réseaux, Revue d'Économie Industrielle 51, pp. 104-112.
Guillot F., Simon P. (1991) Coût de recherche et développement d'une nouvelle molécule, Analyse Financière 3, pp. 30-35.
Hagedoorn J., Schakenraad J. (1994) The effect of strategic technology alliances on company performance, Strategic Management Journal 15, pp. 291-309.
Hakanson H., Johanson J. (1988) Formal and Informal Cooperation in International Industrial Networks. In: Contractor F.J., Lorange P. (Eds.), Cooperative Strategies in InternationalBusiness, Lexington Books.
Hamdouch A. (1998) Concurrence et coopération inter-firmes, Économie Appliquée LI,
Hamdouch A. (2000) Normalisation, coopération et stratégies concurrentielles. In: Lantner R., Foray D., Fauchart E., Hamdouch A. (Eds.), Normes et compétition internationale. Econo-mica, Paris (forthcoming). Hamdouch A., Depret M. H. (2000), Coalitions industrielles, réseaux de firmes et dynamiques de concurrence-coopération dans les secteurs en cours de globalisation : le cas de l'industrie
pharmaceutique, Cahiers de la Maison des Sciences Économiques 2000-24, Université de
Paris I – Panthéon-Sorbonne, 50 p.
Hamdouch A., Perrochon D. (1999a) Les dynamiques d'interaction entre firmes dans les sec- teurs intensifs en R&D : un cadre d'analyse des principales configurations théoriques. Col-loque de l'ADIS : Coopération industrielle : diversité et synthèse, Sceaux, May. Hamdouch A., Perrochon D. (1999b) Processus d'innovation, nature des relations inter-firmes et dynamique sectorielle : le cas de l'industrie pharmaceutique. Working paper CRIFES-
MATISSE 97, Université de Paris I, February, 17 p.
Henderson R., Cockburn I. (1997) Taille des entreprises et productivité de la recherche du sec- teur pharmaceutique. In: Jacobzone S. (Ed.), Économie de la santé : trajectoires du futur,INSEE-METHODES 64-65, Economica, Paris, pp. 186-210.
Imai K., Baba Y. (1991) Systemic innovation and cross-border networks, transcending markets and hierarchies to create a new techno-economic system. In: Technology and productivity:the challenge for economic policy, O.E.C.D., Paris.
IMS (1999) Forecast of global pharmaceutical market growth. (www.ims/insight/global). M.-H. DEPRET AND A. HAMDOUCH Iung N., Rupprecht F. (1997) Productivité de la recherche et rendements d'échelle dans l'indus- trie pharmaceutique française, Document de travail Insee D3E, G97/16.
Jacquemin A. (1994) Capitalisme, compétition et coopération, Revue d'Économie Politique 104, pp. 501-515.
Joly P.B., Mangematin V. (1995) Les acteurs sont-ils solubles dans les réseaux ?, Economies et Sociétés W-2, pp. 17-50.
Katz M., Shapiro C. (1994) Systems Competition and Network Effects, Journal of Economic Perspectives 8, pp. 93-115.
Koenig G., Van Wijk G. (1992) Alliances interentreprises : le rôle de la confiance. In: Noël A. et al. (Eds.), Perspectives en Management Stratégique. Economica, Paris, Vol. 2.
Kotabe M., Swan S. (1995) The role of strategic alliances in high-technology new product development, Strategic Management Journal 16, pp. 621-636.
Langlois R. (1984) Internal organization in a dynamic context: some theoretical considerations.
In: Jussawala M., Ebenfield H. (Eds.), Information and Communication Economics: NewPerspectives. North-Holland, Amsterdam.
Lazega E. (1996) Arrangements contractuels et structures relationnelles, Revue Française de Sociologie XXXVII, pp. 439-456.
Lemieux V. (1998) Les coalitions : liens, transactions et contrôles. P.U.F., Paris.
Lesourne J., Orléan A. (1998) Advances in Self-Organization & Evolutionary Economics, Economica, Paris. Maillat D. et al. (1991) Réseaux d'innovation et dynamique territoriale : un essai de typologie, Revue d'Économie Régionale et Urbaine 3-4, pp. 407-432.
Mairesse J., Mohnen P. (1990) Recherche-Développement et productivité : un survol de la litté- rature économétrique, Économie et Statistique 237-238, pp. 99-108.
Martin E.; Rupprecht F. (1998) Les enjeux des médicaments génériques, Économie et Statis- tiques 312-313, pp. 55-66.
Massard N. (1997) Externalités, apprentissage et dynamique technologique : un retour sur la notion de réseau. In: Guilhon B. et al., Economie de la connaissance et organisations : entre-prises, territoires, réseaux. L'Harmattan, Paris, pp. 336-355.
Maupertuis M.A (1997) Nature et rôle de la coopération interfirmes dans le processus de chan- gement technologique : l'exemple de la biotechnologie. In: Froehlincher T., Vendemini S.
(Eds), Connivences d'Acteurs, Contrats, Coopération inter-entreprises et métamorphoses desOrganisations, Presses Universitaires de Nancy, pp. 131-149.
Mothe C. (1996) L'appropriation des résultats au sein de consortia de R&D. In: Noël A. et al. (Eds.), Perspectives en Management Stratégique, Tome 4. Economica, Paris, pp. 381-406. O.C.D.E. (1985) L'industrie pharmaceutique : questions liées aux échanges. OCDE Publi- cations, Paris.
Panzar J., Willig R. (1981) Economies of Scope, American Economic Review 71, pp. 268-272.
Paulré B. (1997) Evolutionnisme contemporain et auto-organisation, Économie Appliquée L,
Perrin J.-L. (1991) Réseaux d'innovation, milieux d'innovation et développement territorial, Revue d'Économie Régionale et Urbaine 3-4, pp. 343-374.
Perrochon D. (2000) Le recentrage dans l'industrie pharmaceutique : vers une redéfinition des métiers de la santé. In: Paulré B. (Ed), Les stratégies de recentrage dans les secteurs de hautetechnologie. Chimie, Pharmacie, Informatique, Télécommunications, Ministère de l'Éco-nomie, des Finances et de l'Industrie, Direction Générale de l'Industrie, des Technologies, de NETWORKS AND COALITIONS IN THE PHARMACEUTICAL INDUSTRY l'Information et des Postes, Collection Études, February, Paris, Editions de l'Industrie,pp. 125-190. Pfeffer J., Salancik G.R. (1978) The External Control of Organizations: A Resource Depen- dence Perspective. Harper and Row, New-York.
Pharmaceutical Executive (2000) Top 50 Pharmaceutical Companies of 1999, 10 p., (http:// Pharmaprojects (2000) What are the latest industry developments?, Figure 3: Top 25 companies based on the number of active products in their R&D pipelines, (http://www.pjbpubs.co.uk/pharma/index.html). PhRMA (2000) Pharmaceutical Industry Profile 2000, (http://www.phrma.org/publications/ industry/profile00/). Planque B. (1991) Note sur la notion de réseau d'innovation, Revue d'Économie Régionale et Urbaine 3-4, pp. 295-320.
Porter M.E. (1993) L'avantage concurrentiel des Nations. InterEditions, Paris. Porter M.E., Fuller M.B. (1986) Coalitions and global strategy. In: Porter M.E. (Ed.), Competi- tion in global industries. Harvard Business School Press, Cambridge (Mass.), pp. 315-343.
Powell W.W. (1990) Neither market nor hierarchy: network forms of organization, Research in Organization Behaviour 12.
PriceWaterhouseCoopers (1999) Pharmaceutical executive briefing 1998, PWC. Quéré M., Ravix J.L. (1996) Les relations science-industrie : analyse des procédures d'accès à des activités nouvelles. In: Ravix J.L. (Ed.), Coopération entre les entreprises et organisationindustrielle. Editions du CNRS, Paris, pp. 267-284. Raugel P.J. (1993) Les sociétés spécialisées en biotechnologies : stratégies, environnement industriel et financier. In: Scriban R. (Ed.), Biotechnologie, Lavoisier, Paris, pp. 773-813.
Rhône-Poulenc-Rorer (1998) Rapport d'Activité.
Richard D., Senon J.L. (1996) Le médicament. Flammarion, Paris.
Richardson G.B. (1972) The Organization of Industry, Economic Journal 82, 327, pp. 883-896.
Rivière P. (1997) Esquisse d'une typologie des réseaux. In: INSEE, Les réseaux d'entreprises :
des collectifs singuliers, Insee-Méthodes 67-68.
Rosier F. (1999), La course aux "snips" bat son plein, Biofutur 194, pp. 84-86.
S.N.I.P. (1997) L'Industrie pharmaceutique, ses réalités. SEFEFIM, Paris.
S.N.I.P. (1999) Les chiffres clés. (www.snip.fr/industrie/presentation).
Sako M. (1991) The role of "trust" in Japanese buyer-supplier relationships, Ricerche Econo-
miche XLV, pp. 375-399.
Schumpeter J.A. (1961) The Theory of Economic Development. An Inquiry into Profit, Capital, Credit, Interest and the Business Cycle. Oxford University Press, Oxford. Sharp M. et al. (1994) Transferts de technologie et politique de l'innovation : le cas des biotech- nologies. In: Sachwald F. (Ed.), Les défis de la mondialisation innovation et concurrence.
Masson/IFRI, Paris, pp. 155-212.
Spikler B. (1993) Multinational pharmaceutical companies: principles and practices. Raven Staropoli C. (1998) Cooperation in R&D through a network, an "organizational gamble"? An empirical analysis of Rhône-Poulenc Rorer-Gencell, Technology Analysis & Strategic Mana-
gement 10, pp. 511-527.
Stengers I. (1985) Les généalogies de l'auto-organisation, Cahiers du CREA: Généalogies de l'auto-organisation 8, November.
M.-H. DEPRET AND A. HAMDOUCH Tarabusi C.C. (1993) Globalisation de l'industrie pharmaceutique : évolution technologique et concurrence dans la perspective de la triade, STI 13, pp. 139-180.
Tarabusi C.C., Vickery G. (1996) La mondialisation de l'industrie pharmaceutique. In: O.C.D.E., La mondialisation de l'industrie : vue d'ensemble et rapports sectoriels, pp. 81-125. Teece D.J. (1992) Competition, cooperation, and innovation – Organizational arrangements for regimes of rapid technological progress, Journal of Economic Behavior and Organization 18,
pp. 1-25.
Thorelli H.B. (1986) Networks: between markets and hierarchies, Strategic Management Journal 7, 1, pp. 37-51.
Von Neumann J., Morgenstern O. (1944) Theory of Games and Economic Behavior. Princeton University Press.
Waelbroeck-Rocha E. (1997) La cohérence des politiques de l'Union européenne en matière de commerce, de concurrence et d'industrie dans le secteur pharmaceutique. In: Jacobzone S.
(Ed.), Economie de la santé : trajectoires du futur. INSEE METHODES 64-65. Economica,
Paris.
Yovits M.C., Jacobi G.T., Goldstein G.D. (1962) Self-organizing systems. Spartan Books, to access this journal online:

Source: http://ejess.edpsciences.org/articles/ejess/pdf/2000/03/depret.pdf