MANAGING FRAGMENTATION BETWEEN INTERNATIONAL TRADE AND INVESTMENT LAW AND GLOBAL PRIORITIES FOR NONCOMMUNICABLE DISEASE PREVENTION IN FOOD AND ALCOHOL

This article examines three ways in which instruments developed by international health organizations might be used to manage fragmentation between health and international trade and investment law in the context of regulating food and alcohol as noncommunicable disease risk factors. These are: by modifying or overriding trade and investment obligations, affecting interpretation and fact finding, and establishing cooperative inter-institutional processes. The article assesses the advantages and disadvantages of each of these approaches for managing fragmentation under the rules of treaty interpretation and dispute settlement. It argues that while much of the discussion of fragmentation between trade and health focuses on treaty interactions, many of the actual uses of health instruments are not necessarily dependent on their formal legal status. It then proposes several features of international instruments that might strengthen or support these uses, whether through binding or through non-binding instruments.


INTRODUCTION
The prospect of 'fragmentation' or 'policy incoherence' between trade and investment law and health objectives has been an acute concern in noncommunicable diseases ('NCD') prevention over the last decade. Much of this has been driven by high profile disputes about tobacco control in the World Trade Organization ('WTO') and under the investor-state dispute settlement mechanisms of bilateral investment treaties, including investor-state disputes brought by Philip Morris companies against Uruguay and Australia's tobacco packaging and labelling laws, and a series of cases brought by four WTO member states against Australia's plain packaging laws. 1 Although all three cases have since been resolved (or reportedly 2. Interpretative uses: use of health instruments in interpretation, fact-finding and benchmarking by trade and investment adjudicators; 6 3. Procedural approaches: use of inter-institutional processes to sensitise trade and investment adjudicators to health concerns, as well as encourage more dialogue between different actors within states. 7 There are a number of detailed existing studies examining fragmentation and coherence between health, trade and investment instruments in terms of the law governing conflicts between treaties. 8 However, the ways in which the WHO FCTC and other non-trade/investment treaties have influenced the deliberations of trade and investment tribunals are much broader than formal legal interactions between treaties. Many of the ways in which health instruments have been used as sources of factual, normative, and institutional authority are formally indifferent to the legal status of the instrument being used, and could also be relevant to nonbinding instruments. Nevertheless, binding multilateral treaties embody other qualities that have also proven to be important to the resolution of trade and investment disputes, including political visibility, institutional support, and international consensus. It is worth considering to what extent these features can be replicated or more strongly emphasised for food and alcohol governance, whether through legally binding instruments or otherwise.
This article assumes that for the near future there will be no relevant amendments of the WTO agreements, and that a significant number of states will still be bound by at least one investment treaty without explicit protection of public health regulation, although it should be noted that several states are in the process of renegotiating investment agreements to provide such protection. 9 As such, its scope is limited to the evolution of trade and investment regimes through case law rather than renegotiation of treaties. 10 It also does not consider the role of health instruments in any respect other than managing the risk of trade and investment disputes. Broader debates about the merits of binding versus non-binding instruments in food and alcohol are, of course, about significantly more than this, and this article does not address other reasons that health bodies might want to adopt new binding instruments. potentially uncoordinated and conflicting subfields, the phenomenon was in fact a sign of international law's vitality. Fragmentation was caused in part by a great increase in the number of areas of international cooperation and the greater use of legal instruments to achieve the goals of such cooperation. The resulting study by the Commission was subtitled 'Difficulties Arising from the Diversification and Expansion of International Law' to reflect both the challenges but also the opportunities presented by an increasingly diverse and complex international legal universe. 11 Among the challenges of fragmentation identified by the Commission were the possibility of conflicting interpretations between different specialised areas of international law, as well as the possibility of states picking and choosing between substantive interpretations and institutional mandates at the risk of coherence to the whole. As international law expanded, governing the relations between its different branches thus became more complex, although the Commission emphasised that the law of treaties and dispute settlement provided many tools for international lawyers to address this complexity. 12 Examples of overlaps between norms and techniques to resolve them were widespread across many subject areas of international law, with the Commission citing examples from interactions between environmental law and trade law; international human rights and humanitarian law; regional and bilateral environmental agreements; and different tribunals interpreting the same rules. 13 This diversification and expansion has also occurred in global health, which has in recent years transformed from a discipline 'primarily focused on technical, medical and professional problems and solutions' to one that also focuses on social, political, and commercial determinants of health. 14  but about the relative power and authority of bodies with different mandates. 29 As such, one of the background assumptions for this article is that 'policy coherence' is largely about whether or not sufficient consideration is given to health in trade and investment adjudication, because there is a significant disparity in enforceability between the two sets of commitments, and thus the legitimacy concerns about their interaction have largely centred around the adjudication of health measures by trade and investment bodies rather than the reverse.
Another background assumption in this article is that fragmentation is a question of politics as much as lawit is a by-product of states (and different organs of states) seeking to advance different interests in different fora. The multiplicity of fora addressing a specific issue can arise both from lack of coordination and from deliberate attempts by states to 'forum shift', including attempts to advance different interests by strategically creating conflicting or overlapping obligations between different regimes. 30 Addressing fragmentation therefore requires normative and political contestation, not just regulation and coordination, and a certain amount of fragmentation is inevitable as long as states continue to have differing political preferences. 31 Nevertheless, how fragmentation should be addressed involves a number of technical considerations, and it is these considerations that are the subject of this article. In the context of tobacco, this has largely focused on the interactions between the WHO FCTC and international trade and investment law under the law of treaties and dispute settlement. 32 Unlike tobacco however, there are no central treaties with detailed policy implementation obligations for food and alcohol. Instead, food and alcohol are governed at the international level primarily by non-binding instruments and by general obligations to protect and promote health in human rights law. Such instruments include:  Human rights treaties, which contain binding obligations to promote health, but are generally not prescriptive as to how those obligations should be met. Under the International Covenant on Economic, Social and Cultural Rights, for example, states have a general obligation to prevent and control disease, which includes the underlying determinants of health (such as adequate nutrition), control of substances harmful to health (such as tobacco or alcohol), and access to health information (such as dietary information or consumer warnings). 33 States also have an obligation to ensure freedom from malnutrition, which encompasses access to nutritionally adequate and safe foods. 34  'combat disease and malnutrition, including within the framework of primary health care, through, inter alia, … the provision of adequate nutritious foods'. 35 General Comment 15 by the Committee on the Rights of the Child defines malnutrition to include both undernutrition and obesity, and recommends that states implement certain WHO instruments to address childhood obesity. 36 Although the status of such General Comments in the interpretation of obligations remains controversial, 37 it has been recognised by the International Court of Justice that they should be given significant weight. 38 As such, it can be considered that the right to health supports states taking measures to address NCD risk factors, including alcohol and food, and in particular that the Convention on the Rights of the Child supports action on childhood obesity. However, most human rights instruments are not prescriptive as to the specific regulatory measures states should take, in line with their nature as broad, universally applicable instruments.  Normative non-binding instruments, which aim to provide an agreed agenda for action or a set of recommendations to member states in a particular subject area. Such instruments include those adopted by states through the World Health Assembly ( well as technical packages, such as the SHAKE package for salt reduction. 46 These may make recommendations but are primarily intended to build capacity or serve as a resource rather than as commitments to certain courses of action.
As such, this article looks not only at treaty interactions, but also the role of other international health instruments in the resolution of trade and investment disputes. It considers in particular how much we can apply lessons from the WHO FCTC, a binding treaty with 181 parties, to the very different normative frameworks that govern food and alcohol policy.
Of course, tobacco is significantly different from food and alcohol in other respects, in that it is a uniquely and fairly homogenously dangerous product. 47 By contrast, food is characterised by a diversity of products, some of which we want to encourage people to eat more of (fruits and vegetables) and some of which we want to encourage people to eat less of (junk foods). 48 Meanwhile, alcohol is a product which is generally accepted to be harmful to health overall, but where the risks vary based on dose and manner of consumption. 49 These differences are explored elsewhere in this special issue, which addresses what aspects of the substantive law on tobacco and international trade and investment can apply to food and alcohol regulation, given the more complex nature of regulating alcohol and food for public health purposes. They will not be addressed in significant detail here.

III POTENTIAL FUNCTIONS OF HEALTH INSTRUMENTS IN MANAGING FRAGMENTATION BETWEEN HEALTH, TRADE AND INVESTMENT A Treaty Conflict and Treaty Hierarchy
As can be seen from the above, instruments for NCD governance are generally 'softer' than the treaties which regulate international trade and investment law. The 'harder' instruments in trade and investment law and the more developed dispute settlement systems in such regimes, as compared to the picture sketched above, have often created anxieties within the health sector, and led to calls for stronger instruments in health to compete with trade and investment agreements.
Such proposals have been common in the tobacco control context. For example, the negotiation of the WHO FCTC included proposals and draft text for a clause that would specifically allow the WHO FCTC to take precedence over trade agreements, although such a clause was not ultimately included. 50  normative counterweight to address the health impacts of increased trade in unhealthy foods and alcohol. 52 As such, the first potential use of health instruments to consider is their use to prevail over trade and investment instruments. There are various ways in which a health treaty might be made hierarchically superior to trade and investment agreements, for example by:  Removing disputes about health from the scope of dispute settlement in trade and investment treatiesfor example, by barring the jurisdiction or admissibility of a dispute, waiving the right to bring a dispute under trade or investment instruments, revoking consent to arbitrate where this is a condition of accessing investor-state dispute settlement, and instruments purporting to terminate or amend the agreement providing jurisdiction;  Removing health-related measures from the scope of the substantive obligationsterminating or amending the treaty, rendering the obligation inapplicable or unopposable for a class of disputes, reading in exceptions, or generally stating that obligations in a health treaty are to take precedence over any conflicting obligations in trade or investment treaties;  Requiring formal deference by trade and investment tribunals to decisions of health bodies, for example, by requiring them to adopt the same findings of fact as health bodies.
At first glance, this seems to be the simplest way of addressing potential health, trade and investment interfaces. However, whether a health treaty can remove or nullify the risk of trade and investment dispute settlement in the absence of corresponding amendments to trade and investment agreements is in fact a fairly complex question. Most of the rules concerning treaty conflict as well as the jurisdiction of specialised tribunals such as trade and investment bodies have arisen in the context of a decentralised, largely 'contractual' system between theoretically equal sovereign states. They depend in large part on political coordination by states to address potential overlaps. For overlaps between multilateral, 'law-making' treaties that set out comprehensive regimes of both rights and obligations, 53 there are numerous open questions as to what the adoption of an instrument in another forum can do. These include questions about how they interact with the jurisdiction and rules of trade and investment tribunals under the applicable law, what happens if not all parties to trade and investment agreements are party to a new health treaty, and whether general conflict clauses will capture all forms of conflict that are of concern to health actors. 1

Jurisdiction and Applicable Law Rules of Trade and Investment Tribunals
The first issue is the extent to which trade and investment tribunals have jurisdiction to consider 'other' instruments that purport to override or contract out of obligations under trade and investment treaties. Specialised tribunals take their jurisdiction from the instrument that creates 52  them, which typically is limited to the interpretation or application of that instrument. The limits to this jurisdiction can sometimes also be reinforced by provisions that specify the law that the tribunals are to apply. For example, WTO panels only have jurisdiction to hear disputes about the interpretation or application of the 'covered agreements', and rulings under the WTO dispute settlement system 'cannot add to or diminish the rights and obligations provided in the covered agreements'. 54 Although the provisions of the WTO are not to be interpreted in 'clinical isolation' from the rest of public international law, 55 so far, panels and the Appellate Body have been relatively unconvinced by arguments that other instruments modify the WTO agreements or remove the jurisdiction of its panels.
In The WTO Appellate Body has so far been relatively unwilling to recognise that members can contract out of WTO obligations on an inter se basis. In Peru -Agricultural Products, the Appellate Body, found that, at least for conflicting regional or bilateral trade agreements, parties could only modify agreements as between themselves in line with WTO provisions that allow them to do so. 64 Although the judgment does not address non-WTO agreements that are not regional or bilateral free trade agreements, it does seem to suggest that parties can only contract out of WTO obligations to the extent that the WTO rules themselves provide for doing so.
There is some scope to the use of other international instruments to interpret WTO obligations, although the extent of overlap required to do so is not fully resolved. Article 31(3)(c) of the Vienna Convention on the Law of Treaties provides that 'there shall be taken into account … any rules of international law in force between the parties'. 65 The Convention does not specify whether this means all parties to a treaty, or only the parties to the dispute. In the WTO context, this has led to some uncertainty over how other treaties affect the interpretation of the WTO covered agreements if the membership of the treaties does not fully overlap. Panels have previously interpreted that 'parties' means all of the parties to the WTO agreements. 66 However, the Appellate Body has since determined that the relevant question is the extent to which an interpretation taking into account the existence of the other agreement reflects the common intention of the WTO membership as a whole. 67 Potentially, this may not require all members, although it does appear to require enough members to affect the understanding of the WTO agreements on a multilateral basis. 68 Additionally, it can only be used to the extent that the two treaty obligations can be read togetherarticle 31(3)(c) of the Vienna Convention cannot be used to create an interpretation that contradicts the text of a WTO provision. 69 3

Defining What a 'Conflict' Is
A final issue, particularly for 'conflict clause' proposals, is that in order for a treaty to prevail in the event of conflict, a tribunal needs to find a conflict in the first place. This can be a more complex question than it first appearsnot every instance where the implementation of one treaty is in tension with the implementation of another is necessarily a conflict in the legal sense of the term.
The way in which treaty conflicts should be understood and resolved has been a long-running and difficult problem across a number of areas of international law, 70 and it is not the intention of this article to cover the topic comprehensively. However, even a relatively brief review of examples shows that a tribunal with a mandate under one regime can be adverse to finding that it conflicts with another, and can sometimes consider the obligations to accumulate rather than conflict, even if, in practice, states would be faced with a choice as to which to prioritise. For example, a tribunal might find that it is technically possible to comply with both treaties by adopting a different measure. This can be the case if one treaty imposes an obligation of result, and another imposes an obligation of means. For example, in some cases concerning the intersection between the right to water and international investment law, tribunals have found that the obligation to fulfil the right to water did not conflict with obligations towards foreign investors, because there were other means open to the state that allowed it to comply with both treaties. 71 As such, most investment versus right to water jurisprudence only takes the right to water into account to a very limited extent. 72 It can also be the case where alternatives, even relatively onerous ones, are open to the state, allowing it to comply with both obligations. For example, the UN Security Council targeted sanctions regime is formally hierarchically superior to the European Convention on Human Rights, by virtue of UN Charter article 103. However, European Court of Human Rights jurisprudence on the interaction of the European Convention on Human Rights with UN Security Council resolutions has generally required a state to take quite extensive steps to try to resolve the conflict before it can invoke article 103. In Nada v Switzerland, for example, the ECHR considered a conflict between Switzerland's obligation to respect Convention rights and its obligation to implement targeted sanctions restricting the claimant's freedom of movement and thus his enjoyment of the right to private and family life. It found that, despite the very limited discretion it had to do so, Switzerland should have done more to ensure that its implementation of the sanctions regime was adapted to the particular personal and medical needs of the applicant, and that it could have done more to ensure that he was removed from the UN sanctions list in a timely manner after he was found to no longer present a security risk essentially it had to try to reduce or remove the conflict before it could rely on the conflict clause. 73 Although this approach is a technique of human rights bodies arising from the special character of human rights obligations (as well as the particular accountability problems arising from the lack of review of Security Council actions), it does demonstrate that a clause establishing a hierarchy does not, per se, exempt a state from considering the requirements of the hierarchically lower treaty.
More broadly, two instruments may not be considered to interact at all, because they do not have the same subject matter. For investment law, the parties to an investor-state dispute will not necessarily be the same as the parties to a state-state treaty. 74  The cases outlined above do not necessarily represent the only or 'correct' approaches to identifying and resolving conflict. The approach adopted in many of the right to water and investment law cases, for example, seems to assume that a conflict only occurs when it is impossible to comply with both normsan approach that has been criticised in other contexts for always resolving conflict in favour of the most prohibitive norm, and for obscuring how a conflict is resolved by framing it as how the conflict is defined. 77 Broader definitions of conflict note that it can arise not only where obligations are mutually exclusive, but also where obligations contradict provisions providing for rights. 78 However, even these broader approaches may not capture the kinds of 'conflict' that many 'trade and health' or 'investment and health' proposals aim to address. For example, a legal definition of conflict that includes both rights and obligations still requires that a right can be mapped onto a particular obligation and the overlaps identified. By contrast, many 'trade and' or 'investment and' conflicts also concern conflicts of objectives, such as conflicts between provisions which require the attainment of an objective while not being prescriptive as to the means of doing so; or conflicts between provisions whose requirements are themselves the subject of dispute. In such cases, when the relevant course of action is undertaken to fulfil a treaty obligation but is not specifically identified under the treaty or its subsidiary instruments, the conflict (and conflict clauses) can equally be avoided by downplaying the extent to which the treaties overlap, potentially generating unhelpful jurisprudence that diminishes the importance of the 'other' treaty. This is not necessarily a reason not to adopt such a clausetribunals could of course resolve the conflict in many other waysbut it does demonstrate the need to consider what it means for one treaty to prevail over another in quite precise and detailed terms. 4

Conclusion -Treaty Conflict and Treaty Hierarchy
These questions of jurisdiction, applicable law, interpreting conflicts between incongruous parties, and defining what kinds of conflicts matter suggest that designing a health treaty regime that will effectively exclude the risk of trade and investment dispute settlement is a matter of some complexity, potentially requiring coordinated changes across multiple regimes. A simple statement that one treaty prevails over another is not necessarily a guarantee that a case cannot be heard under the hierarchically inferior treaty.
Of course, establishing a hierarchically superior treaty can be important in other ways, including by affecting how states treat the relative importance of such obligations. Ultimately, it is states that are responsible for managing conflicting obligations. As such, explicit regulation of conflicts can be useful for setting priorities within states, and for empowering civil society advocacy to hold states accountable for legal obligations as well as political commitments to health. Conflict clauses may well be important expressions of state priorities and political commitments, or have important consequences in domestic law. And treaties can obviously be 76  valuable for a variety of reasons other than their effect on trade and investment adjudication. 79 But their impact on the risk of disputes is uncertain in the abstract, and requires careful identification of the potential conflicts and the legal mechanism for resolving them, phrased in a way that can be assimilated into the considerations of the trade and investment tribunals hearing challenges to their jurisdiction or arguments that their instruments have been overridden.

B Treaty Interpretation, Fact Finding, and Benchmarking
A second way in which trade and investment tribunals take into account 'other' law is through using them to interpret and apply concepts in trade and investment law. There are several ways in which 'non-trade' instruments have been considered in the jurisprudence of trade and investment tribunals, including to inform the ordinary meaning of words, as evidence of good faith regulatory purposes by states, to benchmark standards of reasonableness or proportionality, and as sources of fact. 80 1

Ordinary Meaning
Although the formal use of article 31(3) of the Vienna Convention has been relatively circumscribed in the WTO context, an interesting parallel development has been the use of other instruments to interpret the ordinary meaning of words. For example, in US -Shrimp, the Appellate Body used the Convention on Biological Diversity to interpret the meaning of 'exhaustible natural resources' in the exception to include living resources in GATT article XX(g), 81 while in EC -Biotech, after finding that the parties were not congruous and article 31(3) did not apply, the panel continued to consider a range of outside resources in terms of the light they shed on the 'ordinary meaning' of the words in article 31(1), considering it to 'provide evidence of the ordinary meaning of terms in the same way that dictionaries do'. 82 The use of other instruments to interpret ordinary meaning appears to be one that is not prescriptive about the kind of instruments that can be used, although unlike article 31(3)(c), it is not mandatory. 83 2

Evidence of Regulatory Purpose
A related use of 'other' international instruments is to determine what the purpose of a measure is, for the purposes of determining how public health exceptions should be applied. The fact that a measure implements a commitment in health or is in line with a consensus position of a health institution is generally treated as strong evidence of a bona fide public health purpose.
In police powers and provided fair and equitable treatment. 84 In US -Clove Cigarettes, the panel accepted that a ban on flavoured cigarettes made a material contribution to the legitimate objective of protecting public health, because it was a measure recommended in the WHO FCTC partial guidelines for articles 9 and 10 of the FCTC. 85 3

Measures of Reasonableness
Non-trade or non-investment international instruments may also be used to benchmark reasonableness standards. There are many points in both trade and investment law where opentextured terms such as reasonableness, proportionality, legitimacy, justifiability and so forth must be interpreted. In trade law, these include: In investment law, they include:  Whether or not a measure is a valid exercise of police powers; 90  Whether or not a measure is 'reasonable' or 'arbitrary' under the fair and equitable treatment standard; 91  What an investor can 'legitimately expect' as a component of the fair and equitable treatment standard. 92 Normative instruments in many areas of international law have been used to benchmark such concepts. Examples include:  The affirmation in Philip Morris v Uruguay that the FCTC is a 'point of reference' for determining whether or not a tobacco control measure is 'reasonable', and its finding that the guidelines supported the reasonableness of Uruguay's large graphic health warnings on cigarette packages; 93  The use of the right to water to frame the 'legitimate expectations' of the investor in Urbaser v Argentina; 94  The use of environmental treaties to interpret the concept of fair and equitable treatment in Chemtura v Canada and Glamis Gold v United States; 95  The consideration of a range of materials from health and environmental law to assess whether or not Brazil's import ban on retreaded tyres was 'necessary' to protect human health under GATT article XX(b) in Brazil -Tyres; 96  The use of the FCTC Partial Guidelines on articles 9 and 10 to determine whether or not a US ban on clove cigarettes was more trade-restrictive than necessary for the legitimate objective of protecting public health in US -Clove Cigarettes. 97 4

Factual authority
Additionally, normative instruments are often used as evidence or sources of factual authority in trade and investment disputes. For example, in Philip Morris v Uruguay, the tribunal found that the WHO FCTC guidelines on article 11 provided evidentiary support for Uruguay's graphic health warnings and single presentation requirement in relation to whether the measure was 'arbitrary' under the fair and equitable treatment standard. It considered that Uruguay was entitled to rely on evidence based and cooperative international processes under the WHO FCTC rather than conducting local studies on the likely impact of its measures. 98 In US -Clove Cigarettes, the Partial Guidelines on articles 9 and 10 were used to reinforce scientific evidence that flavoured cigarettes were attractive to youth, and to support the evidence in favour of the US's ban on characterizing flavours in cigarettes. 99 5

Conclusion -Use of International Instruments in Interpretive Processes
Across all of these uses, a key feature is that the other instrument is not treated as one that conflicts or modifies the trade/investment agreement, but rather, one that sheds light on how it should be interpreted or applied. Non-trade or non-investment instruments therefore must be assimilated to an equivalent concept in trade and investment law, and they cannot override the treaty text itself.
However, as the above discussion shows, provided there is an entry point in the treaty to do so, there are many ways in which other international instruments can influence the interpretation of a trade and investment agreement. As detailed above, there are several such entry points in the substantive standards of trade and investment treaties, and many of the trade and investment provisions which are relevant to NCD prevention measures have fairly malleable meanings.
International health instruments will therefore be important in the outcomes of trade and investment and health cases on the merits. On the other hand, there are fewer such entry points at the jurisdictional stage, which means that interpretive approaches cannot easily remove the jurisdiction of trade and investment tribunals altogether. C

Processes and Institutional Governance
The final way in which instruments are taken into account is through the various procedures they establish. Instruments often come with institutions, actors and processes, such as Conferences of the Parties (COPs), Secretariats, or implementation conferences. Such institutions can be fairly active in supporting states facing trade and investment dispute settlement, whether by observation and monitoring, information sharing, providing capacity building and other support, assisting in internal coordination, issuing statements in support of members, or acting as third party interveners in cases. 100 Many of these activities have had a significant influence in tobacco litigation. For example, WHO, the WHO FCTC Secretariat, and the Pan American Health Organization ('PAHO') filed amicus curiae briefs in support of Uruguay in the Philip Morris v Uruguay arbitration, outlining how Uruguay's measures were supported by the WHO FCTC and scientific evidence. 101 These amicus briefs were an important source of evidence considered by the Tribunalthe panel majority largely accepted WHO, the Convention Secretariat, and PAHO's conclusions on whether Uruguay's single presentation requirement was a reasonable measure to implement the WHO FCTC. 102 Similarly, in 2010, the WHO FCTC COP adopted the Punta del Este Declaration, which affirms the sovereign right to regulate for public health, to express support for Uruguay shortly after the commencement of the Philip Morris arbitration. 103 The Court of Appeal of England and Wales has cited the Punta del Este Declaration in relation to the UK's plain packaging litigation, to confirm that provisions providing for flexibility in the application of the TRIPS Agreement apply to tobacco control as well as in their original context of access to medicines. 104 Additionally, institutions can be highly influential in internal priority setting, and in shaping how states participate in other regimes. For example, after a number of investment tribunals decided to exercise jurisdiction in cases involving intra-EU BITs, the European Commission began infringement proceedings against EU members that were party to intra-EU BITs, resulting in the mutual termination of several such BITs. 105 Less drastically, the FCTC COP has adopted a number of decisions calling on parties to improve internal coordination with respect to tobacco control and trade and investment law. 106 Institutions may also themselves participate in other regimes, providing input and information based on their areas of specialised expertise, and potentially shaping other institutions' understandings of relevant factual matters or shared legal concepts. Such processes are important because they can shape how the relationship between different regimes are understood in the day-to-day business of implementing, enforcing, and developing norms, and can play a role in such understandings beyond those relatively rare occasions where a formal dispute arises. 107

IV FEATURES STRENGTHENING THE POTENTIAL USES OF HEALTH INSTRUMENTS
As such, there are many ways in which international instruments outside trade and investment might affect the way in which trade and investment tribunals make decisions with implications across regimes. How many of these might also apply to non-binding as well as binding instruments?
For obvious reasons, only binding instruments are relevant to formal conflict or hierarchy between treatiesa non-binding instrument cannot prevail over a binding one as a matter of formal law. Under either an interpretive or a process-based paradigm, however, the scope of instruments that can be considered is wide, and does not appear to be limited to instruments that are binding on both parties. For example, in US -Clove Cigarettes, the panel cited the WHO FCTC partial guidelines on articles 9 and 10 even though neither the US nor Indonesia was party to the WHO FCTC. 108  That said, there are many features of treaties, deriving from their status as multilaterally negotiated instruments, which make them more likely to be used by trade and investment tribunals. As detailed below, features of the WHO FCTC which have proven useful in legal challenges against tobacco control measures include the fact that it is high profile and clearly the authoritative instrument in its field, the fact that it is clear and specific about what parties must do to protect public health, the fact that it is evidence based and encompasses processes for scientific exchange and cooperation, and the fact that it is supported by a range of institutions and processes. A

Political Visibility
A key achievement of the WHO FCTC has been to raise the profile of the importance of global tobacco control. As an instrument with 181 parties, which has a governing body and secretariat which are dedicated to advancing its goals, has legally binding implementation requirements, and feeds into other processes such as human rights bodies and the Sustainable Development Goals, 113 the WHO FCTC is a highly visible commitment to tobacco control. This is a key advantage of legally binding instrumentsthey are difficult to ignore, and they require tribunals in other areas to at least turn their minds to whether or not they should take into account coherence across both institutional mandates and international law as a whole. Binding instruments often also require more internal processes, encouraging greater multisectoral coordination within states, and it can often be easier to convince non-health sectors of government of the need to implement them, given their binding nature. Finally, visible expressions of commitment send signals to other actors, and can therefore be important to concepts in investment law such as whether or not investors have 'legitimate expectations' about certain forms of regulation. 114 B

Normative Specificity
Another key advantage of some legally binding instruments is that they need to be implemented, and thus generally specify what states need to do. Thus, they tend to provide more normative specificity as to whether a measure is needed, making it relatively easy to link the objectives of a measure to the instrument that it implements. This can often be important both for establishing whether or not a measure is a health measure, as well as in establishing the necessity and proportionality of a measure, particularly in the context of standards that ask tribunals to consider what alternative measures could be taken. The fact that the WHO FCTC and its guidelines specifically recommend the implementation of large graphic health warnings, for example, was a key part of the Philip Morris v Uruguay tribunal's finding that Uruguay's graphic health warnings covering 80 per cent of the front and back of the principal display areas of tobacco packs was 'reasonable'. The WHO FCTC has also been important in a number of domestic cases for countering industry arguments that there are 'alternatives' to particular tobacco control measures. 115 Conversely, some binding obligations, such as the human right to water, have proven quite difficult to invoke in other fora, because complainants have been able to point to other measures states could have taken instead. 116 C

Scientific Authority and International Consensus
As international instruments are often used as a source of evidence, the scientific authority of such instruments is important to their use in other regimes. The relevance of the WHO FCTC guidelines as reflecting 'the best available scientific evidence and the experiences of the parties' and the WHO FCTC COP as a forum for scientific and technical cooperation have both been cited in tobacco litigation to support the evidence base behind a measure. 117 Additionally, international instruments can embody a shared international consensus on certain factual questions, which can support both their factual authority and their role as benchmarks in standards of reasonableness or proportionality.

Institutional Support
Finally, processes for managing fragmentation are important, whether they take the form of amicus curiae briefs, capacity building, cooperative processes, or otherwise. Legally binding instruments often provide for such processes explicitly, setting up institutions that are then able to advance the goals of a treaty. 118 Treaty implementation also frequently involves domestic level review at the ratification stage, as well as ongoing processes for implementation, which can help establish systems to ensure policy coherence at the domestic level. Non-binding instruments do not set up such institutions by default. However, many non-binding instruments do include extensive processes to support implementation, and are managed by existing international institutions with the capacity to cooperate with other institutions. 119 Relevant programmes of work and/or further cooperative arrangements could be developed through resolutions of the WHA and/or the UN General Assembly. 120

V CONCLUSION
Overall, the relationship between health instruments and trade and investment dispute settlement is complex. It is difficult for health bodies alone to fully address intersections with trade and investment regimes, because in a decentralised system of obligations and adjudicators, much of how overlapping obligations should be considered and prioritised has been left to states to coordinate internally. Managing fragmentation is not simply a case of writing that health should prevail over trade and investment law into a new health treaty and declaring the task finished.
Nevertheless, health instruments can exert significant influence on the development of trade and investment law through interpretive and supportive processes, and this does not necessarily require instruments to take a specific form. Treaties, such as the WHO FCTC, combine legal authority, specificity, political commitment, scientific evidence, and institutional support in ways that have proven powerful in managing cross-regime conflict. But it is also the case that these features are present in varying combinations across non-binding instruments with application to food and alcohol. For example, the right to health in the Universal Declaration of Human Rights 121 125 are both normatively specific and evidence based. And of course, the scientific authority of instruments does not depend on their legal statusnon-binding instruments can be more authoritative in this respect because they can be updated more easily as study or exchange of information takes place (although of course, the 'hardness' of an instrument can often reflect that there is a strong consensus in the first place). International food and alcohol policy are likewise supported by processes, whether through the WHA or the WHO Secretariat, which may have a role to play in supporting policy coherence across health, trade and investment law. As important as it may be to develop stronger instruments around food and alcohol, it is also important not to underestimate (or to allow industry actors to diminish) what we can do with what we have.
To conclude, using health instruments to promote greater coherence across health, trade, and investment is not just a question of their formal status, but also the functions we want them to serve. As states take further action to address the growing burden of NCDs, it will therefore be important to consider how international instruments can best be used to support coherence across trade and investment law, where the gaps might be, and what might be the best way of filling them.