Transnational Corporate Ties : A Synopsis of Theories and Empirical Findings

For sociologists and analysts of the world economy, it is certainly a matter of fact that transnational corporations are by no means isolated actors in an economic “war of all against all” (Th omas Hobbes). Rather, transnational corporations are frequently members of corporate networks, often of global reach. Although the literature on globalization emphasizes the increasing economic power of these networks and postulates the formation of a transnational capitalist class, there is still a lack of empirical fi ndings. Th is article starts with a review of theoretical perspectives (resource dependence, social capital, coordination of markets, fi nancial hegemony, class hegemony, inner circle, and transnational capitalist class) which focuses on the functions and structures of corporate interlocks at the national and the transnational level. Th e subsequent section off ers a synopsis of studies concerning transnational corporate networks. Th ese analyses of corporate ties (interlocking directorates, fi nancial participations and policy group affi liations) suggest the emergence of transnational economic elites whose members, however, have not lost their national identity. In the fi nal section, the theoretical perspectives will be assessed and some prospects are sketched out. Finally, it will be argued that the disintegration of the world society, which is considerably driven by rent-seeking corporate networks, can only be restrained if a potential global regulatory agency will be anchored in a post-Washington consensus.

fact that transnational corporations are by no means isolated actors in an economic "war of all against all" (Th omas Hobbes).Rather, transnational corporations are frequently members of corporate networks, often of global reach.Although the literature on globalization emphasizes the increasing economic power of these networks and postulates the formation of a transnational capitalist class, there is still a lack of empirical fi ndings.Th is article starts with a review of theoretical perspectives (resource dependence, social capital, coordination of markets, fi nancial hegemony, class hegemony, inner circle, and transnational capitalist class) which focuses on the functions and structures of corporate interlocks at the national and the transnational level.Th e subsequent section off ers a synopsis of studies concerning transnational corporate networks.Th ese analyses of corporate ties (interlocking directorates, fi nancial participations and policy group affi liations) suggest the emergence of transnational economic elites whose members, however, have not lost their national identity.In the fi nal section, the theoretical perspectives will be assessed and some prospects are sketched out.Finally, it will be argued that the disintegration of the world society, which is considerably driven by rent-seeking corporate networks, can only be restrained if a potential global regulatory agency will be anchored in a post-Washington consensus.
In general, corporations are not isolated actors in an economic "war of all against all" but members of corporate networks of global reach.Although the literature on globalization emphasizes the increasing economic power of these networks and postulates the formation of a transnational capitalist class, there is still a lack of empirical findings.The article starts with a review of theoretical perspectives (resource dependence, social capital, coordination of markets, financial hegemony, class hegemony, inner circle, and transnational capitalist class) which focuses on the functions and structures of corporate interlocks at the national and the transnational level.The subsequent section offers an outline of em-pirical studies concerning transnational corporate networks.These analyses of corporate ties (interlocking directorates, financial participations and policy group affiliations) suggest the emergence of transnational economic elites whose members, however, have not lost their national identity.In the final section, the theoretical perspectives will be assessed and some prospects are sketched out.Finally, it will be argued that the disintegration of the world society, which is considerably driven by rent-seeking corporate networks, can only be restrained if a potential global regulatory agency will be anchored in a post-Washington consensus.

Michael Nollert 290
Transnational Corporate Ties 

theoretical perspectives of national corporate ties and networks
Th e analysis of networks of corporate power can be linked to several theoretical perspectives which are well known to scholars who address corporate ties and networks (cf.Mizruchi 1996;Nollert 2005).Although most of these perspectives focus on the national level, there is no a priori argument against applying them to the transnational level.On the contrary, even if we accept the well-known presumption that globalization leads to a devaluation of the national level, none of these perspectives will be falsifi ed when transnational corporate ties are increasing.
Table 1 sorts the theoretical approaches according to their level and their units of analysis.Th e theory that postulates the rise of a transnational capitalist class is diff erent from other perspectives because it primarily addresses linkages across national borders.Th erefore, we will fi rst summarize arguments of perspectives focusing on the level of the nation state.
Corporate ties as an instrument for corporations and managers.Studies which focus on the corporation as an actor see networking as an instrument to reduce market risks (Pfeff er and Salancik 1978).Th e resource dependence perspective assumes that fi rms operate within an insecure environment comprised of other fi rms.Hence, fi rms are not seen as autonomous but rather as constrained by interdependencies with other fi rms.In order to survive economically, corporations are thus expected, on the one hand, to control other fi rms and, on the other hand, to use interlocking directorates with fi nancial institutes to safeguard the supply of capital.Moreover, corporate ties provide a stable means of communication among interdependent fi rms.Also, the theory expects that linkages between non-fi nancial and fi nancial corporations reproduce themselves if a tie is broken due to the retirement or death of a linker.
If the individual person is the unit of analysis, interlocks are regarded as the strategy of corporate directors to accumulate social capital.Th us, this perspective assumes that the social networking of managers at board meetings, in private clubs, and in policy groups provides a resource which can be transformed into economic capital (Bourdieu 1983).Hence, the social capital theory postulates that top ma nagers use their social contacts to get better paid jobs or executive com pensations which cannot be justifi ed by market rules (see Bebchuk and Fried 2004).
Corporate networks pointing out coordinated markets and fi nancial hege mony.A good starting point for the analyses of corporate network structures is still Adam Smith' s Th e Wealth of Nations (1979Nations ( [1776]]) which points out that economic actors in general are not adherents of the principle of free markets.Rather, these actors fear competition; and therefore they try to coordinate markets by fostering alliances with competitors.
Th e economic history of the core countries demonstrates that Smith' s assumption obviously received a bigger echo in the Anglo-American world than in continental Europe.Indeed, whereas corporate networking in the sphere of Rhenian capitalism (Albert 1991), and especially in the small countries, was seen as a legitimate strategy enabling corporations to strengthen their common economic power in world markets, the United States, for example, introduced rigorous legal barriers to prevent collusion and anti-competitive interlocking.
Hence, Alfred Chandler (1990) distinguished two models of capitalism: the competitive one, which is characterized by strict laws on competition and weakly coordinated markets, and a cooperative one, in which collusive behavior and networking among corporations are tolerated.According to Chandler, the United States exemplify the pattern of competitive capitalism in the 20th century because the Sherman Act in 1890 prohibited conspirational agreements between companies, e.g.price cartels.In consequence, Section 8 of the Clayton Act of 1914 even prohibited interlocks between fi rms competing in the same markets.Vice versa, at the end of the 19th century, Germany clearly represented a model of cooperative capitalism because the 1897 verdict of the Supreme Court (Reichsgericht) declared cartel agreements as legally binding and enforceable contracts.Rudolf Hilferding (1981Hilferding ( [1910]]) also emphasized that representatives of German banks on boards of competing industrial companies cemented this system of collusion.However, the question whether interlocks between competitors are eff ective in promoting collusion is still open.
In recent years Peter A. Hall and David Soskice (2001) revitalized this distinction by pointing out the diff erences between liberal and coordinated market economies.Th ey argue that in coordinated market economies, fi rms depend heavily on non-market relationships, such as networks, to coordinate their rela- ate the claim that banks and insurance companies reach very central positions within the networks (cf.Nollert 1998).
Corporate networks as structures for class formation.We should also be aware of the theoretical perspectives which do not focus on the corporate networks but on the network of individuals linking corporations.Th e so-called class hegemony model presumes that interlocking directorates are the instrument by which the class of top managers and owners can transform itself into a class-for-itself.In other words, inter locking directorates or memberships in business associations, private clubs and policy groups are seen as the basis of the elite' s cohesive pursuit of economic interests in the political fi eld (Mills 1956).Th e inner circle theory (Useem 1984) specifi es this perspective insofar as only individuals with many positions in diff erent companies are expected to lobby for the common interests of all private enterprises.Most country studies still support this view (e.g., Useem 1984;Bearden and Mintz 1987;Windolf 1997;Nollert 1998Nollert , 2005)).Th us, it is a matter of fact that the so-called big linkers, or multiple directors, share not only social characteristics and political views, but also an active commitment to furthering the common interests of big companies.However, the sociological community is still divided over the question of the impacts of these elites on the integration of society.Proponents of the class hegemony and critical elite theories argue that elites only pursue their own selfi sh interests and therefore further economic inequality and social disintegration (e.g., Mills 1956).In contrast, the functional elite theories reply that elites as well as a distinct system of social stratifi cation is necessary and in the best interest of society as a whole (e.g., Keller 1963).
Th e transnational-capitalist-class-hypothesis.Unfortunately, we have to recogni ze that the literature on transnational corpo rate networks, fi rstly, hardly ever refers to the theoretical approaches mentioned above, and, secondly, rarely deals with empirical data.Indeed, this literature is confi ned to presenting hypotheses, such as the emergence of alliance capitalism (Gerlach 1992;Dunning 1997) or network capitalism (Castells 1996) or the formation of a transnational capitalist class (Sklair 2001;Robinson and Harris 2001).
In recent years, analysts of the world society focused on the debate regarding the emergence of a transnational class.Th e starting-point of this debate was marked by the contributions of Richard Barnet and Ronald Mueller (1974) as well as Stephen Hymer.Already in 1979, the latter author argued that in the long run, the owners and managers of multi national enterprises will constitute a powerful social class with global reach.
Th ere are at least three hypotheses referring to the consequences of the transnationalization of production processes: tions with other economic actors (other fi rms, trade unions).In contrast, liberal market economies are characterized by hierarchically organized fi rms which exchange goods and services in a context of competition and formal contracting.Th erefore, corporate networking is assumed to be far more frequent in coordinated than in liberal market economies.
Th e most prominent theory addressing the structure of corporate networks is the bank or fi nancial hegemony theory (Mintz and Schwartz 1985;Glasberg 1996).In line with the resource dependence argument that production companies cannot survive without a supply of fi nancial capital, this theory postulates that banks and insurances occupy a hegemonic position in domestic economies.Yet, the theory also emphasizes that this hegemony is not based on the control of non-fi nancial companies by managers on the board, but on mutually interlocking directorates among fi nancial and non-fi nancial companies and on high legitimacy of the fi nancial sector:¹ "Interlocking directorates are not a source of hegemony but a method for managing discretion.…[B]ankdominance in this context refl ects the dominant position of fi nancial institutions in capital-fl ow decision making."(Mintz and Schwartz 1985: 250).In other words, the central position of fi nancial institutions is based on sending directors to the supervisory boards of non-fi nancial companies as well as on the cooptation of directors of non-fi nancial companies.In line with Hilferding (1981Hilferding ( [1910]]), Mintz and Schwartz also assume that banks use their interlocks to contain inter-company disputes.Th ese control and cooptation strategies are not problematic in the eyes of the national com peti tion agency as long as the related communication does not establish a basis for collusive behavior (car tels, mergers).Taking into account the above-mentioned arguments, it is no sur pri se that the competition laws of many Western countries (e.g.USA) and of the Euro pean Union focus particularly on the horizontal linkages between corpo ra tions.
Empirical analyses of corporate networks at the national level strongly support both Chandler' s typology as well as the hegemony theory.Th us, comparative analyses of data on interlocking directorates show, on the one hand, that corporate networks in the Anglo-American world are less dense and centralized than those in the Netherlands or in Switzerland (see, for example, Windolf and Nollert 2001;Nollert 2005).On the other hand, most country studies substanti-¹.From a neo-Marxist point of view, linkages result not only in the reduction of dependency but, primarily, bring about an increase in control executed by the resourceful corporation.Since fi nancial capital is a crucial resource in modern capitalism, one can assume that fi nancial institutions dominate networks (Hilferding  []).

Michael Nollert 294
Transnational Corporate Ties  • Th e importance of the nation state is declining.
• Th e signifi cance of the geographic division of the world into North and South, Core and Periphery is declining.• A transnational capitalist class is emerging which transforms into a class-for-itself.Furthermore, these three consequential processes are assumed to reinforce each other.However, since we are primarily concerned with the structure and immediate function of corporate networks, only the third hypothesis needs to be addressed here.Transnational interlocking directorates and policy group affi liations are expected to constitute a global upper class that will lose its national base and integrate the core sectors of the national economies.In short, it is argued that communication between owners and managers of transnational companies creates mutual trust, the potential to monitor or even to exercise control over another company, and a common identity that shapes the members' behavior more than their national identities.Th us, William I. Robinson and Jerry Harris (2000;Robinson 2004) claimed that the process of transnational class formation goes hand in hand with a new geographi cally undefi ned division of the world into a global bourgeoisie and a global proletariat.In addition, Leslie Sklair (2001) specifi ed that the emerging "transnational capitalist class" is composed of four main interlocking groups: corporate executives, globalizing bureaucrats and politicians, globalizing professionals and merchants, and the media.And Van Apeldoorn (2000: 157) even claimed that the European Roundtable of Industrialists must be interpreted as an elite platform for an emergent European transnational class.
However, considering those theoretical approaches which focus on corporate ties at the national level, one must concede that most proponents of the transnational-capitalist-class-hypothesis do not refer to analyses of interlocking directorates among corporations.Also, there are good reasons to criticize both Sklair' s neglect of the wealthy owners and the global rich (Beaverstock et al. 2002) as well as the lack of empirical evidence in the study of Robinson and Harris (2000).Furthermore, neither of these theoretical approaches can substantiate that corporate leaders form a community operating similar to the inner circles within nation states.In short: until we can identify a social network whose members share a transnational identity and pursue common political interests, we cannot verify the existence of a transnational class.

preliminary empirical evidence
In order to test the hypotheses derived from theoretical perspectives focusing on the network level, scholars have examined interlocking directorates by means of social network analysis (Scott 1991;Wasserman and Faust 1994).In general, these analyses are expected to answer the question of power (cf.Mills 1956) and the ways in which power is attained, reproduced or refused in interactions.However, class formation is not only based on inter locking corporate directorships but also on common memberships in business organizations, private clubs, and policy groups.Indeed, meetings at British and American clubs, for example, strengthen class cohesiveness in the same way as board meetings in Continental Europe do.Because there are only a few transnational private clubs, the only study which empirically addressed the question of transnational class formation (Carroll and Carsons 2003) had to focus on global policy groups.
According to the model of coordinated markets, the structures of corporate networks should strongly vary in levels of density and centralization.Th e density and centralization is expected to be substantially higher in coordinated economies than in competitive market economies.Th e fi nancial hegemony theory supports this hypothesis as well.However, this theory also suggests that banks and insurances are over-represented in the core of the network.Focusing on the networks of persons, the inner circle theory is supported if the analysis indicates the existence of a network of persons who pursue the interests of the business community.Th e transnational-capitalist-class-theory fi nally postulates the emergence of a network of top managers who share common interests which are no longer constrained by geographic location.
Considering the arguments against the background of the transnationalcapitalist-class-hypothesis and the premise that interlocking directorates indicate the existence of a class-for-itself, we must admit that network analyses alone cannot prove whether the members of the network share a transnational identity and engage in politics.Certainly, other methods have to be considered if one wants to answer questions of whether interlocking directorates result in intercorporate control and/or the construction of an transnational business community, whether they infl uence the political decision-making process, whether they reduce the extent of economic competition, or whether corporate networks intensify the disparities between the core and the periphery of the world society.By admitting these limits, however, intercorporate network analysis can answer the question whether interlocking directorates across national borders exist and what the structural propensities (density, centralization) of these networks are.After an outline of the intercorporate network analysis, I will therefore discuss a couple of seminal and recent analyses; furthermore I shall discuss an analysis of the interlockings represented in the European Roundtable of Industrialists.
Analyses of corporate networks.Analyses of corporate networks generally start with the defi nition of a list of major companies and their boards of directors.Th en, all interlocks in the inter-company matrix are identifi ed and counted.Two

Michael Nollert 296
Transnational Corporate Ties  companies are interlocked if a director sits on the boards of both companies.However, it should be noted that corporate law varies considerably in Europe.In some countries, we fi nd the one-board-system, which does not separate between operational and supervisory func tions (for example, in Great Britain).In contrast, the company law of Germany and of the Netherlands obliges companies to have a board of directors as well as a su pervisory board.Th erefore, one can speak of an undirected linkage if a person is mem ber of the su pervisory boards of the corpo rations A and B. If a manager of cor poration A is member of the super visory board of corporation B one may code a direc ted linkage.In order to guarantee com parability and to sim plify the coding process, the two committees are considered two parts of a single super committee in the case of the twoboard-system.
As soon as all linkages are identifi ed, it is also possible to create a second matrix which documents the ties between the individuals (Bearden and Mintz 1987).Two people are regarded as linked if they are members of the same board of directors.
After having fi nished the coding process, the structural features (such as the density, centralization or clique structures) of both the intercorporate and the interpersonal networks can be determined.Furthermore, network analysis helps to distinguish between those actors who occupy central positions and those at the periphery of the network.
Of course, network analyses can also examine fi nancial participation ties.Figure 1 presents a sample of well-known national ownership network patterns: the star, pyramid, circle, and reciprocal clique.
Th e star symbolizes a dominant fi rm that is surrounded by a number of satellite companies.German combines used to have such a structure.If the arrows are inverted, the structure becomes that of an inverted star.Th is pattern is very common in the USA and the UK where many institutional investors often own small shares of large companies.Th e pyramid symbolizes a dominant fi rm which owns shares in other fi rms, which in turn dominate further fi rms.Th is pattern is very common in Belgium and in Germany.Circle structures can be identifi ed in France.Th ey help to maintain family control of large enterprises.Finally, the Japanese "keiretsu" provides a famous example of the reciprocal clique.
Pioneer and recent studies.To the best of my knowledge, Meindert Fennema provided the fi rst analysis of transnational corporate networks in 1982.He studied the interlocking directorates of 176 of the largest companies from 12 countries in 1970 and in 1976.While in the 1970 network, the North Atlantic companies and the Japanese companies belonged to diff erent components, Fennema identifi ed a large component of Western fi rms in the 1976 network which by then included two Japanese companies.Th is study also indicated that both the number and the proportion of interlocks that cut across national borders increased in the 1970s.Th e most central compa nies in the overall network were Akzo and Shell (both from the Netherlands) and Deutsche Bank and Bayer (both from Germany).Also, the analysis showed that German, Dutch, Swiss and French fi rms at least doubled their number of trans national linkages.
Fennema and Huibert Schijf published another seminal study in 1985.In contrast to the previous study, they analyzed the transnational fi nancial participation of more than 50 of the 250 largest fi rms in nine countries.Th eir analyses showed, on the one hand, that corporations with the home base in the US participated fi nancially in 116 fi rms, whereas for companies in Italy, Belgium and Austria the fi gure was less than 10.On the other hand, fi rms in the other eight countries participated fi nancially in only two companies in the US, but in 72 in Belgium and 40 in Italy.Th e most central corporations in the network were Philips (10 participations), British Petroleum (8), Exxon (7), Chrysler, Mobil Oil, Shell (each 6).
In 2002, Fennema published another study together with William Carroll in which they tried to answer the question of whether a transnational capital- According to their fi ndings, the global corporate elite consists of 622 individuals who direct at least one corporation and one other organization (corporation or policy group).However, 64.3 of these elite members were only national linkers without policy group directorships.Six transnational linkers who participated in two or more policy boards, belonged to the inner circle of this elite: Bertrand Collomb (Lafarge-Copée, France), Minoru Murofi shi (Itucho Corporation, Japan), Percy Barnevik (ABB, Sweden/Switzer land), Peter Sutherland (British Petroleum), Paul Allaire (Xerox, USA), Etienne Davignon (Fortis, Belgium).As a result these six linkers connected eleven European, six North American fi rms and one Japanese company.It is also noteworthy that all four Europeans on this list were members of the European Roundtable of Industrialists in 1994.
Th e European bias in the global network holds if the directorships of fi ve national linkers with two or more policy board memberships, and one corporate board member with two or more policy board memberships are included in the inner circle.Regarding the 105 individuals who were transnational linkers or had at least two policy board memberships, representatives from the US (21), the UK (16.2), Germany (14.3) France (11.4), the Netherlands, Belgium, Canada and Japan predominate.
Furthermore, an analysis of the number of corporate elite in policy groups indicates that the Trilateral Commission is the most central meeting point.While 91 members were affi liated to the commission, only 15 were associated with the WEF.Finally, the Swiss-based industrial conglomerate ABB, whose directors sat on all fi ve policy boards, was identifi ed as the most central corporation in the network.
Th e network of the European Roundtable of Industrialists.Let us fi nally turn to the fi ndings of my analysis of interlocking directorates among companies which are represented in the European Roundtable of Industrialists (Nollert 2000a;Nollert 2005: chap. 14).Several interviews with representatives of the most infl uential European policy group supported the assumption that the emergence of the network of the European Roundtable of Industrialists was the result of a collabo ration between Etienne Davignon, at the time Commis sioner of the European Commu nity for Internal Market and Indu strial Aff airs, and Pehr Gyllen hammar, the chief executive offi cer (CEO) of Volvo.Various CEOs were invited to join, and in April 1983, the fi rst meeting took place in Paris.Since this initial meeting, the Roundtable has convened regularly.Regular meetings also take place bet ween the Roundtable and the Commis sion of the Euro pean Commu ni ties.Th e Round table' s secretariat is now based in Brussels and has published se veral publications on a variety of pan-European subjects.
ist class is emerging.In order to answer this question they compared the networks linking 135 of the largest industrial companies plus 41 of the largest banks in 1976 and in 1996.Surprisingly, mean centrality-as measured by the degree of the companies' involvement and the overall number of interlocks-slightly decreased.In contrast, the proportion of transnational interlocks increased from 22.8 to 24.5.
Hence, they had to concede that almost three-quarters of all interlocks still linked companies within national borders in 1996.Yet, in 1976 the key fi rms were Dutch and German (Royal Dutch-Shell, Akzo, Deutsche Bank) and in 1996 Swiss, French, British and American companies (Credit Suisse, Novartis, ABB) also belonged to the core of the network.
Regarding the sector of core companies, the study suggests that fi nancial companies do not predominate in the transnational network as is the case in the national ones.Th e most central fi nancial institutes are Credit Suisse, the Dresdner Bank, the HangSeng Bank, the Banque Nationale de Paris and the Canadian Imperial Bank of Commerce.
However, in a recent analysis Kentor and Jang (2004) challenged the fi ndings of Carroll and Fennema which suggest that the transnational business community has not grown in the last decades.Th eir comparison of the interlocking directorates of the Fortune Global 500 fi rms between 1983 and 1998 shows both a signifi cant increase in the total number of interlocking directorates and increasing trans national interlocks.While in 1983, there were 875 linkages among the 500 largest fi rms (by revenue) in the world, in 1998 there were already 1097 linkages.During this period, the number of transnational interlocks increased from 120 (about 16) to 181 (about 20)-this is more than 50.Both authors attribute this growth to the economic integration process within the European Community.It is hardly a surprise that Carroll and Fennema (2004) immediately responded to this fi nding.Th ey pointed out methodological problems and criticized the lack of adequate theorizing of the structural and historical conditions of corporate interlocks.More specifi cally, they argue that the Fortune list favors high-volume merchandisers such as Walmart.Financial institutions, such as banks and insurances, are therefore strongly underrepresented.Furthermore, they state that their fi ndings do not necessarily challenge the view of an emerging transnational capitalist class which can persuade politicians and domestic publics that the structural power of transnational corporations should not be restrained in the long run.
Moreover, we should appreciate the analysis of the 1996 networks of corporate leaders presented by Carroll and Carson (2003).Th is analysis included the network of the 300 largest companies (plus 50 companies), the boards of the International Chamber of Commerce (Paris), the World Economic Forum It should be noted that many members of the Roundtable already knew each other prior to its formation.For example, Giovanni Agnelli (Fiat), the brother of co-founder Umberto Agnelli, was friends with the owner of Volvo, Peter Wallenberg.Pehr Gyl lenhammar (Volvo) and Bernard Hanson (Renault) both belonged to the associa tion of the automobile manu fac turers.
Th e board of the Geneva International Manage ment Institute included both Gyllenhammar and Stefan Schmid heiny (former Roundtable member).Gyllenhammar also met Giovanni Agnelli and Kenneth Durham (Unilever) as advisors to the Chase Manhattan Bank.Fi nally, Gyllen hammar and Davignon were allied with the consulting company of the former US Sec retary of State, Henry Kissinger.One of the important clients of the Kissinger Asso ciates was Umberto Agnelli.
Another venue of Roundtable members is the so-called "Bilderberg Group" which retains its importance even today.Th e British businessman Joseph H. Retinger who was one of the founding members of the European Movement initiated the global policy group.In 1952 he proposed to his friend, Prince Bernhard of the Netherlands who at that time sat on the boards of Shell-Royal Dutch and the Belgian holding company So ciété Générale, that he summon a secret conference for the elites of the NATO member-states.In the USA, the Rockefeller family who controlled Chase Manhattan Bank and Standard Oil principally supported this proposal.Th e fi rst meeting took place in 1954 at the Bilderberg Hotel in Oosterbeek.Usually more than 100 per sons (about 80 are from Europe) take part in the annual discussions.In the 1990s, representatives on the Steering Committee included Roundtable mem ber Etienne Davignon, former Roundtable members Percy Barnevik and David de Pury (both ABB) and Hilmar Kopper (Deutsche Bank) who interlocked several Round table companies (see Table 2 be low).Giovanni Agnelli belonged to the advi so ry committee.At the 1996 conference in Toronto, participants included the Italian Commis sioner Mario Monti, Giovanni Agnelli, Barnevik, Bertrand Collomb (Roundtable member, La farge Coppée), and Morris Tabaksblat (Roundtable mem ber, Unilever).
Informal contacts were also made through the Trilateral Commission, in which Gio vanni Agnelli, Etienne Davignon, and Poul Svanholm (also a Roundtable member) par ticipated in the early 1980s.Th e Trilateral Commission was convened for the fi rst time in 1973 by mem bers of the Bilderberg Group.Th e Trilateral Commission was designed to provide an opportunity for a selection of promi nent politicians, business leaders, and academics from Europe, North America, and Japan to think about and discuss coordinated action on issues of global im portance.In con trast to the discussions of the Bil der berg Group, however, the results of the Trilateral Commission' s proceedings were pub lished.
A re view of the economic and fi nancial newspapers of the European Union (Wirtschaftswoche 1991) also noted that Round table businesses are over-represented in the European network of boards of directors, which at the begin ning of the 1990s was still quite weakly inter connected.Th us, among the seventeen per sons identifi ed as having at least two seats, four Roundtable members and four non-mem bers can be found who have two seats in Roundtable businesses.Included among those holding three or more seats are André Leysen (representative of Gevaert in the Round table) and Etienne Davignon (representative of the Société Générale in the Roundtable).Among those with two seats are Carlo de Benedetti (Olivetti) and Helmut Maucher (Nestlé).In addition, Roundtable businesses are connected through Hilmar Kopper (Solvay, Pil kington), Henry C. Bodmer (Fiat, Pirelli), François Laage de Meux (Olivetti, Société Gé nérale) and Niklaus Senn (VEBA, Siemens).
In addition, a comparative analysis of the 1984 and 1994 interlockings among the members of the European Roundtable indicates that the intercorporate network expanded during the 1980s and became more dense.In 1984, only 15 personal linkages have been identifi ed, whereby three of them are di rected and twelve are undirected.Th ere are eight linkages between corporations from the same country.Th us, the extent of transnationalization of the network was 46.7.Fur thermore, it is noteworthy that eight of these linkages concerned people who are not members of the Round table.However, not only the frequency of linkages was modest, but so were the relation densities.Only fi ve corporations (Uni lever, ².Umberto Agnelli, FIAT, Italy; Sir Peter Baxendell, Shell, UK; Carlo Benedetti, Olivetti, Italy; Wisse Dekker, Philips, Netherlands; Kenneth Durham, Unilever, UK; Roger Faroux, St. Gobain, France; Pehr Gyllenhammar, Volvo, Sweden; Bernard Hanon, Renault, France; John Harvey-Jones, ICI, UK; Olivier Lecerf, Lafarge Coppée, France; Helmut Maucher, Nestlé, Switzerland; Hans Merkle, Robert Bosch, German; Curt Nicolin, ASEA, Sweden; Louis von Planta, Ciba Geigy, Switzerland; Antione Riboud, BSN, Franc; Wolfgang Seelig, Siemens, Germany; Dieter Spethmann, Th yssen AG, Germany.

Michael Nollert 302
Transnational Corporate Ties  companies in the network were Bayer and VEBA from Germany and Solvay from Belgium.A familiar concept, which designates the subgroups of a network, is the 1-clique.Th is re fers to groups of actors that are all directly interconnected.In 1994, seven 1-cliques could be identifi ed with at least three companies.Four of these included fi ve compa nies: (1) Bayer, Daimler-Benz, Pilkington, Solvay, VEBA; (2) Danone, Lyonnaise des Eaux, Petrofi na, Société Générale, Saint-Gobain; (3) Danone, Petrofi na, Société Géné ra le, Saint-Gobain, Sofi na; and (4) Danone, Petro fi na, Société Générale, Sofi na, Sol vay.In addition, comparatively large 2-cliques can also be identifi ed, i.e., groups of linked actors that include at most two mutual paths.Th e biggest 2-clique includes ten com panies; along with Bosch, Daimler-Benz, Gevaert, Krupp, Philips, Pilkington, and Sie mens, this group includes the three most central enterprises: Bayer, Solvay and VEBA.
Th e stability of the cliques is substantiated by the block procedure and ncore ana lysis.Blocks designate components that do not disintegrate into two subcomponents after the rejection of one actor.Th at is to say, all block members are connected to each other by at least two paths.Th e block contains no communicatively privileged "brid ges" (Burt 1992).Th e only block with more than three Roundtable members includes the following twenty-four companies: Bayer, Bertelsmann, Bosch, BP, Daimler-Benz, Da none, Fiat, Gevaert, Krupp, Lafarge Coppée, Lyonnaise des Eaux, Petrofi na, Philips, Pilkington, Royal Dutch Shell, Société Générale, Saint-Gobain, Siemens, Sofi na, Sol vay, Th yssen, Total, Unilever and VEBA.Th e central enterprises are therefore not only directly linked but also interconnected through numerous other paths.Th us, the net work itself would not fall apart with an exit of the representatives of Sofi na and Solvay, the two companies with the outstanding betweenness scores.Nestlé, which also dis tin guishes itself through a com paratively high betweenness score, is not however con tained in the block.Indeed, the Swiss food conglomerate forms a bridge between the network' s core (Solvay) and the peripheral businesses ABB and Hoff mann-La Roche.Here, it should be noted that the link between Solvay and Nestlé is not to be attributed to Roundtable member Maucher but rather to the president of the manage ment of the Credit Suisse bank, Rainer Gut.
Another index for the cohesion of a sub-network is the n-core.Th is is a group of ac tors who are directly connected with at least n group members.In the pre sent data set we identifi ed a 4-core which includes the nine following enterprises: Bayer, Da none, Daimler-Benz, Lyonnaise des Eaux, Petrofi na, Pilkington, Société Générale, Sol vay, and VEBA.In other words, each of the listed nine businesses is directly con nec ted with at least four other businesses.
Th ese diff erent analyses of cohesion show that the central enterprises of the net work did not con stitute separate but only weakly connected sub-networks.
Sie mens, Th yssen, ABB, Volvo) or 29.4 of the 17 corporations represented in the Round table were linked.Taking into account the multiple linkages, the density of the network was 4.4 (6 linkages) of a possible 136 linkages {(17 * 16) / 2}.Finally, there was not a single fi nancial interconnection and only fi fteen managerial intercon nections.
In 1994, there are 122 linkages (see Figure 2), whereby 37 are directed and 85 are undirected.Additionally, it is noteworthy that 52 of these linkages are comprised of individuals who do not belong to the Roundtable.Similar to 1984, double mandates in managements (4) are a negligibly small category.In 1994, the density of the network is already 6.5.Responsible for the comparatively small increase is the fact that numerous multiple linkages exist anew and the denominator of the density quotient meanwhile reaches 990 {(45 * 44) / 2}.In contrast to 1984, it is possible to identify 49 linkages between corporations from diff erent countries.However, because the num ber of transnational interconnections has increased far less than the number of all interconnections, the extent of transnationalization decreases to 40.2.Accor ding to Figure 2  Transnational Corporate Ties  Which is also to say, the core of the network of Roundtable companies is only slightly fragmented and is dominated by German, Belgian and French businesses.Th e division of the presi dency of the Roundtable refl ects the two axes that had already become visible in the research of Fennema (1982) and Fennema and Schijf (1985).Th us, President Monod represents enter prises linked to the Suez Group (French and Wallonian-Belgian bu sinesses) and Vice-Pre si dent Leysen represents those enterprises linked to the Deutsche Bank (German, Dutch and Flemish-Belgian busi nes ses).As second vice-president, Mau cher plays the role of a neutral ' old boy' .
Considering the interpersonal network, we can also recognize an expansion of the network.In 1994, the linkers' network already encompassed 69 persons who were responsible for 237 direct linkages.Th at is: with 2346 possible linkages, the density of the network is an impressive 10.1.Surprisingly, the most central linker in the network was not a member of the Roundtable (cursive); it was Hilmar Kopper, the former director of the Deutsche Bank (see Table 2).³However, the comparison of the degree and betweenness ranking list implies that these cen trality indicators correlate only moderately.⁴Re garding the sum of ranks, the former EU commissioner Etienne Davignon fol lows the German bank representative.Beside Kopper and Davignon, who also join the Steering Committee of the Bilderberg Group, only André Ley sen (Ge vaert), Yves Boël, Bruno Ro ger and Gérard Worms occupy high positions in both lists.
From the list of the enterprises, we can see that the boards of Solvay, Sofi na, Petrofi na, Société Générale, VEBA and Lyon naise des Eaux are the most popular venues of Roundtable members.Com paring the list of names with the list of linkers in the study of the Wirtschaftswoche (1991), one is struck by the fact that in the circle of the 17 identifi ed persons with at least two man dates, four Roundtable members and four non-members which connect Roundtable companies are included.Inter alia, Leysen, Davignon, Carlo de Benedetti (Olivetti), Maucher as well as the non-members Kopper, Henry Bodmer (Fiat, Pirelli), François Laage de Meux (Oli vetti, Société Géné rale) and Niklaus Senn (VEBA, Sie mens) be long to this circle.
Finally, we also observe an increase of fi nancial participations among the companies represented in the Roundtable.In 1994, Bayer fi nancially participated ³.It is possible that a manager of a non-member business might sit on the board of a Roundtable business.Similarly, a Roundtable representative might sit on the board of a non-member business.In our analysis, those links between a Roundtable member and a non-Roundtable member are taken into account in addition.We do not analyze links among non-Roundtable businesses.
⁴. Th e simplest and most familiar measure to defi ne the centrality of a corporation is the degree.Th e degree corresponds to the absolute number of all direct contacts.Th ereby, it is not important whether the contacted actors are connected with each other.
Starting from the assumption that such interconnections are redundant, it is necessary to consider an indicator in addition to the degree that weights the non-redundant contacts stronger than the redundant ones.Th is claim is fulfi lled by the index of betweenness.Unlike the measurement of degree, betweenness indexes the ex tent to which an actor is located between unlinked actors.In comparison with the degree, betweenness has the advantages of indicating the centrality in the whole network and giving priority to the non-redundant contacts.Non-redundant are, for example, the contacts from British Petroleum (BP) to BAT and to British Steel (see down to the right in Figure ), because no contacts exist between the latter two.On the contrary, Krupp (see top left in the table) has contact exclusively with corporations which are interrelated with each other (for example VEBA and Bayer) and therefore, obtains a betweenness of .Actors who are linked only with actors who are connected to each other get a betweenness score of  (see Was serman and Faust : ff .).Also, the Fiat holding compa ny Ifi l had a 5.8 fi nancial involvement with Danone (formerly BSN).Considering indirect links, the greatest number of fi nancial links is found in Belgium where the French hol ding company Compagnie de Suez, as parent company of the Société Générale, has in direct inter connections to French Roundtable companies.On the one hand, Suez is linked through its 5 share in St. Gobain and its 17 share in Lyonnaise des Eaux.On the other hand, St. Gobain controls over 6.3 of the stock and 10 of the voting rights in Suez.On April 11, 1997, the boards of Suez and Lyonnaise des Eaux approved a merger, thus deepening the interconnectivity.
Th e apparently weak interconnection among German businesses may be deceptive; the three large banks Deutsche Bank, Commerzbank and Dresdner Bank, are represented in advisory bodies of Roundtable enterprises and are shareholders of Roundtable companies, thus forging indirect per sonal and fi nancial con nections between the German member companies.Not to be forgotten is Peter Wallen berg' s Investor AB which controls two Roundtable com pa nies: ABB and Ericsson.Th e new chair person of the board of Investor AB is Percy Bar nevik, the former head of the ABB conglo me rate.According to Weltwoche (No. 17, 24. April 1997), the nomi nation of the former head of Wal lenberg' s Electrolux the head of Volvo went too far-even for Wallenberg, a founding mem ber of the Round ta ble who com plained: "Swe den ap pears to occupy a rather unique position if one con siders how unilateral and mono lithic we have become, and how highly power is concen trated with us" (ibid.: 24).
Referring to the theoretical perspectives presented above, these empirical fi ndings support, at least at the level of Europe, both hypotheses.A transnational capital class is emerging and there is an inner circle of multiple directors who share an interest in an intensifi ed European integration.However, the fi nancial hegemony hypothesis is not fully confi rmed because there are no chief executives of big banks and insurances in the Roundtable.Yet, we know that many members of the Roundtable are members of supervisory boards of fi nancial corporations.Th erefore, the interests of big European banks and insurances obviously will not be neglected.Also it would be very risky to argue that all members of the Roundtable belong to the inner circles within the European countries.Hence, the transnational class hypothesis might be challenged by the fact that many members of national inner circles do not sit on boards of foreign companies.

discussion and prospects
Th e reconsideration of the theoretical perspectives in Table 1, which concentrate on the structure and function of transnational corporate networks, allows no consistent conclusion in the light of the preliminary empirical fi ndings.According to the resource dependence theory, the transnationalization of production processes implies a revaluation of transnational resource transfers.Th us, an increase of transnational linkages and a relative decline of domestic linkages are expected.Although we know that trans national corporate networks and transnational elites exist, there is no strong evidence that the formation of transnational networks goes hand in hand with the de-structuration of national networks.Yet, there is empirical evidence that at least the national networks of Switzerland and Germany disintegrated to a certain extent in the late 1990s.For instance, my analysis of the net works' develop ment of the 50 lar gest Swiss com panies between 1995 and 1997 shows a reduction rather than an in crease of relation densities.In fact, the num ber of lin kages, as well as the density and centrality, declined slightly.Moreover, the number of persons who interconnect with at least three companies has dropped from 32 to 23.At the same time, the density has remained unchanged (33) and the centrality has even decrea sed slightly (from 23 to 18); thus a signifi cant disintegration of the elite network can be diagnosed.For Germany too, a tendency towards disintegration is reported.For instance, Win dolf (2000) indicates that the number of interlocking directorates of the 15 most central corporations was nearly halved between 1993 and 1999 (see Heinze 2001).Beyer (2003) draws a similar conclusion.His comparison of the cores of German networks in 1992 and 2001 documents a reduction in density of 50 for the most central corpo rations of the year 1992, and a reduction of nearly 30 if the diff erently composed network cores are compared.
In view of the supposed erosion of the "Deutschland AG" (see Heinze 2001), a number of authors consider possible convergence processes, that is, the transformation of Rhenian capitalism to the Anglo-Ame rican system (e.g., Lütz 2000;Streeck 2001).Even if this convergence process is impeded by institutional obstacles (Windolf and Nollert 2001), and even if Kogut and Walker (2001) argue that deregulation, techno logical change, and the integration of markets do not inevitably question the stability of intercorporate ownership, it is foreseeable that the personal linkages on the national level have become less important in Rhenian capitalism over recent years.
Turning to the fi nancial hegemony perspective, we also have to admit that network analyses do not suggest a hegemonic position of banks and insurances in the global economy.In most analyses which consider fi nancial institutions, the core is dominated by industrial fi rms.In contrast to the national networks which  Nobel), Denys Henderson (Imperial Chemical Industries), and the Swiss Fritz Gerber (Hoff mann-La Roche) and Reto Domeniconi (Nestlé).However, we have to concede again that even the cosmopolitan linkers in the inner circle do not necessarily share a transnational identity.Furthermore, the rejection of the Multilateral Agreement on Investment due to strong protests of civil society organizations, for example, and other de-legitimating moves from below clearly question the omnipotence of transnational corporate elites.Finally, the question of whether transnational corporate networks off er creative responses towards the re-integration of the world society has to be answered.Reconsidering the dissent in sociological elite theory, we can imagine two scenarios.On the one hand, we can argue that elite networks will work like are still organized around big infl uential fi nancial institutions, the transnational network seems to work like a superstructure that primarily secures the communication between the national networks at the moment.
Th e remaining theoretical perspectives obviously get more empirical support.First, we cannot deny that directors of big companies currently accumulate social capital all around the globe by interlocking directorates as well as by meetings in global policy groups.
According to recent empirical fi ndings, there is also some evidence for the hypothesis that a transnational elite or class is emerging.Th ere is no doubt that a transnational corporate elite exists and forms a global communications structure by mainly linking corporations based in Europe and North America (see also Van der Pijl 1984, 1998).However, most studies deny a hegemonic role of this group and also suggest that the identity of the members of this group is still anchored in nation states (see also Hartmann 1996Hartmann , 1999)).Moreover, until there is a world state, the power of this class will continue to depend on the nation state.In other words: the transnational class-in-itself has not transformed into a hegemonic class-for-itself.Beaverstock et al. (2002) are also right when they complain about the neglect of the global rich in the network perspective and in the research on the trans national capitalist class.Finally, even Robinson and Harris (2000) concede that the transnational capitalist class is not a coherent bloc but a group of at least three interconnected fractions.Th ese include the free-market conservatives who are inspired by proponents of the Mont Pelerin Society which calls for a complete global laissez faire.Furthermore, there are the neoliberal structuralists who favor a global superstructure that could at least stabilize the global fi nancial system.Finally, we have the so-called neoliberal regulationists who demand-in line with German ordoliberalism-a global regulatory apparatus that restrains the power of transnational corporations and prevents social disintegration (see also Gill 2003).
Th ere is much more empirical evidence for the inner circle theory, which assumes that directors who hold positions in many corporations (big linkers) use them to engage in national or global policy groups.Th us, on the one hand, the analyses of Carroll and Carsons (2003) clearly identify an inner circle of cosmopolitan linkers who represent the political interests of transnational corporations.On the other hand, we also know that in the 1980s the European Roundtable of Industrialists succesfully lobbied for more intensive integration of the European Community in Brussels (Bornschier 2000).Moreover, the International Chamber of Commerce closely cooperated with the United Nations in the late 1990s (see Balanya et al. 2000).Finally, we cannot deny the infl uence of the global policy groups listed in Table 3 as well as the infl uence of less important transnational organizations such as the Mont Pelerin Society, the European Roundtable of Industrialists and therefore promote the integration of the global economy.On the other hand, the rise of transnational elite networks can be seen as the source of concentration of economic and political power which threatens democracy, social welfare and cohesion.
According to the coordination of markets-perspective, the impact of corporate networks primarily depends on the regulatory framework which will prevail in the global economy.Th is framework encompasses the property rights, the governance structures that defi ne competition rules (e.g., anti-cartel and antitrust laws) and the company law, and rules of exchange that defi ne who can transact with whom, the transaction medium (e.g., bills, money) and the enforcement of contracts.In line with the simple distinction between coordinated and liberal market systems, we can ideal-typically distinguish between a coordinated and a liberal world economy.Whereas the coordinated world economy would be characterized by an intense networking of corporations, interest groups, and a state, the model of a liberal world economy presumes strong competition between corporations and a weak state, that is, a minimal regulatory frame work.
Trying to forecast the eff ects of these types of economy on the level of societal integration, one can expect that a coordinated world economy, where corporate networks and a redistributive state cooperate, could enhance social integration while in a liberal world economy transnational corporate networks could primarily integrate production processes and national elites.Th ere is no doubt that most transnational corporations prefer a liberal economic order to an order in which a regula tory apparatus interferes in market mechanisms.Many representatives of trans national corporations even reject Adam Smith' s assumption that markets always need pro tection against networks of corporate power.In fact, they prefer a neoliberal order which rather corresponds to Friedrich von Hayek' s assumption that the "spontaneous order" of free markets primarily should be protected against the strong state.Because neoliberal competition theory, inspired by the Chicago school, refuses a state agency which prevents oligopolistic and monopolistic market structures, this economic order certainly implies a strengthening of networks of corporate power.Moreover, this order which was consolidated in the early 1980s under the program of the "Washington Consensus" allows rentseeking of well-organized and well-equipped interest groups which leads to a disintegrating redistribution of economic resources from the poor to the rich (see Nollert 2003;Shugart, Tollison and Yan 2004).In other words: "Set in a world of tradition, class, privilege, power, and diff erential organization costs, rent seeking most likely promotes more signifi cant inequalities in the distribution of income" (Tollison 1997: 518).Also, this shift of power from the nation state to transnational corporations might undermine the idea of democracy insofar as citizens will not be able to impede corporate power or initiate redistributive policies anymore.In other words: transnational corporations are expected to move their activities to a new host country if they feel challenged by democratic decisions or political protest at the local level.However, in face of the rise of transnational social movements such as Attac or the Social Forums and the new liberal regulationism (see Stiglitz 2002) which calls for a global regulatory agency anchored in a post-Washington consensus that limits collusive and rent-seeking behavior of corporate networks and considering the destruction of the global lysine, vitamin, sorbate and graphite electrodes cartels (see Connor 2003), there is a realistic hope that further global disintegration might be restrained.
Figure 1 -Patterns of Network Structures ABB members are in italics.

Table 1 -Theoretical Perspectives on Corporate Ties Level of Analysis Network Units of Analysis Actor
According to GreenCowles (1995), the fi rst list of potential members of such a group was drawn up in 1982 by the Commission and Volvo staff .Th e planning group in cluded Fer nand Braun, Director-General for Internal Market and Industrial Aff airs, Bo Ekman and Michael Hinks-Edwards, cor porate plan ners with Volvo, and Pierre De fraigne, Chef de Cabinet to Commis sioner Davignon.Seventeen of Europe' s top busi ness men² were recruited to the group and a prelimi nary meeting of top associates came up with six key areas for the agenda of the European Round table: the internal market, infrastructure, technology, jobs, environment and fi nance.
clubs, alumni clubs, and of a wide variety of other transnational networks such as the advisory boards of the global investment company and US defense contractor Carlyle (seeBriody 2003).In the latter, we fi nd former US Secretaries James Baker and Frank Carlucci, former British Prime Minister John Major, former Secretary of the United Nations Perez de Cuellar, former president of the Deutsche Bundesbank Karl Otto Pöhl, and, until October 2003, former US President George Bush Sr., as well as prominent represen tatives of the European corporate network (European Roundtable of Industrialists) such as the former Commis sioner of the European Union and Head of Société Générale de Belgique, Etienne Davignon, Oscar Fanjul (Unilever), Cees van Lede (Akzo- service