Organizing International Politics: Sovereignty and Open Systems

Christopher Ansell and Steven Weber

Center for German and European Studies

June, 1997

Introduction

Is sovereignty changing and if so what are the implications for world politics? This question ought to be at the center of intersecting discussions between comparative and international politics. International relations (IR) theories do provide "falsifiable" propositions about sovereignty, the nature of the state, and the politics that follow from that. But many of those propositions have been or are in the process of being falsified by events on the ground. It is fine to talk about "challenges" to sovereignty from above and below and beyond, but only if we know what it is, that is in fact being challenged.

This paper places sovereignty within a broader set of questions, about how world politics is "organized." We specify a baseline perspective on sovereignty and explain why scholars and political leaders continue to take it seriously. We then introduce a distinction from organization theory, between closed and open systems, and argue that sovereignty can be reconsidered in a useful way by moving from a closed system perspective to an open system perspective. This generates a forward-looking research agenda that goes beyond sterile arguments over the erosion or resilience of sovereignty, toward more important questions about the nature of boundaries and the possibilities of constituting new actors whose relationships would follow different logics. We conclude with some theoretical implications and examples from emerging regional structures in Europe.

I. Sovereignty and Globalization

The sovereign state is distinguished in the ideal by several characteristics. The state claims exclusive jurisdiction on a piece of territory over activities, which are then labeled domestic. There is mutual recognition by states of these claims. Mutual recognition constitutes each state formally as an equal actor. The state monopolizes the legitimate use of violence to sustain its authority internally and to protect it from external intrusion. The system of states remains anarchic, without legitimate forms of central governance autonomous from states.

This is a clearly delineated closed-system perspective on the organization of world politics. The central issue is property rights, which are clearly defined by hard boundaries that distinguish domestic from international and authority from anarchy. States are sharply demarcated from "the environment" and from each other. Since they control their own resources, the primary thing that they want to do with each other is to transact, to exchange.

The emphasis on exchange is a powerful theme in almost all the recent IR literature on the emergence of the sovereign state. In Hendrick Spruyt's (1994) account, the world is made up of independent actors who are drawn to each other for the purpose of making trades. Sovereignty emerges out of that because the state, rather than non-territorial and functionally organized alternatives (like City-Leagues) was able to facilitate the deals. States won because they could more efficiently and with lower transaction costs make deals with each other for joint gains (as "preferred partners") than could alternative forms of political organization. David Strang (1991) explains the extension of the state system outside Europe as an outcome of transactions between existing states and their former colonial possessions. Where there was competition for the benefits of exchange, the solution was to recognize and "create" a state. Jan Thompson (1994) argues that states conquered, extruded, or agreed to forego non-state violence so that they could effectively conduct inter-state transactions and maintain the exchanges that provided necessary and beneficial resource flows between them. States took central control of violence in order to meet the demands of accountability that contracts with other states placed on them.

Exchange logic (somewhat ironically) also is central in recent arguments about the transformation of sovereignty. What sometimes goes under the loose conceptual heading of "globalization" is a collection of anecdotal and some systematic evidence that many of the ideal-typical characteristics of sovereignty don't describe reality at the end of the 20th century, because those characteristics have been undermined by transactions (Ruggie, 1993). ₁ Phillip Cerny (1995) summarized three underlying presumptions in this literature:

1. an increasing predominance of structures and processes (mostly economic, but some political as well) that are transnational and larger in scale than states.

2. these structures and processes already do or soon will have a greater impact on peoples' lives than does the state.

3. these structures and processes are empowering individuals, firms, NGOs, and other groups to act effectively and autonomously in world politics.

If these presumptions are correct to any significant extent, the question then becomes what kinds of politics would follow from that. Some authors take the position that "globalization" implies fundamental changes in the relevant actors of world politics and by implication in the politics that happens between them. Others see no fundamental change in who the principal actors are, or what their relationships are like.

The debate bogs down because it is nearly impossible to adjudicate between these claims. There exists no agreed metric for what constitutes fundamental change in the actors. Without that, the argument rests on deduced behavioral effects; that is on the nature of politics in the international system. But because states and sovereignty permit such a wide range of behaviors, and because those who argue in favor of fundamental change have not spelled out precisely what kinds of politics would follow in their vision, this doesn't help much either. The erosion of sovereignty becomes a yes/no discussion without much significance for empirical research. And while some analysts claim the technically-driven, functional obsolescence of the nation-state to be a nearly obvious fact, what is even more obvious is that political leaders and mass publics all over the world still identify states as their primary legitimate political association. Is sovereignty nothing more than a symbol?

This is a critically important question because the conception of what sovereignty actually is, lies at the center of much of the theoretical apparatus in IR. Sovereignty specifies fundamental principles of order in the international system. It postulates ways in which that order affects the behavior of actors. In some instances, sovereignty posits who the actors are, how much power they have, even what kinds of interests they pursue. Sovereignty usually invokes a story about how these principles came to be and how they persist (Krasner, 1989). If world politics were coming to be organized according to principles other than state sovereignty, IR theories would have to consider different logics of explanation.

II. Sovereignty and Open Systems

The question we pose is quite general. It abstracts one step away from the sovereignty debate as it is currently configured. It is, simply, how is world politics coming to be organized? To answer this question we move away from the yes/no character of sovereignty, and drop the implicit assumption that there is a zero-sum relationship between states and globalization -- that what global processes gain, the state necessarily loses. A broader conception of what kinds of organizational principles are possible in international politics comes out of theories about "open-systems."

The study of organizations in the 1970s moved decisively away from representations of closed systems and toward "open-systems". The major advantage of open-systems theories is that they take on porousness of boundaries as a central focus of investigation. We use porousness to mean the existence of fluid and multidimensional boundaries, in contrast to the solid and unidimensional borders that constitute the ideal-typical sovereign state that fits comfortably in closed systems theory.

Increasingly porous boundaries are a major feature of modern international politics. Property rights and liabilities, as well as authority and control over transactions, are increasingly ambiguous. What is an "offshore" market, a "global" factory, a "cross-national" production network, or "international" finance? These awkward terms are an attempt to come to grips with the realms of transactions, as well as other kinds of relationships, that are disconnected functionally from a piece of territory. Some may take place on a piece of territory -- after all, as Richard O'Brien (1992) points out, "everybody has to be somewhere" and that is particularly true of a piece of physical matter like a factory. It is even true, in some sense, of a piece of data which represents a piece of money, which sits on the hard drive of a computer, which sits on a desk -- somewhere.

But the significance of that somewhere becomes a different matter when it might as well be anywhere, and can become so, at very low cost. Is it meaningful to say that the US Federal Reserve owns the property rights to dollars traded in so-called Euromarkets, for example, when a transaction takes place between the London office of Nomura Securities and an oil trader in Kazahkstan? These transactions don't need to be -- and in fact probably are not -- "located" anywhere, at least not in the standard geographical sense of the term.

"Location, location, location" is what real estate agents say are the three determinants of the price of a house. Location does give a lot of information about what it is possible to do (and even, in a social sense, what it is possible to be) for the potential buyers and residents. But if it were all just about location, and "real estate" were as "real" as it sounds, the action of exchange -- buying and selling a house between willing transactors -- would be easy. It isn't easy. Part of the reason is simply imperfect and asymmetrical information about the commodity that is about to be exchanged. The more interesting, and analytically complicated part of the reason is uncertainty about property rights. Who "owns" the house -- in the sense of having residual rights and claims and responsibilities for one or another aspect of the property -- is never as clear as it seems. The actual experience of buying a house is much more like a combination of archaeology and sociology than it is like market economics. And analytically it is much messier than the phrase "housing market" seems to imply.

But just as liens and liabilities and title complications make it more complicated for people to engage in real estate exchange, introducing porousness of boundaries muddies the waters of traditional IR theories. It is a derogation from the clean logic of property rights and exchange in the anarchic "market" of international politics, undermining a key metaphor in IR. It lessens the power of arguments about the dominance of systemic effects deriving from international structure. Porousness chips away at the theoretical world we think we know but does not seem to replace it immediately with anything in particular, which is why it has been so hard for IR theory to take stock of its significance.

Open systems thinking in OT puts porousness of boundaries at the center of the discussion. It shifts the focus of debate to the conceptualization of newly relevant and emergent actors other than the state. Part of the promise of OT is that it allows for stable and significant forms of organization that lie off the continuum of states and markets. In other words, sovereignty created two categories of organizing principles: politics, with coercion, power and violence at its center; vs. economics, which is about pareto-improving voluntary exchange. Forms of organization in the middle, or off the continuum entirely, appear to be vague, unstable, and marginal. But that is a statement about a particular conceptual structure, not necessarily about the real world. In conceptualizing how other actors come to be and what it is that they do, OT develops a larger variety and richer understanding of possible logics of organization for a system made up of these actors.

III. Organization Theory and Open Systems

Exchange logic is a clean way of understanding organization in a world where constitutionally separate units transact in things that they own without encumbrances or liens, among themselves, and across solidly demarcated boundaries. But if property rights are muddy and boundaries are porous, the logic of exchange -- even between constitutionally separate units, if they continue to exist -- gets very complex. Modern OT tries to integrate porousness directly, via "open systems thinking."

The move towards "open systems" thinking in organization theory began in the 1960's and 1970's, in part as a delayed reaction against Max Weber's theory of bureaucracy and Frederick Taylor's Principles of Scientific Management. ₂ Instead of viewing the organization as an autonomous and sharply bounded entity, open systems approaches see organizations as simultaneously continuous with and demarcated from their environments. They are continuous with their environments by virtue of the constant and necessary interchange of resources with their surroundings. But they are at the same time demarcated from the environment by a disjuncture in the degree of interaction. Internal interactions are more densely packed than are interactions between the organization and its environment.

This is a fluid process. The boundary between organization and environment is a dependent variable of that process -- dependent in the sense that it emerges from the interchange, and variable in the sense that it can move or reformulate itself as patterns of interaction shift. In open systems, interchange across porous boundaries is what constitutes the unit and marks it off from the environment. The boundary is enacted, not assumed. Boundaries, then, become the major developmental causal force for organizations and the systems in which they exist -- much like membranes are the developmental causal force for cells in biological systems.

James Thompson (1967) argued that models of rational organization naturally gravitate toward "closed systems" because rationality presumes an internalization of control of over tightly interdependent variables. ₃ Moving toward open systems generally means giving up formal notions of rationality, because the "actor" part of the "rational actor" is now contingent. It would be easier to make sense of this analytic shift in international politics if states were primarily functional organizations, if they were organized around getting tasks done in an efficient manner. But that is not the case now and arguably has never really been the case -- states are multifunctional, territorial in nature, and hence highly inefficient. As boundaries become more porous, and as non-state actors increasingly take on functions previously done by states (and in many cases do so more efficiently) the theoretical advantage of the sovereign rational actor model becomes less and less compelling.

What arguments can be deployed to replace it? Transaction-cost economics (TCE) takes an important step towards open systems thinking by adopting the "transaction" as the central unit-of-analysis (Williamson, 1996). This, in one sense, turns boundaries into a dependent variable. The boundaries of organizations depend upon the nature of the transactions that are taking place. And presumably those boundaries can move and reconfigure themselves as the nature of the transaction changes. We say presumably because TCE by itself is not a theory of how change takes place. It is an argument that specifies an optimal configuration of boundaries. And, while it carries an implicit presumption that the world will come to match its predictions, it does not itself specify the means or processes by which that will happen. This is a severe analytic handicap during periods of rapid historical change, when the notion of re-organization having reached some discrete endpoint seems more hopeful or even quaint than realistic, and when the process may actually be a whole lot more interesting and relevant than the prospect of an outcome. ₄ Also important is that TCE is not an argument about symbols or even about meaning. This is a handicap, because it is not easy to see how relationships which have important symbolic components (even if they depend on exchange) could be analyzed in this framework.

Arguments that come under the heading "New Sociological Institutionalism" (NSI), and particularly network theory, are stronger in this area. ₅ The key value of NSI is that it shifts the focus to the level of "organizational fields"-- which are sets of interacting organizations. Legitimacy becomes a central resource. Organizations and environments are seen as mutually constitutive. In practice, this means that boundaries are enacted through continuous processes and can be reconstituted differently over time. Unlike TCE, NSI carries with it no presumption of "optimal" outcomes or unique, efficient equilibria. Inefficient outcomes may be stable, because they will not be challenged, certainly not from within -- because mimicry and legitimation, not competition, are the primary things that go on between organizations.

Network theory carries this inter-organizational perspective one step further. Linkages between organizations, rather than organizations themselves, are now the central analytic focus and linkages are at least partially constitutive of the participating actors. In other words, network theories try to derive boundaries not by assumption, and not from an analysis of transactions, but from relationships, which are assumed to be prior. These relationships certainly include exchange and transactions but can and do go beyond those -- to symbols, meaning, legitimacy and other intangibles. The properties of networks are not reducible to the actors or to the transactions between them. The existence, formation, and reformation of boundaries is an outcome of congealed relationships within the network, nothing more.

The importance of open system thinking is not that it makes the issue of organizational boundaries irrelevant -- quite the opposite. It is that it takes an assumption or set of assumptions about boundaries and drops it, in favor of making boundaries and their behavioral implications contingent. And because boundaries are porous, they are expected to be sites of intensive specialized activity that mediate between the organization and its environment. By making boundaries more contingent and porous, the open-system perspective also forces boundaries to do more real analytical work. Focusing primarily or solely on exchanges between autonomous actors limits that discussion. Open system thinking revives it in at least three ways that could be useful to the IR discussion. The next section takes up these three points -- perspectives on interdependence, the development of boundary spanners, and the constitution of new actors in those processes.

IV. Open Systems

1. Interdependence

IR theory generally treats interdependence as an independent variable. That is, the standard presumption takes interdependence as a condition in the world and tries to explain how it drives or influences behavior. ₆ Interdependence may arise through technological change, through the gradual evolution of long standing political structures, or other means -- but that process is not the major analytic focus. The question becomes how do states and other international actors respond, how do they manage interdependence. There is a standard (but not exclusive) answer -- through bargaining. Explanations then focus on the strategies and structures that states use to control the consequences of a condition in their environment. On this logic, interdependence is a risky situation which (despite its potential value in some circumstances) states would ideally not be in .

Cooperation and conflict, terminology from game theory that evolved into stock metaphors of IR, are each a generic kind of response to interdependence. So are regimes -- moderately institutionalized sets of norms, principles, rules and decision making procedures that provide information and other conducive means of managing interdependence over time (Krasner, 1983). Exchange logic plays a central role in this story, since regimes are partly about capturing economies of scale in the production of the kinds of reputation that states need for relatively efficient exchange in an anarchic environment. Superpower nuclear arms control as it was thought about during the Cold War was in many ways the minimalist exemplar of this logic. It was about two states locked in an unfortunate but inescapable situation of interdependence, responding with limited measures to reduce the risk that interdependence might literally blow up in their faces.

Bargaining to manage interdependence is seen as a strategic game where interests, power, and constraints are the key variables. The two-level game metaphor (Putnam, 1988) captures this perspective and illustrates (not coincidentally) the conceptual reification of the boundaries between the two games that makes this a closed system perspective. In the two level game metaphor, states are the principal actors in two separately encapsulated games. States have divergent interests, and they bargain among themselves toward "solutions" and cooperative arrangements. This inter-state bargaining game is constrained primarily by a second level game that each state must play out among its domestic political interest groups. Successful outcomes rest in the intersection between the domestic "win-sets" of each major state actor and the international "win-set" of overlap between them. This might work analytically if the parameters of the games, and most importantly the boundaries between the two games, are stable and if the relevant actors and interests can be read off some larger set of assumptions (for example, about economic position). If those conditions do not hold -- that is, if the domestic and international "games" interpenetrate and intermingle in ways that confuse not only what the interests may be, but precisely who the discrete actors are -- then the two level game metaphor is simply irrelevant.

"Complex interdependence" took a single important step toward challenging that roadblock by imagining a world where a multitude of bargaining relationships were being negotiated at once, and continuously (Keohane and Nye, 1977). Quickly this international system would become so complex, multifaceted, and intertwined that two-level games would look more like "spider's webs" of potentially intersecting and overlapping bargains (both to analysts and international actors). But, relationships are still principally about bargaining between discrete actors.

Open systems perspectives take an additional step. The key difference is in arguments that see the interorganizational network of bargains as an "emergent phenomenon" with characteristics that are objects of investigation in their own right. This involves a shift of perspective, from seeing relationships of interdependence in strategic terms, to seeing these relationships in developmental terms. Are the actors themselves changed by the experience of interdependence? This is a central claim of constructivist perspectives in IR theory. It can be taken further. Some network theories argue that positive by-products of relationships, like trust, have significant implications for global patterns of organization (Granovetter, 1985; Putnam, 1993) This opens up the question of where the boundaries are, and why. Firms (or other "assumed" actors) are themselves networks of negotiated interdependencies. In other words, it might not be just the interests of states that change as a result of interdependence. It could be the functional boundaries that demarcate the major actors, as well. Boundaries, in this story, are becoming quite plastic. The next step would be to relax assumptions about the basic functional equivalence of units, in favor of looking at processes of differentiation and the development of divisions of labor both within and without. The market or the international system, and the nature of competition within it, changes, because the identity of the actors within it has changed.

Interdependence can be something that actors create voluntarily and willingly, or even sometimes seek out. Put differently, actors can be creative not just clever in a strategic sense. Rather than being stuck in the realm of "strategic responses to interdependence" they can (and do) innovate to create new forms of organization and novel ways of doing things. Cooperation, in contrast to the limited game theoretic metaphor, becomes a means to acquire resources that are needed to establish new interdependencies. Edstrom et. al. (1984) call this an inside-out perspective, but what it really does is to erase the sharp distinctions between inside and outside. Firms seek out new interdependencies to gain new customers, suppliers, relationships, and ideas, all of which are enriching and utility-enhancing for the firm, but also begin to change its identity. The image is one of institutionalized cooperation as organizational success, not merely a cure for market failure (Lazonick, 1991).

IR theory is reluctant to move in this direction, in large part because there is no comfortable way (in the mainstream, at least) to make the articulated shift away from functional equivalence of units and then to talk about changes in the international environment and research agenda that might follow. The question then becomes, is this intellectual constraint mirrored in the real world? If not, then it makes sense to try to think about interdependence not only as an independent variable but more nearly as open-systems perspectives suggest. Interdependence could be something that states create. They do this not only to take advantage of potential gains, but also simply because they can and are not threatened by doing so. Experimenting with interdependence may not only be more possible and more prevalent, it may also be more important in delimiting the functional boundaries of the relevant actors in international politics.

This leads to a different perspective on conflict and cooperation as categories of behavior. In 1991 one of us wrote that superpower nuclear arms control during the Cold War could be understood as an attempt by two states to seek out interdependence in a realm where they were actually capable of providing for themselves (Weber, 1991). With survivable nuclear second-strike forces in place, these two states achieved historically unprecedented levels of self-reliance for their core security needs. Yet almost precisely at that moment, they moved to make themselves more interdependent by concluding agreements aimed at "stabilizing" the nuclear balance by cooperative means. This "enacted interdependence" changed in important (and unanticipated) ways the political relationship between the US and the USSR. Although the changes may not have been generally beneficial, that does not refute the importance of what happened. The thrust of the effort was to create voluntarily, through enacting interdependence, new forms of organization and ways of doing things. If interdependence could be sought out in a realm as central to states and international relations concerns as basic security and territorial integrity, why not elsewhere as well?

Arguments about alliances, economic regions, integration schemes (including but not limited to the European Union) would then open up to more than variations on the theme of strategic response to interdependence. In current IR literature, alliances are generally seen as a strategic response to power or threat. States are supposed to choose alliances to enhance their security and do so at a minimum cost to their own autonomy. Does this story, which sees interdependence as a constraining independent variable, really capture what has happened in NATO over the last 45+ years (Weber, 1992)? Similarly for the EU, which standard IR literature portrays as a product of internal interdependence between the European member states themselves or external interdependence of Europe as a region with the world economy. According to this perspective, NAFTA exists because of the geographic interdependence between the US and Mexico, and it is principally a management tool aimed at reducing the potential damage that this interdependence does to US interests (economic and importantly in this case, security). None of these arguments is wrong. They all capture a part of what is happening in each case, but only just a part.

The open systems view of interdependence does not ignore or deny the postulate that extant units will desire to retain autonomy. But the reification of boundaries that define autonomy is gone. Now it is a very different discussion -- about vested interests, sticky mindsets and institutions, power relations, and so on -- but not an assumption or an ontological absolute. If the ideal type has porous rather than concrete boundaries, the conceptual distinction between exchange and cooperation lessens. Cooperation retains risks, surely, but risks are only a partial concern because you can't place at risk autonomy that you don't really have to start. The issue in some circumstances is not about interdependence per se, but simply about whom you are going to be interdependent with. Put differently, the question is with whom do you choose to cooperate? The answers are potentially much broader than strategic responses to interdependence suggest. Four possibilities are: people like you (in various attributes); people with complementary resources; competitors; people with whom you do a lot of exchange.

Governance structures to facilitate and manage these relationships will differ depending on the kind of interdependence that is enacted. In closed systems arguments where relationships are primarily about exchange between autonomous actors, governance is likely to be organized at least in part according to the kinds of transaction cost constraints that Oliver Williamson analyzed. But relationships between other kinds of actors can be stabilized in other ways, since the relationships themselves will be based on different principles than those we associate with autonomous states and international anarchy. Trust, for example, or shared normative understandings facilitate governance structures that are focused more on achieving stability, than on deterring defection from agreements by opportunistic and guileful actors who can appropriate and maintain property rights over the gains from cheating.

2. Boundary Spanners

As boundaries are increasingly porous, they are also sure to become sites of intensive specialized activity that mediate between the organization and its environment. All organizations develop such actors, sometimes individuals and sometimes large subsets of the organization. They aim at modulating, regulating, and sometimes controlling what kinds of resources, signals, information, and ideas pass in and pass out of the semipermeable membranes that are the boundaries of the organization. Over time, at least in stable environments, what starts as a set of specialized functions often develops further into specialized actors. These are boundary spanners.

International organizations are an obvious example in world politics. Conceptualized narrowly as mechanisms to reduce transaction costs for bargains between autonomous states with particular self-regarding interests, IO's fit comfortably within a closed system perspective. Conceptualized more broadly as centers of negotiated meaning in which state actors constitute and reconstitute at least part of what it is that they do and are, IO's are boundary spanners with direct causal influence not only on what kinds of deals are possible but also on who the actors are and what they want to deal about. This broader perspective has a long theoretical tradition, that is gaining increasing support from contemporary empirical research in what is now being called "constructivist" IR theory.

Clearly, de jure definitions of sovereignty have been deeply subject to consensual meanings negotiated out in IO's. To join the United Nations is a symbolic act, but it would be much more than that as well for an entity like the Palestinian National Authority. IO's are part of what states use to mediate between themselves and the international environment. But IO's don't do that solely from the outside, as external or exogenous actors. They reach through boundaries and into the domestic structure of the actors as well. Martha Fennimore (1996) calls this "teaching" (of states by IO's). It is that, in terms of knowledge and meaning, as well as the transferring of standard kinds of resources that have to travel through boundary spanning structures like Foreign Ministries, science bureaucracies (in one of Fennimore's cases), and central banks. These are not new ideas or arguments, but they have been recently invigorated along with the historical expansion of both the number of IO's and the functions (as well as realms of activity and meaning) that IO's are taking on.

The IR literature has paid less attention to some other kinds of boundary spanners and particularly those in the private sector. The modern management consultant, for example, might be thought of as a kind of epistemic community but one without any likelihood or inclination to formalize its influence in international regimes (which may be why the IR literature pays less attention). John Meyer (YEAR) exaggerates slightly when he argues that the management consultant carries knowledge that claims to be "at the disposal of the general human good anywhere and for all time". But he is right to suggest that these boundary spanners carry something that they (and many others, including the clients) believe is as valuable to a South Korean car manufacturer as to a French biotechnology firm. Transferring resources in the traditional sense across borders is only part of this function. Boundary spanners carry knowledge and ideas, too. They also carry legitimacy, which in a narrow sense might be seen as an accepted set of procedures and techniques that spell out how to go about a particular production process, as captured in the phrase "best practice."

Legitimacy can be seen as well in a broader sense, as an accepted set of statements about what it is possible to be, what kinds of identities make sense. The "modern state" is defined in part by a science bureaucracy (even if there is no domestic research establishment), by an airline flagship carrier and by certain military capabilities (among other things) even if there is no need for either or sufficient resources to support them. The modern business firm is similarly tied to legitimacy expectations, whose agents are the management consultants, international banks, venture capital firms and other boundary spanning actors of modern capitalism. Individuals too may be increasingly subject to these pressures, through the expansion of global information and entertainment providers (another key set of boundary spanners) who work through cables, satellites, and now the internet. There are new identities: the "global manager" who moves more or less seamlessly from running a factory in Jakarta to running a distribution center in Sao Paulo; or the "global teen" who may live in New Delhi, Beijing, or New York but regardless wears Levi's jeans, listens to the latest U-2 album on a Sony Walkman while sipping a Coca-Cola. Increasingly porous boundaries make possible this detachment of identity from location. Surely, new identities are continuously in battle with older, retro-nationalistic, and pseudo-tribal identities that by no means have disappeared. Globalization does not wipe away old vested interests or the imagination that is at the heart of imagined communities -- witness the breakup of Yugoslavia. But boundary spanners do bring forward a new menu of choices, some substantially different in content and structure from what came before.

3. Constituting New Actors

The porousness of boundaries and the increasing influence of boundary spanners does not lead necessarily to homogenization of ideas, cultures, or even organizational forms. Driving forces toward isomorphism are only part of the story. Surely old lines of fracture and differentiation that previously were protected behind national borders are subject to challenge. This leads to periodic waves of concern about a dulling of identity and difference. In the European Community after 1986 it was common to hear worries that a single market would mean homogenization toward a single, least common denominator culture where British beer would be cold and French cheese pasteurized. A few years later the same anxiety went global in Fukuyama's (1992) "end of history." In 1997 the argument turns up in fears about globalization, the supposed "triumph" of the market, as in Kuttner's (1997) Everything for Sale: the Virtues and Limits of Markets or Greider's (1997) One World Ready or Not: the Manic Logic of Global Capitalism. The anxiety tends to divert attention from another and possibly more significant aspect of porous boundaries and the growth of boundary spanners. This is the development and constitution of new actors, separate and different from each other, who congeal within the networks that link together the old organizations, whose boundaries are indeed breaking down.

Open systems perspectives have increasingly come to see networks that link boundary spanners together as themselves being quasi-organizations -- or "network organizations" that develop their own corporate identities. ₇ In some cases, as with "brokers" that span "structural holes" in networks -- the network organization is one of the most powerful actors in terms of its ability to influence overall outcomes (Nohria, 1992).

Woody Powell's (1990) classic study of academic publishing firms captures the essential logic of this process, by showing how boundary spanners along with their clients develop into semi-autonomous organizations of their own. Editors here are the boundary spanners. While they try to mediate between the publishing firm and the authors, it is often unclear in their minds (and the minds of others) whether their role is to represent the firm to the client, or the client to the firm. Frequently they become disgruntled and move to a new firm, taking with them their retinue of authors as well as their links with other editors. In the extreme case, as Crozier and Friedberg (1980) suggest, what happens is the "colonization of the organization by the gatekeepers" or the setting up of independent monopolies by the gatekeepers -- changing the boundaries of the relevant actors.

Contemporary discussions of international finance are beginning to move in this direction. Finance is probably the most "globalized" industry to date, in terms of geographical spread and independence of transactions from physical boundaries, time and traditional avenues of exchange. But, debates about the consequences of this change have tended to an almost surreal dichotomy. One side tends to argue that finance is escaping from states into a nearly perfect "market world" where transaction costs approach zero, information flows freely, and everyone is empowered to act. States may struggle to retain control, but they are doomed to lose this battle for reasons parallel to why state sovereignty came to characterize the international system several hundred years earlier -- the logic of exchange. Even the regulation of property rights is gradually being taken out from under the control of states as finance finds ways to repackage property as diversified portfolios of risk -- whose boundaries are arbitrarily configured in the minds of the mathematicians who develop new "products." The flip-side of this state vs. market debate is that internationalization occurs at the behest of states: states have granted some freedoms to finance because it serves state purposes; states still control the most important factors in the equation and could mostly "reverse" this on-going liberalization "process" if it no longer served their important interests. ₈ These portraits are set up as two ends of a continuum, with the supposedly "realistic" assumption being that the present reality locates somewhere along that line.

The open systems perspective suggests that financial globalization can be seen not just as a loss of old political ordering power in favor of the market, but as encouraging the constitution of new types of powerful international actors. This is fundamentally consistent, actually, with the way in which the emergence of order has traditionally been understood by the principle of sovereignty. Boundaries are centrally important in the creation of order because they construct actors and it is the routinized interaction among actors that produces order. Open systems thinking enlarges this conception, by emphasizing that different types of boundaries that emerge through the more intense interaction of the gatekeepers might lead to the constitution of different types of actors.

What might constitute new boundaries and new actors in international finance? An example would be informal or tacit knowledge, embedded in cultures and ways of doing business, that underlie the creation of new personal networks effectively bounded by borders of impacted information. States as semi-bounded economies still contain many of the value-creating activities that are the basis of finance, and these boundaries continue to mediate the activities and organization of transnational relationships to some degree. Personal networks in finance act as boundary spanners, creating network organizations that exploit arbitrage opportunities partially created by the still-present boundaries of national economies. Put differently, transnational networks do not represent the simple collapse of sovereignty -- in fact, they owe their existence to the continued relevance of national boundaries.

This way of viewing the problem makes it easier to elaborate the claim that finance networks are gradually becoming more important as motivating forces for behavior. The important step in enriching this claim is to analyze more precisely the differences that might result from that shift, by focusing on the ordering principles that are being established in that system. Perhaps the order in this world is more like the "old boys networks" that trade municipal bonds, than it is like the seamless vision of post-sovereignty global finance. In other words, it is closer to an embedded order than to a competitive one. ₉ The new old boys network is a different kind of corporate actor. It cuts across states and across trans-national firms. The finance professional in the offices of Merrill Lynch on Wall Street is more closely tied to and interdependent with a trader who works for Daiwa in Tokyo, than she is to either the United States or the person sitting at the next desk in New York. Her job is to seek out new partners and interdependencies to expand her opportunities and investment base, but she only does that within the borders of her impacted information set and personal network, because it is very difficult to step outside. Within the network there is more trust than opportunism. People make asset specific investments without relying on formal governance structures. Smart entrepreneurs may indeed design new ways of formalizing cooperation in this world -- one cartoon example is the "virtual corporation" which has no physical presence and changes personnel for a single transaction. These innovations (by definition, really) are hard to see before someone invents them (could anyone have "predicted" the emergence of the matrix corporation or multi-divisional firm?) But this kind of innovation is precisely the essence of change. It is organizational success, not market failure, as Lazonick (1991) put it. Innovation doesn't "perfect" the market; instead it builds organizational capabilities that supersede market coordination to some degree.

V. Implications of the Open System Metaphor

The concept of an "open system" is an organic metaphor. It implies that the "system" cannot be mechanically decomposed into component parts without losing some critical information about the organization of both the units themselves and the behavior of the system as a whole. An "open" system is one that is integrated through an organization of flows across the porous membranes of individual units For a social system, the analogy is the flow of resources, people, and ideas, across the boundaries of organizations and social groups. A major attraction of this "open system" view of sovereignty is that it breaks with the more mechanical conception of sovereignty of realists. In classical realist imagery, both the units (states) and the system (international system) are jointly defined by the principle of sovereignty. The behavior of both the states and the international system can be adequately understood by decomposing the system into its component parts (sovereign states).

The power and increasingly the limitation of this mechanical view is that sovereignty is always measured against the baseline of an anarchic international system. This baseline is inherent in the ability to decompose the system into its component parts without loss of information. When the international system is anarchic and institutionally "thin", this mechanical conception yields important insights into the behavior of states. But with the increasing organization of international relations and the complexity and density of flows between states, the international system becomes institutionally "thick," which is then read as an erosion of sovereignty. In contrast, an open system view suggests sovereignty is becoming "contextualized" within the setting of a larger system of activities and organizations. Sovereignty then is just one principle among a set of principles that define states and the international system--not the sole principle.

There are two very immediate advantages to contextualizing sovereignty within the framework of a broader open system. First, the open systems view of sovereignty goes beyond the worn dichotomy of "states versus markets" that dominates the current discourse on globalization. Although often couched as a critique of realism, this discourse takes the realist conception of sovereignty at face value: the international system is anarchic and therefore the cross-border mobility of capital undermines the principle of state sovereignty. Globalization, therefore, replaces an international system of states with an international system of markets that have minimal (and decreasing) socio-political contexts. In contrast, by understanding sovereignty in contextual terms, the open systems view is closer to a Polanyian view: sovereign states and international markets are mutually embedded (Polanyi, 1944).

The second advantage of the open system perspective is that it provides an alternative to equating internationalization with the erosion of sovereignty. From an open system point of view, new forms of international cooperation do not imply the erosion of sovereignty. Realists, of course, make such an argument, but their conception of sovereignty forces them to argue at the same time that states control international organizations. In contrast, the open system view suggests that sovereignty may remain a basic structural principle around which the international system evolves, though it is likely to recede in importance relative to other principles of structuring international relations. We can expect the role and meaning of sovereignty to evolve, but that doesn't mean it will erode. Sovereignty will remain the "deep structure" around which an open system takes shape. Support for this view comes from the haste with which the new Post-Soviet republics donned the mantle of state sovereignty. Flows of international resources and legitimacy require constructing states around the principle of sovereignty, and surely that is an important reason why political leaders take it seriously. Sovereignty is the price of admission to the open system.

An open system view suggests caution in claims about the "challenges to sovereignty" posed by globalization. As a phenomenon, globalization encompasses a whole set of activities that escape the traditional purview of state control. These activities probably originate or end up within the borders of some nation-state. The failure of states to control them is claimed to constitute a challenge to sovereignty. But sovereignty is only "challenged" if this discussion assumes that a sovereign state should have ultimate control over everything taking place within its borders. Many activities associated with globalization are novel, such as the transmission of electronic data within multinational firms. For the most part, states have given up trying to directly control these transmissions (Drake, 1993). Most likely, they would not be able to do so even if they wanted to. However, this constitutes an erosion of sovereignty only if states once had the ability to regulate the flow of information between firms in different countries. For the most part, they really did not.

A private firm provides a good example of the way in which sovereignty remains viable even when control of situational variables is highly contingent. The firm is sovereign in the sense that it retains the right to decide how to proceed with its business. This sovereignty is, of course, circumscribed. With the exception of company towns, firms generally do not control the communities in which they are located. ₁₀ The employees of firms have families but firms do not ask employees to sever these relationships. The firm depends on employees with certain skills, but those skills are often honed in institutions outside the firm. The firm depends on and complies with legal regulations determined in legislatures and courts. Still, the firm is sovereign over a distinct set of managerial decisions. The firm lacks control over many of the parameters of action within its own midst, but it continues to survive, reproduce, and even prosper. It continues to make authoritative decisions. The boundaries of managerial discretion are often challenged, but no one questions that firms continue to have important areas of sovereignty.

In an open system, sovereignty may be best preserved and reproduced by knowing what not to control and what not to do. Though cities or modern families might provide a good example, we extend our analogy to the firm. A cliché of modern management theory is to "stay lean" and focus on core competencies. Firms are told to stick to their central missions and to off-load whatever lies outside their special expertise. In other words, successful firms know what they cannot easily control and efficiently manage. Increasingly, firms externalize certain functions so that they can better focus their time and energies on the value-added represented by their special expertise. This advice is similar to Thompson's (1967) argument, mentioned earlier in the paper, that the organization can succeed in open system environments when it protects its "technological core." By analogy, in the face of globalization, preserving sovereignty probably means giving up (or not taking on) governing functions that you are not capable of doing well.

The European Union is sometimes described in these terms -- the EU as a strategy designed to preserve (not erode) sovereignty (Milward, 1992). This argument might seem to justify an intergovernmentalist view of the EU. If the EU preserves sovereignty, we should expect to see nation-states having clear control over EU institutions. But the open systems perspective and the analogy to the firm suggests a different interpretation. Successfully reproducing sovereignty is not about controlling the environment, it is about creatively adapting to it. If the firm spins off its subsidiaries or vertically disaggregates its manufacturing operations, it does not then continue to try to directly control these now externalized operations.

EU institutions have typically been set up on a delegative model. States are the principals and EU institutions are seen as agents. Yet it is also clear that this delegative model leads to serious institutional tensions. In the worst case, it leads to institutional stalemate as states layer controls on EU institutions. A major drawback of delegation in a multilateral framework is that it leads to enormous complexity. The result is often precisely the opposite of what is intended: non-transparent forms of governance that work through elaborate policy networks. As more activities require an international response, the delegative model is likely to be stretched to its limits.

A more advanced open systems architecture in Europe will develop when states recognize the limits of the delegative model and begin to take the de facto or de jure independence of international institutions for granted. The advantage of this system is that it would allow the member-states to refocus their energies on their own "technological core." Network theory has begun to recognize that the advantage of the network form is that it allows an advanced division of labor to emerge within a field of organizations while preserving strategic and operational flexibility for individual units (Lawless & Moore, 1989). This division of labor evolves through mutual adjustment of network actors as they come to take for granted that other actors will fulfill certain functions that they would otherwise feel compelled to provide. Again, the argument is that the meaning of sovereignty becomes "contextualized" within the larger open system. If the EU evolved in this direction, we would expect institutional relations to be built around a separation of powers model rather than a delegative principle-agent relationship.

In describing the EU as a network organization, Keohane and Hoffman (1991) emphasized what they call "pooled sovereignty"--essentially sovereignty jointly shared by the network members. This argument still bears the imprint of the intergovernmentalist assumption that all sovereignty is ultimately derived from the state. An open system model of the EU suggests something closer to what has been discussed under the rubric of "multilevel governance" or "neofeudalism." (Ruggie, 1993; Waever, 1995; Marks et al., 1996). In these cases, sovereignty is "distributed" rather than "pooled." "Distributed sovereignty" is closer to what is intended by descriptions of network organization (Powell, 1990; Benson, 1975; Miles & Snow, 1986). Each member of the network acts quasi-independently (e.g., semi-sovereignty) but with the understanding that they are part of a broader enterprise. Each member acts with the knowledge that they have particular rights and skills for acting authoritatively on certain matters. Governance occurs not through "pooling" sovereignty, but rather through the mutual concertation of actors with "distributed" sovereignty. Distributed sovereignty is closer to a polity built around a separation of powers.

The evolution of open systems is likely to depend on the emerging network of relationships between nation-states and international associations. Broadly, we can identify two different paths of evolution that will depend on the degree of membership overlap between international associations--what network theorists call "affiliation networks" (Wasserman and Faust, 1994). At one extreme, the density of associational networks may be low because membership varies greatly across international associations. At the other extreme, density is high because a range of associations have an essentially identical (highly overlapping) membership. As an example of the first case, consider the affiliation network of the US. It is a charter member of OAS, the UN, OECD, NATO, NAFTA, the G7, APEC, and WTO, to name just a few. These associations clearly overlap in membership, but the density is only moderate. It is true that the UN and WTO largely overlap, but OAS, OECD, NATO, and APEC share comparatively few members. At the other extreme, consider the EU. Although it might be counted as a single association, it is really a multipurpose association with densely overlapping membership affiliations (especially given that states can "opt-out" of particular institutional arrangements). And in addition to the overlapping affiliations of the states, there is a corresponding pattern of affiliation among subnational groups: the proliferating associations of "European" trade unions, political parties, regions, environmental groups, and industries. The EU encourages these associational linkages and has consequently become an extremely dense web of institutions and overlapping affiliations.

Both of these patterns might be called "open systems," but they have different implications for the evolving role of state sovereignty. The pattern of low to moderate network density may in fact accentuate the corporate character of states. The state confronts a different group of participants for each association to which it belongs. The relatively specialized character of these associations means that we witness limited issue linkages across associational arenas. Boundary spanners within the member states and in the international associations will be important, but the overlapping relationships between the boundary spanners themselves will be quite limited. Across the range of increasingly numerous associational memberships, the state itself is the only common denominator. The state then is likely to remain the primary common focus of loyalties and commitments.

In contrast, as the degree of membership overlap increases, the likelihood of slipover also increases (Pierson, 1996). Linkages are more easily made across issue areas because the same set of members are engaged in both areas. As these issue linkages increase, so too will the overlaps between the boundary spanners who facilitate extra-territorial coordination. And as the relationships between states becomes more permanent than episodic, boundary spanning roles are likely to be institutionalized. As the activities of the international associations become routinized, the need arises for continuous representation in these arenas. Increasingly, a set of boundary spanners who can tie various levels of governance together is needed.. We might even expect to see something like the French cumul des mandats system emerge, where politicians can simultaneously hold elected offices at various levels--mayor, regional councillor, senator, and member of the European Parliament. Indeed, becoming a member of the European parliament has increasingly been viewed as an attractive career path to domestic political success. As boundary spanners operate across levels, we will expect them to increasingly mix and confuse domestic and international perspectives. The epistemic communities literature argues in effect that the boundary spanners may come to represent these internationalist viewpoints domestically. In the EU, this phenomenon has been most starkly noted in the case of European judges, who have returned to their national courts armed with a legal integrationist perspective (Burley & Mattli, 1993).

A similar phenomenon has been noted with respect to political mobilization around transnational issues. As states become embedded in dense networks of international governance, the agendas of international coalitions will be reflected back into domestic political coalitions. ₁₁ Disputes over transnational issues will polarize domestic political coalitions. Here is where the two-level game analogy breaks down. As applied to the EU by Moravcsik (1993), the two-level game points to the importance of domestic divisions while assuming that national governments are unified. Thus, the EU increases the policy autonomy of national governments by allowing them to play the EU against the divided domestic coalitions. But political agendas may be increasingly worked out by coalitions that straddle the line between domestic and international politics. Certainly, such a dynamic has been apparent in environmental affairs in the last few years (Young, 1994).

This discussion of different types of open systems architecture allows us to partly reformulate the issue of sovereignty. What is really at stake in much of the globalization literature is not whether state sovereignty is eroding, but what we can expect from the nation-state as a political actor. The two-level game argument represents one end of the spectrum, with the nation-state acting corporatively and remaining the critical gatekeeper between domestic and international politics. At the other end of the spectrum, the nation-state becomes simply a decisional arena in which international political coalitions vie for influence over domestic decisions. Here, the state is neither corporate actor nor gate-keeper. Note that sovereignty is not really the issue here. The nation-state may retain full sovereignty and yet have little corporate character. Academic departments in big universities sometimes approach this extreme!

An open system perspective on sovereignty allows for variation along this dimension, but generally expects the future to fall somewhere in the middle rather than at either extreme. The reason is, as we have argued all along, that boundaries are permeable and contingent but not irrelevant. In this middle range, we expect a logic of interaction within sovereign nations and among nations and international associations that we refer to as a "logic of embeddedness." In this scenario, dense institutional networks link domestic and international actors. These networks are themselves the focus of much of the deliberative activities that becomes policy within specific decisionmaking arenas. But these institutional networks are hybrids--composed neither of "agents" of national governments nor of pure cosmopolitans. Rather, the actors within these networks are able to adopt and sometimes meld both perspectives in an on-going process of negotiating the boundaries between states and international associations. Far from seeking to completely undercut state sovereignty, the authority of these brokers is built precisely on the fact that they can successfully operate between boundaries and that the boundaries continue to matter. Thus, they represent the international perspective domestically and the domestic perspective internationally.

References

BENSON, KENNETH J. (1975) The Interorganizational Network as a Political Economy. Administrative Science Quarterly 20: 229-249.

BURLEY, ANNE-MARIE AND WALTER MATTLI (1993) Europe Before the Court: A Political Theory of Legal Integration. International Organization 47 (2): 41-76.

CERNY, PHILLIP G. (1995) Globalization and the Changing Logic of Collective Action. International Organization 49 (4): 595-625.

CONNOLLY, BARBARA (1997) Organizational Choices for International Cooperation: East-West European Cooperation on Regional Environmental Problems. PhD dissertation, University of California, Berkeley.

CROZIER, MICHEL AND ERHARD FRIEDBERG (1980) Actors and Systems: The Politics of Collective Action, translated by Arthur Goldhammer. Chicago: University of Chicago Press.

DIMAGGIO, PAUL J. AND WALTER W. POWELL (eds) (1991) The New Institutionalism in Organizational Analysis. Chicago: University of Chicago Press.

DRAKE, WILLIAM J. (1993) Territoriality and Intangibility. In Beyond National Sovereignty: International Communications in the 1990's, edited by K. Nordenstrang and H. Schiller. Ablex Publishing Company.

EDSTROM, ANDERS, BENGT HOGBERG AND LARS ERIK NORBACK (1984) Alternative Explanations of Interorganizational Cooperation: The Case of Joint Programmes and Joint Ventures in Sweden. Organization Studies 5 (2): 147-168.

FENNIMORE, MARTHA (1996) National Interests in International Society. Ithaca, NY: Cornell University Press.

FUKUYAMA, FRANCIS (1992) The End of History and the Last Man. New York: Free Press.

GRANOVETTER, MARK (1985) Economic Action and Social Structure: The Problem of Embeddedness. American Journal of Sociology 31 (3): 481-510.

GREIDER, WILLIAM (1997) One World Ready or Not: The Manic Logic of Global Capitalism. New York: Simon and Schuster.

HELLEINER, ERIC (1994) States and the Reemergence of Global Finance: From Bretton Woods to the 1990s. Ithaca: Cornell University Press.

KEOHANE, ROBERT O. AND STANLEY HOFFMAN (1991) Institutional Change in Europe in the 1980s. In The New European Community: Decision-making and Institutional Change, edited by Keohane and Hoffman. Boulder, CO: Westview Press.

KEOHANE, ROBERT O. AND JOSEPH NYE (1977) Power and Interdependence: World Politics in Transition. Boston: Little, Brown.

KRASNER, STEPHEN (1983) International Regimes. Ithaca: Cornell University Press.

KRASNER, STEPHEN (1989) Realist Praxis -- Neo-Isolationalism and Structural Change. Journal of International Affairs 43 (1): 143-160.

KUTTNER, ROBERT (1997) Everything for Sale: The Virtues and Limits of Markets. New York: Alfred A. Knopf.

LAWLESS, MICHAEL AND RITA MOORE (1989) Interorganizational Systems in Public Serve Delivery: A New Application of the Dynamic Network Framework. Human Relations 42 (12): 1167-1184.

LAZONICK, WILLIAM (1991) Business Organization and the Myth of the Market Economy. Cambridge: Cambridge University Press.

MARKS, G., L. HOOGHE AND K. BLANK (1996) European Integration from the 1980's" State-centric vs. Multi-level Governance. Journal of Common Market Studies 34 (3): 341-378.

MEYER, JOHN (YEAR) TITLE

MILES, R. AND C. SNOW (1986) Organizations: New Concepts for New Forms. California Management Review 28 (3): 62-73.

MILWARD, A.S. (1992) The European Rescue of the Nation-State. Berkeley: University of California Press.

MORAVSCIK, A. (1993) Preferences and Power in the European Community: A Liberal Intergovernmentalist Approach. Journal of Common Market Studies 31 (4): 473-524.

NOHRIA, NITIN (1992) Information and Search in the Creation of New Business Ventures: The Case of the 128 Venture Group. In Networks and Organizations: Structure, Forum and Action, edited by Nitin Nohria and Robert G. Eccles, pp. 240-161. Boston: Harvard University Press.

NOHRIA, NITIN AND ROBERT G. ECCLES (1992) Networks and Organizations: Structure, Form and Action. Cambridge, MA: Harvard University Press.

O'BRIEN, RICHARD (1992) Global Financial Integration: The End of Geography. New York: Council on Foreign Relations Press.

PERROW, CHARLES (1979) Complex Organizations: A Critical Essay. New York: Random House.

PIERSON, PAUL (1996) The Path to European Integration: A Historical Institutionalist Perspective. Comparative Political Studies 29 (2): 123-163.

POLANYI, KARL (1944) The Great Transformation. Boston: Beacon Press.

POWELL, W.W. (1990) Neither Market Nor Hierarchy: Network Forms of Organization. Research in Organizational Behavior 12: 295-336.

PUTNAM, ROBERT D. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press.

PUTNAM, ROBERT (1988) Diplomacy and Domestic Politics: The Logic of Two-Level Games. International Organization 42 (3): 427-460.

ROSENAU, JAMES (1990) Turbulence in World Politics: A Theory of Change and Continuity. Princeton, NJ: Princeton University Press.

RUGGIE, JOHN (1983) International Regime, Transactions and Change: Embedded Liberalism in the Postwar Economic Order. In International Regimes, edited by Stephen Krasner. pp. 195-232. Ithaca: Cornell University Press.

RUGGIE, JOHN (1993) Territoriality and Beyond: Problematizing Modernity in International Relations. International Organization 47 (2): 139-174.

SCOTT, RICHARD W. (1992) Organizations: Rational, Natural and Open Systems. New York: Prentice-Hall.

SPRUYT, HENDRIK (1994) The Sovereign State and Its Competitors: An Analysis of Systems Change. Princeton, NJ: Princeton University Press.

STEINBRUNER, JOHN D. (1974) The Cybernetic Theory of Decision. Princeton, NJ: Princeton University Press.

STRANG, DAVID (1991) Anomaly and Commonplace in European Political Expansion: Realist and Institutionalist Accounts. International Organization 45 (2): 143-163.

THOMPSON, JAMES D. (1967) Organizations in Action. New York: McGraw-Hill.

THOMPSON, JANICE (1994) Mercenaries, Pirates and Sovereigns: State-Building and Extraterritorial Violence in Early Modern Europe. Princeton: Princeton University Press.

WAEVER, OLE (1995) Identity, Integration and Security: Solving the Sovereignty Puzzle in E.U. Studies. Journal of International Affairs 48 (2): 1-43.

WASSERMAN, STANLEY AND KATHERINE FAUST (1994) Social Network Analysis: Methods and Applications. Cambridge: Cambridge University Press.

WEBER, STEVE (1991) Cooperation and Interdependence. Daedalus 120 (2): 183-201.

WEBER, STEVE (1992) Shaping the Postwar Balance of Power -- Multilateralism in NATO. International Organization 46 (3): 633-680.

WEBER, STEVE AND JOHN ZYSMAN (forthcoming) Why the Changed Relation Between Security and Economy Will Alter The Character of the European Union.

WILLIAMSON, OLIVER E. (1996) The Mechanisms of Governance. Oxford: Oxford University Press.

YARBROUGH, BETH V. AND ROBERT M. YARBROUGH (1992) Cooperation and Governance in international Trade: The Strategic Organizational Approach. Princeton: Princeton University Press.

YOUNG, ORAN (1994) Institutional Bargaining: Creating International Governance Systems. Ithaca, NY: Cornell University Press.

Footnotes

Note 1: Yet, Ruggie does not argue that transactions lead unproblematically to social organization; there is an intermediate step of how people think about transactions -- the development of "social epistemes." Back.

Note 2: For two relatively comprehensive essays on the development of organization theory, see Scott (1992), and Perrow (1979). Scott provides a categorization of schools of organization theory according to the closed-open system dichotomy. Note that there is some debate as to whether Weber is rightfully seen as a closed system theorist (Scott, 1992, p. 102, note 103). Back.

Note 3: Thompson (1967:12, and 4-10) felt that "closed system aspects of organization are most clearly seen at the technical level, and the open systems qualities appear most vividly at the institutional level." Back.

Note 4: See Yarbrough and Yarbrough (1992) and Connolly (forthcoming). Back.

Note 5: An overview of NSI is found in DiMaggio and Powell (1991). Back.

Note 6: Keohane and Nye (1977) set the tone for an explosion of literature about interdependence in the 1980's. Back.

Note 7: The literature on network organizations is starting to explode, but a good starting point is Powell (1990) and the overview volume by Nohria and Eccles (1992). Back.

Note 8: See O'Brien (1992) and Helleiner (1994). Back.

Note 9: See Ruggie (1983). Back.

Note 10: Communist countries have sometimes been exceptions, but the economic collapse of this model -- even in China, where the communist party remains in power -- says something about its unworkability. Back.

Note 11: See Weber and Zysman, forthcoming. Back.