Do we need an Arctic Ocean framework agreement?

In international society, much like in other social settings, framework agreements are constitutive arrangements that focus on the creation of governance systems defined in functional terms, spatial terms, or both. These arrangements feature the formulation of broad principles, often framed as rights and duties, and the establishment of bodies endowed with the authority to apply such broad principles to concrete situations. For the most part, framework agreements are not regulatory in character, though they are often intended to provide an institutional platform on which to build more substantive regulatory systems. Prominent examples of framework agreements operative in international society today include the 1982 UN Convention on the Law of the Sea (UNCLOS), the 1985 Vienna Convention for the Protection of the Ozone Layer, the 1992 UN Framework Convention on Climate Change (UNFCCC), and the 1992 Convention on Biological Diversity (CBD). More substantive regulatory arrangements built on these platforms commonly take the form of supplemental measures (for example the 1991 ban on driftnet fishing in international waters adopted in the form of a UN General Assembly resolution in the case of UNCLOS) or protocols to the original convention (for example the 1987 Montreal Protocol on Substances that Deplete the Ozone Layer). Such regulatory arrangements vary substantially in terms of their effectiveness. Consider the contrast between the Montreal protocol and the 1997 Kyoto protocol to the UNFCCC in these terms. But in each case, it is helpful to draw a distinction between the framework agreement itself and various regulatory supplements in efforts to understand the structure and functioning of the governance systems in question.

Because UNCLOS is at the centre of the current debate about addressing needs for governance in a rapidly changing Arctic, it is worth delving a little more deeply into the character of this convention as a framework agreement (United Nations 1983). UNCLOS formalises numerous concepts relating to ocean governance and lays down principles relating to human uses of marine resources. It recognises coastal state jurisdiction in exclusive economic zones (EEZs), codifies the right of innocent passage even within EEZs, calls for peaceful uses of the high seas, creates what is known as the ‘Area’ covering the seabed beyond coastal state jurisdiction, and so forth. In all, UNCLOS runs to 320 articles grouped into 17 parts that cover issues ranging from traditional concerns such as the limits of the territorial sea, the right of innocent passage in areas under coastal state jurisdiction, and freedom of the high seas to emerging issues including deep seabed mining and the protection of the marine environment.

Despite its length, UNCLOS does not deal with a wide range of issues in sufficient depth to resolve disagreements or disputes relating to its application to specific situations. To illustrate, the convention does not spell out detailed rules regarding the delimitation of boundaries between adjacent or opposite states, the regulation of marine fishing, the scope of peaceful uses of the high seas, the operational meaning of the phrase ‘straits used for international navigation,’ or the precise criteria to be used in evaluating the claims of coastal states to jurisdiction over the continental shelves beyond the limits of their EEZs. Instead, UNCLOS establishes a number of bodies that are intended to play a role in resolving issues involving applications of the principles of the convention to concrete situations. These include the International Tribunal on the Law of the Sea (ITLOS), the International Seabed Authority (ISA), and the Commission on the Limits of the Continental Shelf (CLCS). Because UNCLOS operates under the auspices of the United Nations system, moreover, it is possible to make use of more general UN procedures in addressing issues relating to the application of the principles of UNCLOS to specific situations. This includes the reference of specific disputes to the International court of justice (ICJ) as illustrated by the role of the court in resolving the dispute between Canada and the United States regarding jurisdiction over the waters of Georges Bank. Like most framework agreements, then, UNCLOS seeks to provide mechanisms that interested parties can use to bridge the gap between broad principles and the concrete circumstances of specific situations.

Those who argue that there is a need to negotiate a distinct Arctic Ocean framework agreement subscribe, implicitly if not explicitly, to the following propositions: 1) UNCLOS is deficient as a framework agreement covering the Arctic Ocean, 2) it is reasonable to expect some appropriate set of actors to reach agreement on the terms of an Arctic Ocean framework agreement and 3) the parties to such an agreement will exercise the political will needed to make use of such an agreement to address a range of emerging Arctic issues. One argument regarding the deficiency of UNCLOS can be disposed of at the start. The United States has not become a party to UNCLOS and therefore is unable to make use of its procedures regarding matters like the delimitation of its jurisdiction over the continental shelf beyond the limits of its EEZ. In practice, however, the United States accepts most of the provisions of UNCLOS as a matter of customary law (with the exception of the system governing deep seabed mining in part XI). There is no reason to expect that the United States would ratify an Arctic Ocean framework convention as an alternative to UNCLOS when it comes to dealing with Arctic issues. In fact, many observers believe that pressure to address issues now emerging in the Arctic could mobilise the political pressure needed to persuade the US Senate to ratify UNCLOS.

Is UNCLOS deficient as a framework agreement within which to address emerging Arctic Ocean issues, as some interested parties (for example the European Parliament, a variety of environmental NGOs) assert? Or is it perfectly adequate to provide a basis for addressing these issues, as other interested parties (for example the five Arctic coastal states) proclaim? There is no simple answer to these questions. The general principles articulated in UNCLOS are broad and fully applicable to the Arctic Ocean. The process of delimiting jurisdictional boundaries on the continental shelf beyond the limits of the EEZs of the Arctic coastal states, for example, is proceeding in conformance with the procedures established in article 76 of UNCLOS. The convention also contains a special provision, article 234, on ice-covered areas designed to allow coastal states ‘. . . to adopt and enforce non-discriminatory laws and regulations for the prevention, reduction and control of marine pollution from vessels in ice-covered waters'. Drafted largely in the 1970s, UNCLOS does not emphasise some of the language that has become familiar in recent years regarding such things as large marine ecosystems, ecosystem-based management (EBM), and the importance of paying attention to traditional or indigenous ecological knowledge. What we can conclude, in this context, is that the provisions of UNCLOS are broad enough to allow for a variety of specific regulatory arrangements aimed at addressing the conditions obtaining in the Arctic. On the other hand, it is possible to make the case that an Arctic Ocean framework agreement would draw the world's attention to the transformative changes occurring in the circumpolar Arctic and, in so doing, sharpen the case for adopting a precautionary approach to human activities in the region.

The situation becomes more difficult when we turn to the question of identifying an appropriate set of actors to negotiate the terms of an Arctic Ocean framework agreement. The problem here is easy to identify but hard to solve. Because it would encompass areas and activities beyond the jurisdiction of the Arctic coastal states, the process of developing such an agreement would have to be open to participation on the part of non-Arctic states. This would be good news from the perspective of non-Arctic states like China, Japan, and Korea as well as intergovernmental bodies like the European Union. But it goes far toward explaining the clear and frequently stated opposition of the five Arctic coastal states to any move in this direction. What is more, this matter of participation raises difficult questions about the role of members of the Arctic Council that are not Arctic coastal states (Finland, Iceland, and Sweden) and about the participation of non-state actors like the indigenous peoples' organisations that have the status of permanent participants in the Arctic Council. Without sorting out these tricky questions relating to participation, any effort to negotiate the terms of an Arctic Ocean framework agreement will be stillborn.

In my judgment, an effort to move forward on this track under the conditions prevailing at this time would become a dead end and might well divert energy from other more promising initiatives. UNCLOS, like any constitutive agreement, can be difficult to apply to newly emerging issues. Yet the convention is broad enough to provide a serviceable framework for addressing most issues relating to the Arctic Ocean. Efforts to hammer out the terms of an Arctic Ocean framework agreement are not only likely to prove frustrating; they are also likely to divert attention from efforts to address a number of more focused issues (discussed in the next section) that may well lend themselves to issue specific agreements in the near future. This is not to say that we should abandon all hope of devising an Arctic Ocean framework agreement in the longer run. For the moment, however, pursuing this goal is not only politically irrelevant, it may also be counterproductive in the sense that it gets in the way of efforts to hammer out the terms of focused agreements dealing with issues like search and rescue and the protection of the Arctic marine environment from oil spills.

What is the best target to aim for on the integration-fragmentation spectrum?

Those who think about governance systems designed to address complex issue areas or to cover multiple issues in spatially delimited areas have developed the idea of an integration-fragmentation spectrum along with the concept of governance or regime complexes to characterise arrangements located between the two extremes on this spectrum (Raustiala and Victor Reference Raustiala and Victor2004; Keohane and Victor in press). Fully integrated systems are those in which all the issues relating to a functionally or spatially defined area are embedded in a common arrangement and linked together via well defined connections. A fully fragmented arrangement, on the other hand, is one in which every issue is treated separately and there are no explicit connections between or among the elemental regimes dealing with individual issues. Between these extremes lies a range of alternatives in which two or more overlapping but non-hierarchical regimes are connected either through well defined and intentional links or through the development of informal practices that allow for mutual adjustment in the interests of avoiding or minimising interference among the individual elements of the complex. The result is an interlocking structure or network of regimes operating in a broad issue area (Underdal and Young Reference Underdal, Young, Underdal and Young2004).

It is worth noting at the outset that even governance systems admired by many participants in the debate about Arctic governance are actually institutional complexes that are not fully integrated. Perhaps the leading case in point is the Antarctic Treaty system (Berkman and others in press). This frequently praised system encompasses three separate conventions (the Antarctic Treaty itself together with the Convention for the Conservation of Antarctic Seals and the Convention on the Conservation of Antarctic Marine Living Resources) that are not identical with regard to membership, spatial scope, or substantive coverage. Added to this matrix of conventions are a number of closely related yet distinct bodies, including the Scientific Committee on Antarctic Research (SCAR) which is a standing committee of the International Council of Science (ICSU), the Council of Managers of National Antarctic Programmes (COMNAP) which is an association of national managers responsible for Antarctic logistics, and the International Association of Antarctica Tour Operators (IAATO) which is a non-state actor concerned with ship based tourism. What is more, the location of this system on the integration-fragmentation spectrum has shifted over time. If the 1988 Convention on the Regulation of Antarctic Mineral Resource Activities (CRAMRA) had entered into force, the system would have shifted toward the pole of fragmentation. With the negotiation of the 1991 environmental protocol (an extension of the Antarctic Treaty itself) as an alternative to CRAMRA, on the other hand, the system shifted somewhat toward the pole of integration. Most observers are inclined to see the sidetracking of CRAMRA and the negotiation of the environmental protocol as a progressive development in terms of the effectiveness of the Antarctic Treaty system (Joyner Reference Joyner1998). Nonetheless, it is worth noting that CRAMRA contained a number of progressive elements including arrangements that would have allowed for (weighted) majority voting on a range of issues and for the development of a revenue stream independent of the contributions of member states. As the Antarctic example demonstrates, situations of this sort are dynamic. Governance systems can move along the integration-fragmentation spectrum over time as a result of either conscious decisions or of informal practices. And there is no basis for concluding that any particular locus on this spectrum is optimal in some general sense.

Turning to the situation now emerging with regard to the Arctic Ocean, we can identify a number of developments that raise questions of governance and that may become ripe for the development of various sorts of agreements during the foreseeable future. Among the issues that come into focus in this context are: search and rescue, the safety of life at sea, ship design and navigation, oil and gas development, industrial fishing, ship based tourism, the harvesting of marine mammals, environmental protection, ecosystem-based management, and arms control. A little reflection will make it clear that these issues vary in terms of their ripeness for international agreement, the roster of relevant actors, the spatial extent of the relevant problem, the auspices under which negotiations can go forward, and the nature of the arrangements likely to be adopted. A few examples should suffice to indicate the significance of these variations.

Whereas the key players are close to agreement on the provisions of a search and rescue (SAR) regime developed under the auspices of the Arctic Council and based on cooperation among the Arctic coastal states, the expectation now is that the development of a polar code covering the design and operation of commercial ships in Arctic waters will take the form of a mandatory regulatory system operating under the auspices of the International Maritime Organization (IMO). Any agreement on oil and gas development almost certainly will take the form of a set of guidelines (in contrast to mandatory regulations) agreed to among the five Arctic coastal states because such activities will occur during the foreseeable future in areas subject to the jurisdiction of these states and because they all have national regimes covering offshore hydrocarbon development. High seas fishing taking place in a portion of the Arctic is already covered under the provisions of the Northeast Atlantic Fisheries Convention. More focused arrangements cover North Atlantic salmon and pollock in the central Bering Sea. A realistic expectation is that interest in fishing in other parts of the Arctic will give rise to one or more additional regional fisheries management organizations (RFMOs). The harvesting and the conservation of marine mammals are covered by a number of arrangements, including the 1946 International Convention on the Regulation of Whaling, the 1973 International Agreement on the Conservation of Polar Bears and Their Habitat, the 1992 agreement establishing the North Atlantic Marine Mammal Commission, and several bilateral co-management arrangements in the North American Arctic. The rapid rise of ship based tourism in the Arctic has prompted the creation of a northern counterpart to IAATO known as the Association of Arctic Expedition Cruise Operators (AECO).

Whether it would be desirable to create an intergovernmental agreement to deal with tourism is an open question. Even arms control, which many regard as a non-starter in the Arctic in contrast to the Antarctic, may become an attractive option for the Arctic states in the near future. In this connection, it is worth paying attention to prime minister Vladimir Putin's statement in September 2010 that ‘. . . we think that preserving the Arctic as a zone of peace and cooperation is of the utmost importance' (Putin Reference Putin2010).

What can we conclude about the adequacy of this rather complex mosaic of arrangements and about the future of governance in the Arctic? Those who think in terms of governance or regime complexes point out that arrangements of this sort frequently prove advantageous with regard to the twin issues of flexibility across issues and adaptability over time (Keohane and Victor in press). Such a complex, in contrast to a fully integrated governance system, is able to establish distinct regulatory provisions and decision making procedures that are well suited to the character of separable issues. It is also able to adapt on a piecemeal basis, adjusting the provisions of individual regimes or elements to handle changing circumstances relating to specific issues without raising questions about the viability of a larger and more encompassing governance system. It seems clear that these general features of regime complexes are pertinent to the Arctic, at least at this stage in which the governance agenda features an array of issues that are strikingly diverse and the willingness of key players to move forward in addressing specific issues is by no means uniform.

Still, this is not the end of the story with respect to integration and fragmentation. While we have no means of quantifying such matters, it is undoubtedly the case that the Arctic governance system at present is located relatively far toward the fragmentation end of the spectrum. Yet new activities already emerging in the Arctic or expected to develop as the region becomes more accessible will heighten the links between and among a range of distinct human activities taking place in the circumpolar Arctic as well as between Arctic occurrences and developments unfolding at the global level. Nowhere is this more evident than in the realm of environmental concerns. A major oil spill in ice-covered or ice-infested waters could produce profound ecological consequences lasting over an indefinite period of time. The Arctic impacts of global scale climate change constitute one of the critical forces driving the transformation of the Arctic occurring today. These observations suggest two additional points relating to this account of integration and fragmentation with regard to Arctic governance. There is much to be said for highlighting the importance of the ideas associated with EBM as we endeavour to meet needs for governance in the Arctic (McLeod and Leslie Reference McLeod and Leslie2009). Whether this goal is best addressed by seeking to devise a set of region wide principles of EBM for the Arctic or built into a variety of specific regimes in more operational terms remains to be seen. But the virtues of this way of thinking are easy to see with regard to the management of human activities in the circumpolar Arctic. Arguably more challenging is the issue of governing global activities now affecting the Arctic in dramatic ways, whether this is a matter of the release of methane from thawing permafrost, the acidification of cold Arctic waters, or the diffusion of persistent organic pollutants. The challenge here arises from the facts that the problems are caused by human actions occurring far beyond the boundaries of the Arctic and that the voice of the Arctic is often weak or indiscernible in decision making processes relating to such matters.

Clearly, Arctic governance will feature a complex of distinct arrangements during the foreseeable future. But this does not mean that there is no scope for movement along the integration-fragmentation spectrum at this stage. In my judgment, we should not allow ourselves to be distracted by the vision of a fully integrated system, whatever the attractions of such an arrangement may be in principle. Rather, we should think in terms of a step-by-step process, welcoming the development of strengthened governance systems dealing with specific issues (for example a SAR agreement, a mandatory polar code) and remaining alert for opportunities to identify interactions between these issue-specific arrangements and to build linkages between or among them as appropriate. The result during the foreseeable future will be somewhat messy and may seem unappealing to some. Its virtue, on the other hand, will be an ability to make significant progress in governing a range of human-environment interactions under real world conditions.

Would it help to make (some of) these arrangements legally binding?

Those who advocate a move toward legally binding agreements for the Arctic are generally thinking in terms of hard law arrangements. That is, they would like to see the development of one or more treaties relating to the Arctic Ocean that are ratified by the member states and (preferably) incorporated into the domestic laws of each of the members through some appropriate procedure (for example implementing legislation). But it is important to introduce several clarifying observations at the outset. Hard law in this sense anchors one end of a spectrum that ranges from purely political statements that do not have the force of law (for example the 2009 Copenhagen accord on climate change) through various forms of soft law (for example customary practices, codes of conduct, ministerial declarations, executive agreements, and so forth) and onward to formal conventions and treaties. The choice we face in this realm with regard to Arctic Ocean governance, therefore, is not a simple dichotomous one. There is also the prospect of variance with regard to legal status across the individual elements of a governance or regime complex. We may go forward in the Arctic with an array of arrangements including a hard law polar code dealing with commercial shipping, a set of more informal guidelines covering oil and gas development, a code of conduct regarding the activities of military aircraft and ships, and so forth. What is the most unlikely scenario for the foreseeable future is a single integrated governance system for the Arctic articulated in the form of a hard law treaty or convention ratified by all the signatories. But surely we should not allow this to become a deterrent to efforts to address specific needs for governance in a rapidly changing Arctic.

Why is the status of hard law so appealing to many of those who think about emerging needs for governance in an area like the Arctic? It is possible to address this question at two distinct levels, one featuring the appeal of the idea of the rule of law as an overarching approach to governance and another emphasising the attractions of hard law when it comes to eliciting compliance with more specific regulatory arrangements. Many analysts and especially those who are imbued with western thinking about governance in domestic settings take it almost as an article of faith that the rule of law is an essential, arguably necessary, condition for producing what is commonly thought of as good governance (Zaelke and others Reference Zaelke, Stilwell, Young, Zaelke, Kaniaru and Kruzikova2005). Where the rule of law prevails, on this account, good governance is almost sure to follow. Conversely, where good governance occurs, it is to be expected that the rule of law will prevail. Those who subscribe to some version of this paradigm are apt to regard it as axiomatic that we should be aiming to establish legally binding arrangements to meet the emerging needs for governance in the Arctic. Anything short of legally binding arrangements will leave us with a governance system that is not rooted in some overarching conception of what it takes to address needs for governance effectively.

Moving from this philosophical perspective to the nitty-gritty of solving specific governance problems, many commentators express a preference for hard law arrangements on the grounds that subjects are more likely to comply with the requirements of hard law than with the requirements of various forms of soft law, much less mere social practices. The logic behind this proposition is simple, although not entirely persuasive. Many analysts associate hard law with the availability of enforcement mechanisms capable of applying sanctions needed to elicit compliance on the part of subjects. More generally, there is a common view that hard law exercises a normative pull when it comes to compliance. Actors treating hard law requirements as authoritative are likely to comply with them simply because it is the law, and this is especially true in cases where the provisions of legally binding international agreements are absorbed into the domestic legal systems of member states. The significance of this line of reasoning looms large in a setting like international society where the role of law is not backed by the sorts of enforcement mechanisms that we take for granted in a well ordered domestic society.

Does this mean that we should push at all costs for legally binding agreements in efforts to address emerging needs for governance in the Arctic? The answer to this question is sobering. There are good reasons to proceed with considerable care with regard to the legal status of the governance arrangements we create to address Arctic issues. To begin with, the most extensive and rigorous analyses of the effectiveness of international (environmental) agreements raise serious questions about the idea of a normative pull, at least in this setting (Miles and others Reference Miles, Underdal, Andresen, Wettestad, Skjaerseth and Carlin2002; Breitmeier and others Reference Breitmeier, Young and Zürn2006; Breitmeier and others in press). The evidence from these studies suggests that the link between the legal status of governance systems and levels of compliance on the part of their members is weak. On the other hand, a number of studies have shown that indirect mechanisms can play an important role in eliciting compliance, whether or not agreements are legally binding (Dai Reference Dai2007). This is not to say that making international agreements legally binding is irrelevant with regard to effectiveness. But it is clear that reliance on legally binding agreements is neither necessary nor sufficient to achieve success in the development of international governance systems (Lipson Reference Lipson1991).

Given this conclusion, it is important to consider some significant drawbacks that often arise in efforts to negotiate and implement legally binding agreements. When the parties are aiming to reach agreement on the terms of legally binding instruments, the negotiation process is often protracted. Consider the case of efforts to add legally binding commitments to the UNFCCC as a dramatic case in point. Breaking through the logjam is apt to take the form of eliminating or watering down the substantive provisions included in legally binding agreements. This leads to a well known tradeoff between legal status and substantive content. Although it is not inevitable, legally binding agreements are apt to be shallow with regard to content (Barrett Reference Barrett2003; Barrett and Stavins Reference Barrett and Stavins2003). Legally binding agreements do not enter into force and take on the status of hard law until they are ratified by the member states. While some agreements (for example UNCLOS, the Kyoto protocol) can enter into force based on a formula that does not require ratification on the part of all the signatories, the example of the 1959 Antarctic Treaty suggests that entry into force for any broad Arctic Treaty would require ratification by all the signatories or at least by all the Arctic states. This can be a major hurdle. UNCLOS entered into force in 1994, twenty years after the start of negotiations in 1973–1974. The United States still has not ratified UNCLOS, though successive Democratic and Republican administrations have called for ratification, and the US treats most of the convention's substantive provisions as matters of customary law. Even the more focused Kyoto protocol to the UNFCCC took over seven years to enter into force, and the United States is not among the group of states that are formal members of this regime.

One observation that flows from this discussion is that the composition of the membership of any Arctic agreements that are legally binding will be a critical issue. But this raises another set of questions. While it is possible to imagine the five Arctic coastal states agreeing on guidelines relating to offshore oil and gas development or entering into a more formal SAR agreement, it makes no sense to develop a mandatory polar code dealing with commercial shipping in the Arctic without including both the remaining Arctic states and a number of non-Arctic countries (for example China, Germany, Japan, Korea, the Netherlands) that will almost certainly be important players in the development of commercial shipping in the Arctic. Similar remarks are in order regarding marine pollution and industrial fishing, at least in those sectors of the Arctic Ocean that remain high seas. And all this begs the question of how to include key non-state actors, such as the indigenous peoples' organizations that enjoy the status of permanent participants in the Arctic Council, subnational bodies like the Northern Forum and the Barents regional council, and leading environmental NGOs like WWF. My prediction is that it would prove impossible to reach clear agreement on the composition of the membership for any legally binding agreement intended to address a sizable suite of governance needs arising in the Arctic at this time.

Legally binding agreements are often difficult to adjust or to adapt to changing circumstances. As the case of the Antarctic Treaty suggests, this may lead to protracted debates over the addition of new protocols to the original treaty (for example the 1991 environmental protocol) or to the negotiation of separate instruments that do not take the form of elements of the original treaty (for example the conventions on seals and marine living resources). This is not to say that legally binding agreements are never able to adapt to changing circumstances. The case of stratospheric ozone depletion in which the parties reached agreement on the terms of the regulatory provisions of the Montreal protocol just two years after the framework agreement and the meetings of the parties have succeeded in amending specific provisions of the Montreal protocol on a number of occasions provides a particularly striking example in these terms. Nonetheless, legally binding agreements are often difficult to adjust appropriately and efficiently to changing circumstances. It is hard to imagine the creation of an Arctic treaty that parallels the Montreal protocol in allowing the meeting of the parties/conference of the parties to adopt substantive amendments that can enter into force without requiring ratification or some form of formal approval on the part of the individual members.

If we assume that the emerging system of governance for the Arctic will fall somewhere in the middle of the integration-fragmentation spectrum, a relatively clear way forward with regard to the issue of legally binding agreements comes into focus. The individual elements of a governance or regime complex can and often do vary with regard to their legal status. Individual elements frequently vary in terms of membership, and the members of each element can arrive at their own conclusions when it comes to deciding on the legal status of individual elements. The proposed polar code governing commercial shipping may well take the form of a legally binding agreement that becomes a member of the IMO family of conventions. Some jurisdictional issues may be resolved through the negotiation of legally binding agreements involving only two or three states. The September 2010 Treaty between Norway and the Russian Federation concerning Maritime Delimitation and Cooperation in the Barents Sea and Arctic Ocean is a case in point (available in the Arctic Governance Compendium at www.arcticgovernance.org). There are some indications that Canada and the United States may be heading in the same direction regarding the issue of jurisdictional boundaries in the Beaufort Sea. In other areas, a variety of soft law agreements are apt to be the order of the day. The eight Arctic states are likely to address various reforms of the Arctic Council through the medium of ministerial declarations. The five Arctic coastal states may well resort to executive agreements as they endeavour to sort out issues pertaining to jurisdiction over continental shelves beyond the limits of their EEZs. Relevant parties may address specific fisheries issues through the establishment and adaptation of relatively informal agreements, as Norway and Russia have done since the 1970s with regard to the fisheries of the Barents Sea. AECO, a nongovernmental organization similar to IAATO in the Antarctic, is emerging to address regulatory issues pertaining to ship based tourism in the Arctic. There also may be insights relevant to this domain to be drawn from the experience of civil aviation in which an intergovernmental body (the International Civil Aviation Organization or ICAO) shares regulatory responsibility with a nongovernmental body (the International Air Transport Association or IATA). And there is nothing to prevent these arrangements from adapting and adjusting on a case-by-case basis over time with regard to the issue of the legally binding character of the agreements on which they rest.

How should we proceed?

Does this discussion yield any take home messages regarding next steps in the realm of Arctic Ocean governance? My primary concern in the preceding sections has been to deconstruct the argument underlying proposals for an Arctic Ocean treaty rather than to arrive at policy recommendations regarding the future of Arctic governance. Nevertheless, the discussion does yield some conclusions that are of interest as we endeavour to go forward in meeting needs for governance in the Arctic. Four observations seem particularly apposite at this juncture.

It does not make sense, at least at this stage, to invest time and energy in an effort to devise a framework agreement for the Arctic Ocean. The marginal benefit of adding such an agreement to the existing UNCLOS framework would be limited. It seems unlikely that the main players would be able to reach consensus regarding the issue of membership in such an agreement. The Arctic coastal states are united not only in expressing a lack of interest in such an arrangement but also in actively opposing an initiative of this sort. If such an agreement were to take the form of a legally binding instrument, it might not enter into force anytime soon. The effort to make progress toward the adoption of an Arctic Ocean framework agreement would soak up a lot of time and energy and might well detract from the capacity of policymakers to address more specific needs for governance in the Arctic in a constructive and productive manner.

What is now emerging in the Arctic, and is likely to develop further in the coming years, is a governance or regime complex encompassing a number of distinct yet interrelated elements. Progress regarding some of the individual elements of this complex is already apparent. Efforts to reach agreement on a polar code under the auspices of the IMO and a SAR arrangement under the auspices of the Arctic Council are both moving forward and likely to bear fruit in the near future. Other elements of this package, including oil and gas guidelines and arrangements for industrial fishing, will evolve on their own timelines. What we can expect, in this realm, is the growth of a governance complex encompassing a number of elements that differ significantly in terms of their legal character, their functional and spatial domains, and the auspices under which they operate. A major challenge in this regard will be to pay attention to the locus of this evolving complex on the integration-fragmentation spectrum and to encourage movement toward the pole of integration, without sacrificing the capacity of the individual elements to adapt effectively and efficiently to changing circumstances regarding the activities they govern.

The biggest challenge regarding Arctic Ocean governance during the coming years is likely to be to find ways to bring the idea of EBM to bear in the Arctic and, at the same time, to protect and enhance the vitality of the Arctic's indigenous cultures. As experience with the management of marine living resources in the Antarctic makes clear, it is one thing to espouse the idea of EBM in principle; it is another to operationalise the idea and to move it from paper to practice under real world conditions. In the Arctic, the maintenance of biologically diversity (the foundation for EBM) and the protection of cultural integrity are closely connected (Peterson and Johnson Reference Peterson and Johnson1995; Petersen and Poppel Reference Petersen and Poppel1999). A failure to maintain biological diversity and to adopt the ecosystem perspective needed to do so would severely jeopardise the subsistence oriented cultures of the Arctic. Conversely, the traditional ecological knowledge (TEK) of indigenous peoples can play a critical role in helping to understand the requirements of EBM in the Arctic (Berkes Reference Berkes2008). We need to join together western scientific knowledge and TEK to gain the depth of understanding needed to manage human/environment interactions successfully. There is no optimal mechanism for addressing this cluster of issues with particular reference to the Arctic Ocean. But in my judgment, the place to begin is with the idea of stewardship at the conceptual level (Chapin and others Reference Chapin, III, Kofinas and Folke2009) and with the work of the Arctic Council and especially its most relevant working groups at the applied level. The Arctic Ocean review, now underway under the auspices of the Working Group on the Protection of the Arctic Marine Environment, is a constructive example of this approach. It would be even better if this activity were carried out under the auspices of the council's Working Groups on Sustainable Development and on Emergency Prevention, Preparedness, and Response.

Finally, a fundamental challenge for those pursuing stewardship in the Arctic arises from the fact that major forces affecting the health of Arctic ecosystems and the integrity of Arctic cultures (for example climate change, threats to biological diversity, the diffusion of persistent organic pollutants) are driven by human actions taking place on a global scale and governed by global regimes in which the voice of the Arctic is apt to be faint. What this means is that the Arctic needs a champion able to represent the concerns of the region effectively in global governance venues. At this stage, a suitably reformed Arctic Council is likely to be the best mechanism available to perform this role.