Thomas Risse. Handbook of International Relations. Editor: Walter Carlsnaes, Thomas Risse, Beth A Simmons. Sage Publication. 2002.

The end of the Cold War and globalization processes have led to renewed interest in the study of transnational relations and the impact of non-state actors on world politics. Some authors praise the emergence of a global transnational civil society (Boli and Thomas, 1999; Florini, 2000; Held et al., 1999), while others denounce an increasing transnational capitalist hegemony (Gill, 1995). Both positions ascribe to non-state actors quite an extraordinary influence on outcomes in international politics. It is certainly true that transnational actors–from multinational corporations (MNCs) to International Non-Governmental Organizations (INGOs)–have left their mark on the international system and that we cannot even start theorizing about the contemporary world system without taking their influence into account. But there is little systematic evidence to sustain claims that the transnational ‘society world’ has somehow overtaken the ‘state world’ (see Czempiel, 1991, on these notions). Rather than analyzing transnational and inter-state relations in zero-sum terms, it is more useful to study their interactions and interpenetration. As Reinicke put it, ‘governing the global economy without governments is not an option. Yet for global governance to succeed, governments will also have to enlist the active cooperation of nonstate actors’ (Reinicke, 1998: 219). The following review of the literature tries to substantiate this point.

‘Transnational relations’ is a rather elusive concept. If we take the 1971 definition by Keohane and Nye referring to ‘regular interactions across national boundaries when at least one actor is a non-state agent …’ (Keohane and Nye, 1971b: xii–xvi), the concept encompasses anything as long as human agency is involved. Yet, cross-border capital flows, international trade, CNN media broadcasts, international migration, cross-border tourism, the diffusion of values and norms, transnational social movements, INGOs and MNCs are quite different phenomena. It is impossible to theorize about them in any systematic sense. This chapter does not deal with transnational relations in general, but more specifically with transnational organizations and actors with a particular purpose. This refinement still comprises a wide range of regularized transnational relationships, from informal networks exchanging material and/or ideational resources to INGOs and large organizations such as MNCs. Some transnational actors operate globally, for example, the Catholic Church, the International Committee of the Red Cross (ICRC), Amnesty International, Daimler–Chrysler, while others are confined to specific regions of the world (such as the European Environmental Bureau, Asia Watch, or the European Trade Union Confederation). Some transnational actors concentrate on a single issue (such as the transnational campaign to ban landmines), while others follow a multipurpose mission, such as religious organizations.

This chapter distinguishes among transnational actors along two dimensions. The first dimension concerns their internal structure. Some transnational actors are formal organizations (from multinational corporations to INGOs). Others are connected in a far more loose fashion for which I use the term ‘network,’ defined as ‘forms of organization characterized by voluntary, reciprocal, and horizontal patterns of communication’ (Keck and Sikkink, 1998: 8). Some networks simply consist of groups of individuals, others comprise formal organizations. ‘Epistemic communities,’ for example, are networks of individuals and/or organizations based on authoritative claims to consensual knowledge (Haas, 1992b). Advocacy networks comprise actors who share specific values, principled beliefs and a common discourse (Keck and Sikkink, 1998: 2).

The second dimension which is relevant for this chapter differentiates between the motivations of various types of transnational actors. Some, such as MNCs or transnational special interest groups, are primarily motivated by instrumental goals and try to promote the well-being of the organization itself or the members of the group. Others, such as INGOs, epistemic communities or advocacy networks are primarily motivated by promoting a perceived ‘common good.’ This differentiation roughly coincides with the distinction between the ‘for profit’ and the ‘not for profit’ sector as frequently found in the literature. However, it is useful to think of this distinction as a continuum rather than sharply divided classes of actors. The business-sponsored Global Climate Coalition certainly proclaims that it promotes the international public good, while some (I)NGOs seek to make a profit in the humanitarian action sector.

This chapter proceeds in the following steps. I begin with a brief intellectual history of theorizing about transnational relations in world politics. I conclude from this survey that constructing dichotomies between a society-centered and a state-dominated view of international relations is misleading and distracts from interesting research questions. The main parts of the chapter examine the mutual relationship and interaction between the inter-state world, on the one hand, and the transnational world, on the other. I next deal with transnational actors and networks as ‘dependent variables.’ How do states, their institutional structures, as well as their international relations, affect transnational actors, their characteristics and their strategies? The following section changes perspective and looks at the impact of transnational actors and networks on world politics. This is the realm where most empirical research has been carried out in recent years and where we can make some empirically informed theoretical statements. The chapter concludes with some remarks on emerging public–private governance structures comprising states, international organizations and transnational actors.

International Relations and Transnational Actors: An Intellectual History

Neither transnational relations nor theorizing about them started in the post-Second World War era. Multinational corporations with dispersed investments and productions across several political jurisdictions date back at least to the medieval era. During the Renaissance era, ‘family businesses’ such as the Medicis in Florence or the Fuggers in Augsburg held huge investments and production facilities across Europe and had agents in India and China by the end of the sixteenth century (Krasner, 1999: 221). From the sixteenth century onwards, the trading companies of the imperial powers such as the British East India Company and the Hudson Bay Company operated across continents (Held et al., 1999: 238–9). Similar observations hold true for advocacy groups bound together by principled ideas and values. Precursors to modern transnational networks in the human rights and women’s rights areas include the campaign to end slavery in the United States during the early to mid-1900s (Kaufmann and Pape, 1999), the international suffrage movement to secure the vote for women in the late nineteenth century, as well as the campaigns by Western missionaries and Chinese reformers to end the practice of footbinding in China during the same period (Keck and Sikkink, 1998: ch. 2). While these early transnational movements did not enjoy modern communications technologies such as the Internet, their strategies were remarkably similar and sometimes no less effective than those of their modern successors. Krasner concludes, therefore, that rulers ‘have always operated in a transnational environment; autarky has rarely been an option; regulation and monitoring of transborder flows have always been problematic’ (Krasner, 1999: 223).

If the phenomenon of transnational actors is not particularly new, theorizing about them also has its precursors. Yet, scholarship on transnational relations during the eighteenth and nineteenth centuries was much more normative and prescriptive than analytical and descriptive. Take Immanuel Kant, for example. His 1795 ‘Perpetual Peace,’ which has become the mantra of today’s literature on the democratic peace, contains ideas on transnational relations (Kant, [1795] 1983). His statement that the ‘spirit of trade cannot coexist with war, and sooner or later this spirit dominates every people’ (Kant, [1795] 1983: 125 [368]) has been among the first claims about the causal relationship between economic interdependence and world peace. Long before modern human rights treaties proclaimed individuals as subjects of international law, Kant postulated a right of foreigners to hospitality against the government of their host state. Kant’s cosmopolitanism was rather common among liberal intellectuals during the late eighteenth and nineteenth centuries. The modern literature on the democratic peace has largely lost this connection between a democratic society, transnationalism and peace. Yet, liberals such as Tocqueville argued that transnational relations, that is, links among democratic societies of different countries and its citizens, constituted a primary tool to prevent wars: ‘As the spread of equality, taking place in several countries at once, simultaneously draws the inhabitants into trade and industry, not only do their tastes become to be alike, but their interests become so mixed and entangled that no nation can inflict on others ills which will not fall back on its own head’ (Tocqueville, 1994: 660).

While these scholars related the democratic organizations of polities, transnationalism and peace, liberal writers of the nineteenth century such as Adam Smith or John Stuart Mill took up Kant’s ideas about free trade and peace. Yet, the First World War, which was fought among highly interdependent nations, discredited the idea that economic interdependence alone is a sufficient condition for peace in the absence of democracy. Schumpeter’s Sociology of Imperialism constitutes perhaps the most elaborate statement of the inter-war period on the causal relationship between liberal capitalism, economic interdependence and peace. He argued that the essence of capitalism is anti-imperialist, but recognized that capitalist states might pursue aggressive foreign policies if they are usurped by particular economic interests (Schumpeter, [1919] 1953). Schumpeter reacted primarily to Marxist theories of imperialism, particularly Lenin’s writings, who claimed exactly the opposite, namely that imperialist wars resulted from the externalization of the internal class struggles toward the outside world and the eternal capitalist striving for new markets and profit-making. Lenin argued that wars among capitalist states were inevitable in a stage of development ‘in which the dominance of monopoly and finance capital has established himself ’ (Lenin, [1917] 1939: 89). The controversy about the precise relationship between economic interests, capitalism and economic interdependence, on the one hand, and aggressive/ imperialist foreign policies as well as peace and war, on the other, continues until today.

With the emergence of international relations as a social science discipline, scholars increasingly employed analytical rather than purely normative arguments. Mitrany, the founder of modern functionalism and integration theories, argued, in 1943 that technology and technical issues confronting the industrialized democracies in the twentieth century necessitated international cooperation along functional lines. Organizations for functional collaboration would eventually overcome the political institutions of the past including the nation-state (Mitrany, [1943] 1966). After the Second World War, regional integration theory and, particularly, neofunctionalism (Haas, 1958) reformulated the argument claiming that rational economic behavior not only leads to transnational interdependence, but also to the creation of supranational institutions as stable peace orders such as the European Community. It is important to note here that (neo)functionalism never fell into the trap of later theorizing about transnational relations which created a dichotomy and adverse relationship between a ‘society-centered’ and ‘state-dominated’ perspective on world politics. Rather, the argument was about–in today’s terms–the emergence of international institutions and supranational governance structures resulting from, responding to, and facilitating transnational interactions of private actors.

This also holds true for another version of integration theory, one of the most important predecessors of today’s constructivism, the analysis of transnational and supranational community-building by Karl W. Deutsch and his colleagues (Deutsch et al., 1957). Deutsch argued that increasing transaction flows and crossborder communication as facilitated by trade, migration, tourism, educational exchanges and the like, lead to a sense of community among people and to collective identification processes. ‘Pluralistic security communities,’ while retaining the legal sovereignty of their member states, possess a ‘compatibility of core values derived from common institutions, and mutual responsiveness–a matter of mutual identity and loyalty, a sense of “we-ness”, and are integrated to the point that they entertain “dependable expectations of peaceful change”’ (Adler and Barnett, 1998b: 7, quoting Deutsch et al., 1957: 5). In line with the behavioralist orientation of the time, Deutsch and his colleagues measured transnational transactions quantitatively and compared them to the transaction flows inside the countries in order to determine the degree of international community-building. Deutsch’s work in this area was largely ignored until recently when constructivist scholars picked up and reformulated his insights (see, e.g., Adler and Barnett, 1998a).

In the meantime, the question of transnational relations was relegated to the sidelines of theorizing on either side of the Atlantic. Explicit analytical work on transnational actors and relations started during the late 1960s and early 1970s, both in the United States and in Europe (Cooper, 1968; Vernon, 1971). In 1969, the flagship journal of the German Political Science Association Politische Vierteljahresschrift published a special issue entitled ‘Die anachronistische Souveränität’ (anachronistic sovereignty) which contained an essay on ‘transnational politics’ (Kaiser, 1969). Two years later, the journal International Organization followed suit with a special issue edited by Keohane and Nye on ‘Transnational Relations and World Politics’ (Keohane and Nye, 1971a; see also Huntington, 1973; Keohane and Nye, 1977; Rosenau, 1980). These and other works challenged the state-dominated view of world politics. Rosenau in particular attacked the state-centered paradigm of international relations theory promoting the ‘transnationalization of world politics,’ a subject to which he came back ten years later (Rosenau, 1990).

This early literature was theoretically inspired by a broader critique of the concept of the state in political theory and in comparative politics. Liberal pluralist theories defined political systems functionally in terms of the authoritative allocation of values in a given society. Societal interest groups and organizations substantially constrained political actors and the political process was largely conceptualized by conflict and bargaining among these societal groups. The work on transnational relations of the late 1960s and early 1970s transposed pluralist theory to the level of international affairs. But this work did not produce a theory of transnational politics in the sense of testable propositions. It focused on the international political economy, in particular the rise of multinational corporations in the post-Second World War era. INGOs and other transnational actors were not yet the subject of systematic inquiry (see, however, Huntington, 1973; Vallier, 1971). One of the first volumes explicitly dealing with INGOs used the term ‘pressure groups’ suggesting an analysis commensurate with the study of interest groups (Willetts, 1982).

The 1970s also saw a revival of critical political economy attacking transnational economic relations in general and the role of multinational corporations in particular with regard to the North–South relationship. Dependency theory argued against liberal free trade economists that underdevelopment results from the structural dependency and the integration of the developing world in the world economy. MNCs in particular were seen as the main agents preventing the development of an endogenous industry in Southern countries, transferring their profits from the South to the North and exploiting cheap labor in the developing world (see, for example, Amin, 1977; Emmanuel, 1972; Frank, 1967). Dependency theory constituted the first major contribution to the subject of transnational relations by Latin American, African and Asian scholars, even though most of its propositions could not be confirmed empirically (see, for example, Caporaso, 1978; Menzel, 1992; see also Maxfield, Chapter 24 in this volume).

But liberal arguments about transnational relations of the 1960s and 1970s claiming an end of the state-centered view of world politics, did not survive the counter-attack of realism, either. In the 1971 ‘Transnational Relations and World Politics’ volume, Gilpin had already argued against the liberal grain that MNCs were primarily an instrument of American foreign policy and power, not the other way round (Gilpin, 1971, 1975). The late 1970s and early 1980s then saw a revival of (neo)realist theory (Waltz, 1979). Hegemonic stability theory was the realist response to the liberal interdependence arguments.

The result was rather profound, particularly in the United States. Ruggie, Keohane and Nye had originally theorized about international regimes by arguing that economic interdependence led to increased inter-state conflicts to be regulated by international institutions (Keohane and Nye, 1977; Ruggie, 1975). This connection between transnationalism and international institution-building was mostly lost during the early 1980s when regime analysis and neoliberal institutionalism took off. The main controversy between neorealism and neoliberal institutionalism concerned the prospects of ‘cooperation under anarchy,’ that is, of cooperation among states (cf. Baldwin, 1993). In Europe, a state-of-the-art volume of the German-speaking international relations community did not bother to deal with transnational relations, except for a sharp critique of the disappointing accomplishments of the interdependence literature (Kohler-Koch, 1990; see also Zürn, Chapter 12 in this volume).

Two developments of the late 1980s re-opened intellectual space for theorizing about the cross-border activities of non-state actors in the United States and Europe. First, the late 1980s saw the beginning of what would later be called constructivism or sociological institutionalism in international relations (Kratochwil, 1989; Kratochwil and Ruggie, 1986; Wendt, 1987). Kratochwil, Ruggie and Wendt drew attention to the social and ideational rather than simply material structure of international relations (for the most comprehensive statement see Wendt, 1999). Second, the end of the Cold War should not be underestimated in its impact on international relations theorizing. The failure of traditional international relations theory to at least recognize some underlying trends, pushed many scholars away from structuralist theories such as realism and state-centered institutionalism to a renewed appreciation of domestic politics, on the one hand, and of transnational relations, on the other.

As a result of these two developments, the 1990s saw a revival of theorizing about transnational actors, a trend that was further enhanced by the debate on ‘globalization.’ First, Rosenau’s book on ‘Turbulence in World Politics’ constituted a sweeping statement on post-international politics marked by a ‘bifurcation in which the state-centric system now coexists with an equally powerful, though more decentralized, multi-centric system’ characterized by transnational ‘sovereignty-free actors’ (Rosenau, 1990: 11; also Rosenau, 1997). Second, a 1992 special issue of International Organization elaborated the notion of transnational ‘epistemic communities,’ defined as networks among professionals with an authoritative claim to policy-relevant knowledge (Haas, 1992c). The authors used constructivist work on socialization, cognitive evolution and learning in order to theorize about the relationship between consensual knowledge and power (Adler and Haas, 1992). Third, a 1995 volume (Risse-Kappen, 1995b) argued that the impact of transnational actors on outcomes depends on the domestic structures of the polity to be affected and the extent to which transnational actors operate in an environment regulated by international institutions. Fourth, Keck and Sikkink elaborated the concept of transnational advocacy networks and explored their impact in the human rights and environmental spheres (Keck and Sikkink, 1998). Finally, Reinicke’s book on ‘global public policy’ represents one of the first attempts to systematically analyze governance networks involving public and private actors on the international level (Reinicke, 1998).

Compared to the attempts of the 1970s, these latest moves at thinking about non-state actors in world politics share three characteristics:

1. While the empirical literature on transnational relations of the 1970s largely concentrated on MNCs, this focus on the international political economy is now taken over by the literature on globalization (see Chapter 12 by Zürn in this volume). The new transnationalism of the 1990s concentrates more thoroughly on the transnational non-profit sector, such as ‘epistemic communities,’ value-based advocacy networks, INGOs and cross-border social movements.
2. The recent literature is much more about the interaction between states and transnational society than about replacing a state-centered view with a society-dominated perspective. One indicator of this trend is the increasing replacement of traditional regime analysis with its focus on inter-state institutions by a ‘governance without government’ perspective emphasizing non-hierarchical networks among public and transnational actors (see, for example, Cutler et al., 1999; Czempiel and Rosenau, 1992; Kaul et al., 1999; Kohler-Koch, 1998b; O’Brien et al., 2000).
3. As mentioned above, constructivism and sociological institutionalism have influenced recent work on transnational relations. This has resulted in work focusing on transnational actors promoting and diffusing causal knowledge (epistemic communities) and norms (advocacy networks). As to critical theory, neo-Gramscianism and its contribution to the literature on the international political economy has to be mentioned (e.g. Cox and Sinclair, 1996; Gill, 1993).

In the following, I discuss the recent work on transnational actors (TNAs) in more detail. I begin with the impact of the ‘inter-state world’ on the ‘transnational society world.’

The Impact of the Inter-State World on TNAs

The nation-state system and its structuration of the world along territorially defined boundaries has a profound impact on both the nature and the activities of TNAs. The very concept of transnational relations implies an international system composed of nation-states as well as the distinction between state and societal actors within a given nation-state. It makes little sense to talk about transnational actors in a world of empires or in a medieval world of cross-cutting authority structures. However, most empirical work on TNAs remains rather unidirectional by looking at the impact of TNAs on interstate relations, international organizations and international institutions in general. We know rather little about states and international organizations enabling and/or constraining TNA activities.

TNAs as Instruments of National Governments and International Organizations

On the one end of the theoretical spectrum are those (realists) for whom the growth of transnational relations in the contemporary international system essentially reflects the interests of the most powerful states. Gilpin developed this argument most eloquently, namely that it was US post-war foreign policy and the US hegemony in the international political economy that enabled the rise of MNCs and economic interdependence in the first place (Gilpin, 1971, 1975). To ask the counterfactual, would we still experience economic globalization if Adolf Hitler had won the Second World War?

Gilpin’s argument shows some similarities with the claims by critical theorists such as Cox, even though the causal mechanism is different. Cox argued that post-Second World War American hegemony arose from a confluence of three factors, the hegemony of US capitalism and its particular mode of production, the power of the US state, and the consensual nature of Western liberal ideas (Cox, 1987; Cox and Sinclair, 1996). Cox concluded that US hegemony enabled the rise of transnational (economic) relations in the post-war era in the first place. From a liberal perspective, Nye argued in a similar fashion that the days of American hegemony in the world system are far from over, even though its economic preponderance has gone. US ‘soft’ power and the hegemony of US economic, political and cultural ideas continue to secure the current world order (Nye, 1990).

One could develop this argument further and point out that the international free trade order–from the GATT to the World Trade Organization (WTO)–and other international institutions which are ultimately based on inter-state agreements both enable and constrain transnational interactions by regulating them. The legal framework provided by states and international institutions has, thus, constitutive effects on transnational actors and relations. The international refugee regime, for example, defines refugees and their rights in the first place.

While Gilpin’s work focuses on the international political economy, a similar argument has emerged in security studies. Thomson argued in a historical study that state rulers in thirteenth century Europe began authorizing the international use of force by private armies (privateers, mercenaries etc.) in order to accumulate power and wealth (Thomson, 1994). When the unintended consequences of this privatization of international violence became obvious, rulers of the eighteenth and nineteenth centuries struggled against pirates and mercenaries in order to (re)gain the monopoly of the international use of force. The modern state monopoly over the means of internal and external violence did not fall from heaven, but resulted from a sustained conflict between states and TNAs which the states ultimately won (see, however, the rise of global terrorism, of piracy in East Asia and of private armies in ‘failed states’).

In sum, these authors share the view that state power and state foreign policies gave rise to contemporary transnational relations (and globalization, one might add) where it suited their interests, but viciously fought these transnational forces when it did not. Realists would conclude that, when powerful states are pitched against transnational actors, even mighty ones, the former usually win over the latter (overview in Krasner, 1995b: 267–76). This latter assertion, however, does not follow from recognizing state power and international institutions as enabling transnational relations and has been challenged by the recent literature on TNAs.

The question concerning TNAs as instruments of state power must also be asked with regard to the non-profit sector of INGOs and transnational advocacy networks. Unfortunately, there is little empirical work available in this area. Some preliminary observations can be made, though. First, INGOs by and large originated in the Western industrialized world and they are extremely unevenly spread across the world regions. ‘The global stratification structure is clearly reproduced in INGO participation. Residents of resource-rich, technically developed, older, formally democratic Anglo-European countries participate the most; residents of poor, less developed, newer, less democratic countries participate the least’ (Boli et al., 1999: 69). Does the INGO world then represent a ‘global civil society’ (Wapner, 1996) or does it merely reproduce Western enlightenment values such as universalism, individualism, progress and cosmopolitanism? INGOs as part and parcel of a ‘world culture’ dominated by Western liberal hegemony?

Second, many INGOs are more directly dependent on the ‘state world’ than many of them would admit. Particularly in the issue-areas of international development and humanitarian aid, funding for the grassroot activities of INGOs originates to a large extent from public sources. According to the World Bank, public funding for development (I)NGOs increased from 1.5 per cent of their total income in the early 1970s to 30 per cent in the mid-1990s. Some scholars estimate that the dependency of Southern NGOs on public funding by states or International Organizations reaches 80–90 per cent (according to Hulme and Edwards, 1997a: 6–7). The percentage of EU relief aid channeled through INGOs rose from 47 to 67 per cent from 1990 to 1994.

Even in the human rights area where one would expect most INGOs to be heavily critical of state policies, more than half of the organizations claimed to have received public funding (Smith et al., 1998). In the context of the European Union (EU), it has been frequently pointed out that the European Commission both created and funded many transnational organizations in order to be able to deal with societal interests on a European rather than on the various national levels (Greenwood and Aspinwall, 1998; Kohler-Koch, 1994).

TNA dependence on the resources of states and international organizations suggests that it would be preposterous to claim that the INGO world simply represents global civil society against the interstate system. Transnational advocacy groups and epistemic communities often perform tasks that states and international organizations either cannot or do not want to carry out. In the issue-areas of foreign and humanitarian aid, states and International Organizations often subcontract (I)NGOs, because these groups are less bureaucratic, more flexible and can reach those in need of assistance more easily. The World Bank and other organizations have long recognized that strengthening civil society in the developing world through INGO networks contributes to political, economic and social development. In the human rights and environmental areas, transnational actors provide monitoring capacities and supply information to states and international organizations which would otherwise not be available because of concerns about sovereignty rights. In the international economy, states have delegated regulatory authority to transnational private actors, for example, in international standard setting, because they can carry out these tasks more efficiently (Cutler et al., 1999). Last but not least, taking the concerns of INGOs on board also increases the legitimacy of international institutions such as the World Bank or the International Monetary Fund (O’Brien et al., 2000). As Forsythe argued in the mid-1970s with regard to the International Committee of the Red Cross (ICRC), it cooperates with and, therefore, stabilizes the state system in war-making (see Forsythe, 1976).

In sum, transnational advocacy groups and INGOs should not be seen as necessarily in opposition to the inter-state system. Rather, their work often conforms to the interests of states and international organizations. But little is known as to how the increasing role of the INGO world in global governance affects these groups themselves in terms of their institutional structures, and strategies (see the contributions in Edwards and Hulme, 1996; Hulme and Edwards, 1997b). For example, the growing involvement of INGOs in partnerships with IOs might alienate them from their own social base in civil society (Finger, 1994). INGOs working with the World Bank, the IMF, or the WTO, for example, need to moderate their goals considerably, since they have to accept the principal goal of liberalization in order to promote human rights and environmental concerns effectively (O’Brien et al., 2000: 224). This might then lead to increasing tension between more radical transnational social movements and more ‘professional’ and moderate INGOs.

Institutional Similarities between TNAs and Structures of Governance

The works discussed so far adopt an actor-centered perspective to discuss the influence of state governments and international institutions on TNAs. The question was how state actors impact upon transnational actors and how the latter actually perform functions which states or international organizations cannot or are unwilling to carry out. A second, though even less-developed argument takes a more structural perspective and asks how institutional features of states or international regimes and organizations–that is, domestic and international ‘structures of governance’–impact upon institutional characteristics of transnational actors. Krasner, for example, took a sociological institutionalist perspective and argued that the ‘institutional structures of transnational actors must reflect the institutional environment within which they function’ (Krasner, 1995b: 260). Domestic laws, for example, constitute a strong tool forcing transnational actors to adjust their institutional structures to the country in which they operate. As a result, US Honda looks different from its Japanese mother company, even though it is still institutionally different from General Motors or Ford. While the Catholic Church constitutes a quintessential transnational organization which preceded the modern state system, it still had to adjust to the domestic structure of the state in which it operates (see Vallier, 1971). Church–state relations in, say, Germany where the churches enjoy quasi-public status, differ profoundly from the United States where the Catholic Church is treated like any private organization.

Doremus and others have carried this line of argument further (Doremus et al., 1998) by investigating the internal structure and activities of MNCs operating out of Germany, Japan and the United States (see also Pauly and Reich, 1997). They claim that the current talk of ‘global corporations’ and ‘global players’ constitutes an, albeit powerful, myth: ‘Despite intensifying international competition, MNCs are not promoting the ineluctable convergence and integration of national systems of innovation, trade, and investment, nor are they forcing deep convergence in the national economies in which they are embedded. They cannot do so because they themselves are not converging toward global behavioral norms’ (Doremus et al., 1998: 3). Globally operating MNCs do not at all look alike, but maintain distinct institutional features pertaining to their organizational structure and culture which originate from the national institutional environment in which the mother company operates.

If these claims hold true for private transnational actors such as MNCs that command powerful economic resources, similar arguments should be relevant for INGOs and the transnational non-profit sector. To begin with, states and national governments control access to a territory. Moreover, transnational actors operate in institutional environments which are largely determined by the domestic structures of nation-states. Amnesty International in the United States has to abide by different laws than, say, Amnesty International in Germany. As a result, their internal organizational structures are likely to diverge, at least to some extent.

International institutions are also likely to shape organizational features of transnational actors. The EU, for example, represents an international governance structure which has given rise to particular forms of transnational interest organization. The European Commission has actively encouraged the formation of transnational organizations in Brussels, both traditional interest groups and not-for-profit INGOs. The result is a rather pluralist structure of interest organization at the EU level, in contrast to more corporatist structures with strong peak organizations in some member states (see, for example, Kohler-Koch, 1994; Greenwood and Aspinwall, 1998).

Finally, international rule structures such as norms embedded in treaties and international regimes provide an enabling environment for transnational network activities. Take the human rights area, for example. The emerging legalization of the international human rights regime went hand in hand with an increasing professionalization and even bureaucratization of INGOs such as Amnesty International or Human Rights Watch. They changed their character from transnational social movements to professional organizations employing a large staff of lawyers, media experts and country specialists. Yet, the empirical evidence on how institutional structures of the polities–both national and international–in which TNAs operate shapes the latter’s organizational structures and cultures remains sketchy at best. Most of the empirical work so far has concentrated on the question of how and under what conditions TNAs of various sorts have managed to affect the nation-state, international institutions and international organizations. The realist and state-centered legacy in international relations theory required that scholars first established that TNAs mattered before they could study how domestic and international institutions shaped the organizational structures of INGOs and the like. Such work could, for example, draw on the literature on ‘political opportunity structures’ which originated from scholarship on social movements, but has reached work on transnational actors rather recently (e.g. Thomas, 2002; on ‘political opportunity structures’ see Kitschelt, 1986; Tarrow, 1996).

The Impact of TNAs on World Politics

So far, I have discussed the literature with regard to how the ‘state world’ impacts upon the transnational ‘society world.’ As mentioned above, however, most of the empirical work poses the question the other way around and asks what effects, if any, TNAs have on structures and processes of world politics. There is one important difference between scholarly controversies of the 1990s as opposed to the 1970s and 1980s. Most of the contemporary work in international affairs no longer disputes that TNAs influence decisions and outcomes (compare, for example, Waltz, 1979 with Krasner, 1995b, 1999). Rather, current scholarship focuses on the conditions under which these effects are achieved and most of the controversies center around the significance of these intervening factors (e.g. Kaufmann and Pape, 1999; Moravcsik, 2000).

However, a body of literature mostly concerned with studying ‘globalization’ (see Zürn’s Chapter 12 in this volume) goes much further and claims that the transnational ‘society world’ has not only profoundly changed the ‘state world’ but has made obsolete the current order of international relations as an inter-state system (see, for example, Amin, 1997; Gill, 1995; Strange, 1996). We do not live in a borderless world, but political, social and economic boundaries cease to coincide and to be confined to the nation-state (e.g. Brock and Albert, 1995; Ferguson and Mansbach, 1996; Ruggie, 1993; Wolf, 2000). Even talking about transnational as opposed to state actors becomes problematic, the more we accept that the current ‘Westphalian’ system of nation-states is coming to an end and that political authority is increasingly structured along functional rather than territorial lines (Caporaso, 2000). In the following, however, I bracket this discussion, which is taken up elsewhere in this volume (see, for example, Chapter 8 by Bierstekker and Chapter 12 by Zürn). I take a more ‘pedestrian’ and actor-centered approach and ask how and under which conditions specific TNAs such as multinational corporations and ‘non-profit’ INGOs affect outcomes in world politics. I put somewhat more emphasis on the non-profit sector, since the impact of MNCs is treated elsewhere in this handbook (see Chapters 12, 22 and23).

Globalization and MNCs: ‘Global Players’ as Sources of Policy Convergence?

As mentioned above, both the liberal and the critical-Marxist literature on transnational relations of the 1970s focused on the role of MNCs in world politics. At the time, the main controversies centered around the question of whether MNCs contributed to or hindered economic development (overview in Gilpin, 1987: chs 6, 7; see also Chapter 24 by Maxfield in this volume). Realists argued that MNCs were irrelevant for developments, since national government remained largely in control of development policies, even in the less developed world (e.g., Krasner, 1978). Liberals and modernization theory claimed that MNCs had an overall positive effect on economic modernization by guaranteeing an open world economy based on free trade and by exporting capital, know-how and modern values into less developed countries (for an early statement see Huntington, 1968). Critical theorists, particularly ‘dependentistas,’ maintained that, on the contrary, MNCs were among the main culprits of uneven development by essentially extracting resources from developing countries that were desperately needed for economic development (e.g., Amin, 1977; Frank, 1967).

Twenty years later, this controversy has largely disappeared, for two reasons. First, as Menzel claimed, the Third World has ceased to exist (Menzel, 1992). The differentiation process among developing countries led to functionally equivalent paths to economic development (compare, for example, the Latin American experience with South East Asia). As a result, it is impossible to sustain a unifying theory of MNC impact on economic development such as claimed by either modernization or dependency theorists. MNC impact on development varies enormously depending on social, political and cultural structures in target countries (Clark and Chan, 1995).

Second, accounts ascribing an enormous influence of MNCs on less developed countries usually overstate their significance for the local economies. On a worldwide scale, overseas production of firms as a percentage of world GDP has risen from 4.5 per cent in 1970 to 7.5 per cent in 1995, while the sales of foreign affiliates of MNCs have doubled to about one-quarter of world GDP in the meantime (according to Held et al., 1999: 246). These figures already challenge some exaggerated arguments about ‘globalization,’ at least in production. Interestingly enough, the significance of MNCs and their local affiliates is even lower for the developing world, as concerns production as percentage of GDP (6.3 per cent in 1995). While the developing world and Eastern Europe since the end of the Cold War have taken part in the boom of Foreign Direct Investments (FDI) starting in the late 1980s, most FDIs still take place within the industrialized world. In fact, the proportion of FDI stocks in developing economies as compared to the OECD world has fallen from 32.3 per cent in 1960 to 25.3 per cent in 1994 (Held et al., 1999: 249). In sum, FDI and MNC activities are largely concentrated within the industrialized world, both intra- and inter-regional.

As a result of these patterns, the debate about MNC impact on world politics in the 1990s largely concentrated on the developed world in the context of discussions about ‘globalization’ and internationalization (for a useful overview see Held et al., 1999: ch. 5). This controversy is far from over, since the very notion of ‘globalization’ is heavily contested in the literature, let alone the impact of so-called ‘global players’ such as MNCs (see the excellent review in Beisheim and Walter, 1997; also Keohane and Milner, 1996; see also Chapter 12 in this volume). The debate largely concentrates on the effects of MNCs and other transnational market forces on the nation-states, in particular the ability of industrialized countries to conduct their own ‘autonomous’ economic and monetary policies (see also Chapter 22 in this volume).

The ‘convergence hypothesis’ holds that ‘the authority of the governments of all states, large and small, strong and weak, has been weakened as a result of technological and financial change and of the accelerated integration of national economies into one single global market economy’ (Strange, 1996: 13–14; see also Stopford and Strange, 1991). There is widespread agreement that the MNCs’ ability to shift production elsewhere and their capacity as transnational actors to allocate financial and other resources to places promising the highest profit rates severely circumscribe the autonomy of national governments to take economic decisions. The more a national economy is integrated into global markets, the higher the costs of a national economic policy which is not oriented toward liberalizing markets, but toward expansionary monetary and fiscal policies to create full employment. Since the latter policies are usually identified with center-left rather than center-right governments, the former should be more severely constrained in their policies than the latter (Milner and Keohane, 1996: 17–18). The result is a growing convergence of national economic policies toward neoliberalism and monetarism.

Critical theory in the neo-Gramscian tradition agrees with the overall description of recent trends, but explains it differently. Gill and others see an emerging transnational ‘historic bloc’ establishing the hegemony of transnationally mobile capital and relevant capitalist classes. The industrialized nation-states have not been passive by-standers of these trends, but have actively encouraged and contributed to them through, for example, the liberalization of capital markets and the encouragement of FDIs. At the same time and with the demise of Keynesianism, neoliberalism became the dominant ideology of how to run a national economy shaping the world-views of transnational elites, policy-makers and other actors (Gill, 1995; Gill and Law, 1993). In the neo-Gramscian view, it is this confluence of modes of production (transnational), international and national institutions and dominant ideas which constitutes transnational global hegemony. What is less clear in this rather sweeping argument, is how it can be disconfirmed, that is, how do we know transnational hegemony when we see it? Van Apeldoorn has recently applied the neo-Gramscian argument which takes constructivist work on the role of ideas into account, to the European Roundtable (ERT) of businesspeople and CEOs from major European MNCs (Van Apeldoorn, 1999). He claims that the ERT has successfully shaped the economic agenda of the EU toward the neoliberal paradigm, starting with the Single European Act and culminating in the Economic and Monetary Union (EMU). His point is not so much that the ERT lobbied for particular policies, but that it moved the dominant ideology toward an, albeit ‘embedded,’ neoliberal agenda.

The emerging literature on globalization, internationalization and the role of MNCs in the international economy has only started to tackle these questions (see Chapter 12). The more empirical evidence becomes available about the domestic effects of the internationalization of production, the more it becomes clear that generalizations such as the ‘convergence’ hypothesis miss the mark. First, we can observe a transformation of national economic policies rather than a broad ‘retreat of the state.’ Second, the scholarly discussion moves toward specifying the conditions under which nation-states are more or less able to face the challenges of internationalization without giving up social and democratic values. Similar trends toward differentiated arguments can be observed when we look at the literature on the other type of transnational actors to be discussed here, namely advocacy networks and INGOs.

The Power of Principles and Knowledge: Transnational Advocacy Networks and INGOs

There is a growing consensus in the literature that INGOs and other non-profit TNAs make a difference in world politics. Scholars have collected evidence that advocacy networks, epistemic communities and other TNAs can have a substantial impact on state policies, on the creation of international norms and on the diffusion of these norms into domestic practices (e.g., Boli and Thomas, 1999; Checkel, 1997; Evangelista, 1999; Finnemore, 1996a; Florini, 2000; Haas, 1992c; Keck and Sikkink, 1998; Klotz, 1995; Litfin, 1994; O’Brien et al., 2000; Price, 1998; Princen and Finger, 1994; Risse-Kappen, 1995a; Risse et al., 1999; Smith et al., 1997; Willetts, 1996). While these and other works provide evidence that the power of knowledge and of principled beliefs matters in world politics, the more interesting question is why and under what conditions? But many studies do not lead to generalizable conclusions, since they suffer from methodological problems such as case selection on the dependent variable. There are many single-case studies of successful transnational campaigns, while we know much less about failed campaigns (see, however, Cortright and Pagnucco, 1997; Evangelista, 1999). The propositions emerging from the literature can be grouped under the following categories:

• International material and institutional conditions.
• Domestic conditions.
• Complex models linking the international and domestic levels.
• TNA strategies and socialization processes.

International conditions for TNA impact

Realist-inspired authors essentially argue that the more TNAs and coalitions succeed in changing the preferences and policies of the most powerful states, the greater their impact in international affairs becomes. Only great powers enjoy the ability and capacity of affecting outcomes in world politics as a result of which TNAs need to influence their decisions and policies in order to make a difference (e.g. Krasner, 1993, 1995a). The proposition no longer claims to account for state preferences in the international system and, thus, does not explain why great powers sometimes promote INGO goals in international relations. A stronger version of the argument would have to maintain that great powers only promote ‘soft norms’ such as human rights or environmental concerns, if it suits their security and/or economic interests. In this latter formulation, TNA impact becomes more or less epiphenomenal.

Whether in its stronger or in its weaker form, a systematic evaluation of the argument in the human rights area has not shown much evidence for the proposition that TNAs had first to convince great powers in order to influence outcomes and decisions in international affairs (Risse et al., 1999). While it certainly helps if the governments of great powers start promoting the goals of transnational advocacy networks, this is neither a necessary nor a sufficient condition for TNA impact. This work on the impact of transnational networks on the domestic implementation of human rights norms in various regions of the world shows that great powers are rarely decisive in promoting these norms, because they rarely pursue consistent human rights policies. Realism might well explain that great powers do not promote human rights when it does not suit their strategic or economic interests, but precisely for this reason it cannot account for the substantial TNA influence on the implementation of human rights in domestic practices across the world regions.

While the realist proposition has not yet been systematically tested in other issue-areas of world politics, there is circumstantial evidence to challenge it. INGO impact despite great power resistance has been amply documented in the environmental area (e.g. Keck and Sikkink, 1998: ch. 4; Lipschutz and Mayer, 1996; Princen and Finger, 1994), but also in international security. In the case of the international treaty banning landmines, for example, transnational advocacy networks succeeded, even though they were pitched against the fierce opposition of several great powers, including the United States (Price, 1998).

But structure is not confined to the material realm. We need to take the social structure of international politics into account. I argued in my earlier work (Risse-Kappen, 1995a) that TNAs are expected to increase their political influence, the more they act in an international environment which is heavily structured by international institutions. International organizations, for example, provide arenas enabling regular interactions between transnational and state actors. In some cases, they actively encourage (and even finance) INGOs and other transnational coalitions. The European Commission, the World Bank and the developmental sector are cases in point (see, for example, Chabbott, 1999; Imig and Tarrow, 2001). The more international organizations and Western states realized that their developmental policies and foreign aid had to be targeted to the civil societies in Third World countries, the more they came to rely on the INGO world linking the local and the global. The strong collaboration between the World Bank and the INGO world did not result in a less contentious relationship between the two, even though sharp divisions among INGOs emerged concerning how far one should cooperate with the World Bank (O’Brien et al., 2000). The United Nations system provides another arena for INGO participation. The UN world conferences in particular have served as important focal points for the activities of transnational advocacy networks (Clark et al., 1998; see also Weiss and Gordenker, 1996). Moreover, the UN and its various organizations increasingly serve as fora where transnational actors and state officials regularly meet and interact (see for example, Finger, 1994; Willetts, 1996).

It is one thing to argue that international institutions provide arenas in which the activities of TNAs are allowed to flourish. It is quite different to conclude that, therefore, they should have a policy impact on international organizations, international regimes and state policies. The above proposition probably needs refinement. International governance arenas such as the UN system or regional organizations such as the EU certainly provide TNAs with regular access to policy-making. But access does not guarantee impact. As Clark and others showed, NGO influence on policy outcomes of UN world conferences varied significantly according to subject area and countries involved (Clark et al., 1998).

We probably need to differentiate among various phases in the international policy cycle, such as agenda-setting, international norm creation and norm implementation. It is safe to argue that ceteris paribus the influence of transnational advocacy networks has always been greatest during the agenda-setting or ‘norm emergence’ phase of a ‘norm life cycle’ (Finnemore and Sikkink, 1998). Since TNAs provide moral authority and knowledge about causal relationships, they are particularly crucial when it comes to paradigm shifts on the international agenda. One can probably go as far as to argue that there has rarely been a new normative issue on the international agenda which has not been advocated by transnational advocacy coalitions, INGOs or epistemic communities. In the international political economy, for example, an epistemic community put Keynesian ideas of ‘embedded liberalism’ on the international agenda during the negotiations establishing the Bretton Woods system and the GATT (Ikenberry, 1993). In the environmental area, examples include the protection of the ozone layer, global warming, deforestation, wildlife conservation and other questions (Haas, 1992a; Hurrell, 1992; Keck and Sikkink, 1998: ch. 4; Litfin, 1994; Princen, 1995; Raustiala, 1997; Ringus, 1997). Concerning human rights, the origins of almost every single post-Second World War international human rights agreement can be found in the activities of transnational advocacy networks (Keck and Sikkink, 1998: ch. 2; Korey, 1998 see also Chapter 27 by Schmitz and Sikkink in this volume). Examples from international security include the Geneva Conventions, the nuclear test ban debate and–most recently–the treaty banning landmines (Adler, 1992; Finnemore, 1996b: ch. 3; Price, 1998).

Yet, agenda-setting does not equal norm creation. When it comes to international rule-creation and international treaty-making, national governments and IOs assume center-stage again. During this stage of the process, INGOs and transnational advocacy networks need to work through governments or international organizations. Moreover, while the agenda-setting phase might be dominated by well-organized transnational networks and INGOs, they are likely to counter considerable opposition when it comes to transforming principled beliefs or knowledge into concrete norms and rules prescribing appropriate behavior enshrined in treaties and other instruments of international governance and accepted by the international community. As a result, the requirements to build ‘winning coalitions’ with and among state actors usually become extensive and, thus, TNA impact appears to be less pronounced. There are few comparative case studies varying the conditions under which TNAs have an impact on treaty-making or regime creation. The available evidence points to three potential pathways by which TNAs influence multilateral negotiations:

• Through lobbying activities in the domestic society of powerful states such as the United States, thus exploiting ‘two level game’ mechanisms and changing state preferences;
• Through coalitions with international organizations thus pressuring states ‘from above’ and ‘from below’; this particular pathway seems to be pronounced in the EU;
• Through coalition-building with smaller states providing the latter with knowledge and ‘informational power.’

Once international rules and norms are created and international regimes have emerged, these normative commitments need to be implemented in the domestic practices of states and societies. This is by no means an automatic process as numerous studies about rule compliance (or lack of), and rule effectiveness reveal (see, for example, Keohane et al., 1993; Victor et al., 1998). While we still lack systematic comparative studies across issue-areas, the available evidence suggests that transnational advocacy networks and epistemic communities once again assume center-stage in the process by which states and their societies internalize international rules in their domestic practices. There are two reasons for this. First, the legalization process of international norms drastically increases the legitimacy of those actors who demand compliance with them. International institutions and the rules emanating from them empower both domestic and transnational actors in a differential way, thereby enhancing their moral and knowledge power.

Second, international organizations and state agencies must often rely on the monitoring and information capacities of transnational networks and INGOs, because the former are bound by rules of sovereignty and of ‘non-interference in internal affairs,’ while the latter can move more freely. This reliance on TNA expertise and information-gathering capacities is particularly pronounced in issue-areas such as human rights and environment (probably less so in international security and international economy) and probably most relevant concerning international regimes that lack adequate, detailed and intrusive verification procedures (see, for example, Haas, 1992c; Korey, 1998; Peterson, 1997; Smith, 1997).

Domestic conditions for TNA impact

So far, I have concentrated on discussing effects of transnational activities mainly on the international level. Many transnational advocacy networks focus on influencing national policies and national governments in conjunction with international institutions and organizations. One proposition claims that differences in domestic structures explain the variation in TNA policy influence: Domestic structures mediate, filter, and refract the efforts by transnational actors and alliances to influence policies in the various issue-areas. In order to affect policies, transnational actors have to overcome two hurdles. First, they have to gain access to the political system of their ‘target state.’ Second, they must generate and/or contribute to ‘winning’ policy coalitions in order to change decisions in the desired direction … Domestic structures are likely to determine both the availability of access points into the political systems and the size of and requirements for ‘winning coalitions.’ (Risse-Kappen, 1995a: 25)

This argument resembles to some extent the suggestions in the social movement literature that ‘political opportunity structures’ constitute an important factor in explaining the success of new social movements (Kitschelt, 1986; Tarrow, 1996). The domestic structure hypothesis has been evaluated empirically with regard to a variety of countries with diverging institutional setups (case studies in Risse-Kappen, 1995b).

This proposition posits a somewhat inverse relationship between TNA access and TNA impact on the domestic policy-making processes. The more open and the less centralized a political system, on the one hand, and the more pluralist the society, on the other, the easier it should be for TNAs to gain access to decision-makers. The United States probably represents the best approximation of such a domestic structure. Yet, easy access does not equal policy impact. In fact, the coalition-building requirements in open political systems and societies such as the United States are quite formidable. On the other end of the spectrum are extremely centralized political systems such as the former Soviet Union which also dominate state-society relations. Evangelista’s book on the impact of transnational networks on Soviet and Russian security and arms control policies (Evangelista, 1999) confirms the argument that TNA access to the Soviet policy-making structure was extremely difficult to achieve for transnational networks of advocacy and expert groups. Once Soviet leaders were prepared to listen, however, the transnational coalitions exercised an almost immediate policy impact. This explains the variation between the Khrushchev and Gorbachev eras, on the one hand, and the situation under Brezhnev and Andropov, on the other, when access to the top leadership was extremely limited. By contrasting Gorbachev’s Soviet Union with Yeltsin’s Russia, Evangelista also shows that the opening of the Soviet/Russian system multiplied the access points for TNAs, but severely circumscribed their policy impact (Evangelista, 1995). This work constitutes one of the few examples whereby case selection on the independent variable (domestic structure) and keeping other factors constant allows for valid causal inferences on TNA impact.

But, as Keck and Sikkink point out, emphasizing domestic institutional arrangements does not tell the whole story of TNA impact: ‘They cannot tell us why some transnational networks operating in the same context succeed and others do not’ (Keck and Sikkink, 1998: 202). Human rights groups were more successful in changing US policies under Presidents Carter and even Reagan than environmental groups. Their objection points to a weakness of the domestic structure argument. It has so far mainly emphasized formal aspects of political and social institutions rather than the substantive content of ideas and norms embedded in them. Constructivist insights might help to solve the puzzle why some TNAs successfully influence changes in state policies, while other fail, despite similar institutional conditions. A ‘resonance’ hypothesis has been developed by students of international norms trying to explain the differential diffusion in domestic practices (e.g. Checkel, 1997; Cortell and Davis, 2000; Ulbert, 1997): the more new ideas promoted by transnational coalitions resonate or are compatible with pre-existing collective identities and beliefs of actors, the more policy influence they might have.

This proposition can be applied to efforts by transnational coalitions promoting international norms to affect domestic change. A comparison of human rights changes in the Philippines and Indonesia shows, for example, that the arguments by transnational networks resonated well with the Westernized political discourse in the Philippines under Marcos, while similar networks failed for quite a long time to pry open political space for human rights in Indonesia, since their arguments were not compatible with the prevailing nationalist discourse (Jetschke, 2000). In both cases, the domestic structure was strikingly similar during the 1970s (authoritarian rule). The ‘resonance’ argument would also explain the variation in transnational influence on Soviet security policy under Brezhnev as compared with Gorbachev (Evangelista, 1999). Litfin’s critique of the ‘epistemic community’ literature and its application to the case of ozone depletion points in a similar direction (Litfin, 1994). She argues that Ernst and Peter Haas’s conceptualization of consensual knowledge emphasizes too much apolitical and ‘objective’ scientific knowledge. Rather, the claims by scientific communities must be framed in such a way that they are compatible with the prevailing political discourse. Otherwise, they fall by the wayside.

But the ‘resonance hypothesis’ is not unproblematic (see the discussion in Cortell and Davis, 2000). First, assessing the compatibility between trans-nationally diffused ideas and given domestic identities and collective beliefs must strictly concentrate on the discourses rather than on the behavioral practices of actors so as to avoid circular reasoning. Second, if there is a complete match between the new transnational norms and the ideas embedded in a given domestic culture, we do not need conscious efforts by transnational actors to make the norms stick. In other words, a certain degree of cultural misfit or incompatibility is necessary to ascribe causal weight to the activities of transnational actors. Most TNAs, INGOs and epistemic communities alike are in the business of strategic construction (Finnemore and Sikkink, 1998: 269–75). In other words, they deliberately make new ideas and principled beliefs ‘resonate’ with pre-existing and embedded norms and collective understandings. It is very hard to predict beforehand which of these new ideas carry the day. The argument about ideational (in)compatibility is still underspecified in the literature.

Toward complex models of TNA impact

Most recently, scholars have advanced complex models of TNA impact integrating international and domestic levels. This work is particularly relevant for the study of TNA impact on norm implementation and compliance. Scholars have started specifying the conditions and causal mechanisms by which transnational advocacy networks manage to link the ‘global’ and the ‘local’ levels (on norms socialization in general see Checkel, 1999a). Keck and Sikkink have developed the so-called ‘boomerang effect’ model to show how domestic and transnational social movements and networks unite to bring pressure ‘from above’ and ‘from below’ on authoritarian governments to accomplish human rights change (Keck and Sikkink, 1998: 12–13; see also Brysk, 1993; Klotz, 1995). A ‘boomerang’ pattern of influence exists when domestic groups in a repressive state bypass their government and directly search out international allies to bring pressure on their states from outside. National opposition groups and social movements link up with TNAs who then convince international human rights IOs and Western states to pressure norm-violating states. Transnational networks provide access, leverage, information and often money to struggling domestic groups. International contacts can amplify the demands of domestic groups, pry open space for new issues, and then echo these demands back into the domestic arena.

Risse, Ropp and Sikkink have developed a five-phase dynamic model of human rights change consisting of several ‘boomerang throws’ (Risse et al., 1999) and specifying the conditions under which links between domestic opposition groups and transnationally operating networks produce change toward domestic norm implementation and compliance. The ‘spiral model’ of human rights change claims that the mobilization activities of transnational advocacy networks are particularly significant in early stages of the process when domestic groups in the repressive state are too weak or too oppressed to constitute a serious challenge to the regime. At this stage, the information and monitoring capacities of transnational networks as agents of norms change are particularly relevant to mobilize the international community. The more the government is under pressure ‘from above’ and ‘from below’ and forced to make tactical concessions to its critics, the more the center of activities shifts from the transnational to the domestic level. During these later phases of the process, a major effect of transnational network activities is to empower and to strengthen domestic civil society. The spiral model has been successfully evaluated for the human rights area, but there is not yet systematic research for other issue-areas of international relations.

TNA strategies and communicative processes

Structural conditions need to be complemented by agency-centered approaches to account for TNA impact. Work focusing on INGOs and transnational advocacy networks which is largely inspired by moderate social constructivism and sociological institutionalism points to three relevant factors in this context:

1. TNA characteristics, particularly network density, material resources and organizational capacities, but also ideational resources such as moral authority and legitimate knowledge.
2. Target characteristics, such as vulnerability to transnational pressures and uncertainty about cause–effect relationships.
3. Communication processes such as shaming, learning and arguing.

As to the first factor, Keck and Sikkink point out that transnational advocacy networks ‘operate best when they are dense, with many actors, strong connections among groups in the network, and reliable information flows’ (Keck and Sikkink, 1998: 28, see also 206–7). One should add that material resources and organizational capacities of networks and INGOs also contribute to their effectiveness. But the example of Transparency International (TI), a tiny INGO with initially only few professional staff members, which almost single-handedly put corruption on the international agenda, indicates that ideational resources and knowledge might overcome a lack of material power and organizational capacities, at least initially (Galtung, 2000). Within a few years, TI acquired both moral and knowledge power in the area of corruption. The effectiveness of TI and other INGOs depends on ideational resources, particularly moral authority in terms of legitimate claims of representing some international ‘common good’ as well as informational capacities and knowledge. This ability to convert moral authority and excellent knowledge of the issue-area into ideational power explains to a large degree why transnational advocacy networks sometimes win against materially more powerful actors such as MNCs and national governments.

TNA impact not only depends on their own resources and capacities, but also on the vulnerability of their ‘targets’–states, international organizations, or multinational corporations–to network pressures (Keck and Sikkink, 1998: 29, 208–9). Such sensitivity might concern vulnerability to material pressures such as economic sanctions or the cutoff of foreign aid in the cases of many Third World countries. But ‘target vulnerability’ might also imply reputational concerns and normative commitments. States or international organizations might be vulnerable to TNA pressures, because they want to be members of the international community ‘in good standing.’ In other words, the more these actors have committed themselves and their collective identities to the norms advocated by the networks, the more they should be vulnerable to TNA pressures in cases of norm violation. Finally, ‘target vulnerability’ to network pressures might include uncertainty about the situation and about cause–effect relationships. This point has been particularly emphasized by the literature on epistemic communities (particularly Haas, 1992c). In many cases, policy-makers recognize a collective action problem in international life, but simply lack the knowledge to tackle it. Such uncertainty provides a window of opportunity for knowledge-based epistemic communities to exert influence.

Transnational networks as moral and knowledge entrepreneurs use various communication strategies to achieve their goals, the third group of factors to be mentioned here. INGOs rely on social mobilization, protest and pressure. They use strategic constructions such as the re-framing of issues or shaming in order to mobilize people around new principled ideas and norms (Meyer and Tarrow, 1998). Shaming strategies remind actors such as national governments of their own standards of appropriateness and collective identities and demand that they live up to these norms (Keck and Sikkink, 1998: 23–4; Liese, 1999). Advocacy networks and epistemic communities also rely on the ‘power of the better argument.’ They need to justify their claims and to use various communication strategies in order to persuade their audiences to change their interests and policies. Shaming and the re-framing of issues is usually not sufficient to convince others and overcome opposition. INGOs and other transnational actors must engage their audiences in an argumentative process in order to achieve their goals. As ‘teachers of norms’ (Finnemore, 1993), they need to start a reasoned discourse justifying their claims in front of various public audiences. Work on these micro-mechanisms and TNA strategies has just begun. It usually involves detailed process-tracing in (comparative) case studies research. Many case studies are methodologically problematic, however, since they focus on single cases and/or ‘success stories’ of transnational pressures without specifying the scope conditions of their arguments. As a result, we still lack testable propositions on the conditions under which such strategies succeed and when they fail (see, however, Checkel, 1999b). This area certainly deserves further scholarly exploration.

Conclusions: Toward Tripartite ‘Global Governance’?

This survey of more than thirty years of scholarship on transnational actors demonstrates that the significance of cross-border interactions involving non-state actors–multinational corporations, INGOs, epistemic communities and advocacy networks–is no longer seriously contested in an age of globalization. But it would be premature to proclaim the end of the inter-state world as we knew it. The picture emerging from the literature reveals instead complex interactions between transnational actors, on the one hand, and corporate actors on all levels of supranational, international, national, regional and local governance, on the other. Unfortunately, most of the literature is still primarily concerned with proving against a state-centered picture of world politics that TNAs matter. As a result, the more interesting questions–when and under what conditions do they matter?–are rarely asked. Moreover, most research on TNAs focuses on their direct policy impact rather than the structural implications of their activities on international society. This is particularly regrettable, since the goal of many transnational advocacy networks and INGOs is not so much geared to directly shape policies, but to engage in consciousness-raising and, thus, changing societies and building a transnational civil society (Lipschutz and Mayer, 1996; Wapner, 1996). While research on MNCs has always included the structural dimension in evaluating their impact, particularly in the developing world, this aspect is largely missing in studies concerning the non-profit sector. There is no lack of propositions on TNA impact concentrating on institutional conditions (both domestic and international), coalition-building abilities, organizational capacities and resources, and actors’ strategies. Research on transnational advocacy networks and epistemic communities appears to be a fruitful area for probing competing as well as complementary assumptions derived from rational choice liberalism and institutionalism, on the one hand, and sociological and constructivist approaches on the other. But we are still in the early stages of a research program that concentrates more on hypothesis-generating than hypothesis-testing through systematic and comparative case studies. Future research on TNAs should evaluate competing explanations and specify the conditions of TNA impact on the various levels of governance. We also need more studies that turn the ‘dependent variable’ around and investigate the effects of increasing TNA influence in world politics on these non-state actors themselves.

Most important, future research on TNAs needs to take into account that these actors–whether MNCs or principled INGOs–have lost their ‘innocence’ and have become part and parcel of international governance structures. Most previous scholarship has concentrated on the question how transnational actors affect national governments and international organizations through the various channels of exerting influence. Yet, international governance seems to be increasingly characterized by cooperative partnerships involving governmental as well as transnational actors, both MNCs and INGOs (see Cutler et al., 1999; Kaul et al., 1999; O’Brien et al., 2000; Reinicke, 1998; Reinicke and Deng, 2000). Examples include regulations in the financial services sector (Reinicke, 1998: ch. 4), private regimes in the insurance sector (Haufler, 1993), and the Transatlantic Business Dialogue (TABD) initiated by the European Commission and the US government (Cowles, 2000). The UN system is replete with cooperative arrangements including international organizations, national agencies and the NGO community (e.g. Weiss and Gordenker, 1996; Willetts, 1996).

Empirical research on global (or regional) governance by tripartite networks including state actors, firms and advocacy groups has just begun. The evidence is still sketchy and has not yet yielded testable propositions regarding the conditions under which such network structures emerge and how effective they are in international problem-solving. On the one hand, network governance might increase information capacities of both private and public actors as well as lead to new participation and learning possibilities. On the other hand, tripartite governance networks might lead to overly complex negotiation systems and to decision blockages.

Last not least, there is the problem of democratic and legitimate governance beyond the nation-state. How can global governance by increasingly complex tripartite networks solve the dual problems of insuring ‘input legitimacy’ through participation of those concerned by the regulations and of ‘output legitimacy’ through effective and enhanced problem-solving? (On these distinctions see Scharpf, 1999: 16–28; for general reviews see Kohler-Koch, 1998a; Wolf, 2000.) This debate has a long history, starting with the first liberal thinkers on transnational relations for whom transborder interactions were an unproblematic ingredient of liberal democracy and a guarantee for peaceful international relations. But, as Kaiser pointed out already in 1971 (Kaiser, 1971), it constitutes a problem for democratic accountability if transnational governance structures include private actors–be it MNCs or INGOs–who are not elected by and, therefore, not accountable to anybody except, say, shareholders, and members of the transnational organizations. Claims by transnational advocacy networks to represent the international ‘common good’ and ‘global civil society’ have to be taken with a grain of salt when it comes to democratic accountability, repre-sentativity and participation.

Two positions can essentially be distinguished in the emerging debate about the ‘democratic deficit’ of global governance. The first set of arguments is represented by Scharpf’s work on the EU (Scharpf, 1999) and is rather pessimistic about overcoming the democratic deficit of multi-level governance. Increasing democratic participation in international governance networks (‘input legitimacy’) faces the problem that there is no transnational ‘demos,’ no transnational collective identity and no international public sphere as we know them from the nation-states. Thus, ‘output legitimacy’ has to overcome the democracy problem in international governance. But there are limits to increasing the effectiveness and problem-solving capacity of international governance, since it requires positive collaboration among actors with diverging interests (‘positive integration’), while market liberalization (‘negative integration’) tends to be much easier.

The second position is more optimistic. Held’s concept of a ‘cosmopolitan democracy’ (Held, 1995) goes probably furthest in suggesting the strengthening of parliamentarian representation on the level of the United Nations and elsewhere and, the systematic inclusion of transnational civil society and the INGO world into governance mechanisms. The concept of ‘deliberative democracy’ assumes center-stage among those who argue that the problems of democratic accountability in global governance and tripartite network structures can be overcome (see, for example, Schmalz-Bruns, 1999; Wolf, 2000: 213–42). It rests on the assumption that the legitimacy of the political process can be strengthened through public debate and deliberation and through the open exchange of arguments among citizens. The more tripartite networks of global governance are inclusive, their procedures and decisions transparent and subject to public deliberation, the more the democratic deficit of transnational governance can be tackled. Public deliberation might also increase the problem-solving capacity of multi-level governance, that is ‘output legitimacy.’ At least, the concept of deliberative democracy offers a way out to tackle the legitimacy problems of global governance by networks, since it does not require a global ‘demos’ in terms of a strong supranational collective identity.