2. the indus waters dispute

The most recent manifestation of the claim to customary status of the precautionary principle is found in the Indus Waters arbitration, which was initiated by Pakistan in 2010 under the Indus Waters Treaty 1960 between India and Pakistan (the first dispute between the parties to be settled under the Treaty).^{Footnote 6} The dispute relates to the construction and operation of the Kishenganga Hydro Electric project (KHEP), a 330 megawatt (MW) project located on the Kishenganga River in India-administered Jammu and Kashmir. The parties disagreed as to whether India’s diversion of water conforms with the 1960 Treaty, whether it would have undue effects on Pakistani agricultural and hydroelectric uses, and whether India was entitled to bring the reservoir level of the KHEP below ‘dead storage level’. Central to the dispute was the question if, by diverting the water, India had violated its obligation to ‘let flow’ waters as defined in Article III of the Treaty, or whether this terminology allowed India to carry out certain restricting activities. In the context of diverting water into another tributary, which was provided for in the Treaty, Pakistan asserted that the wording ‘if necessary’ ought to be interpreted in light of the precautionary principle. India countered this argument. It asserted that the principle does not amount to customary law and, moreover, that the many meanings of the principle prevent it from being applied in a consistent manner.^{Footnote 7}

In its partial award of 18 February 2013, the PCA ruled that it was incumbent upon it to ‘interpret and apply [the] Treaty in light of the customary international principles for the protection of the environment in force today’.^{Footnote 8} However, in its final award, the Court found that, irrespective of the status the precautionary principle may have in international law, the Indus Waters Treaty specifically limited the extent to which the Court may take into account sources of law external to the treaty itself.^{Footnote 9} In reaching its decision, the PCA emphasized that:

In taking this approach, the Court indirectly alluded to the contentious nature of the principle by indicating that it belongs in the realm of policy.

3. the precautionary principle

Essentially, the precautionary principle offers decision makers a platform on the basis of which to take decisions when faced with uncertainty. From its origins in German domestic law (or Swedish, according to some),^{Footnote 11} the most commonly cited iteration of the principle is found in Principle 15 of the Rio Declaration, which provides:

Prior to its inclusion in the Rio Declaration, references to the principle and to ‘precautionary measures’ more generally could be found throughout international legal materials: ministerial declarations of the International Conference on the Protection of the North Sea in 1984; the Preamble to the Vienna Convention for the Protection of the Ozone Layer from 1985;^{Footnote 13} the Convention for the Protection of the Marine Environment of the North-East Atlantic (OSPAR);^{Footnote 14} and the United Nations Framework Convention on Climate Change (UNFCCC),^{Footnote 15} to mention just a few. Scholars have identified three functions of the principle:

• • the purely negative function – the principle is used to exclude certain activities;

• • the facilitative function – the principle allows but does not require action; and

• • the obligatory function – the principle embodies a duty to take action.^{Footnote 16}

Upon closer scrutiny, the principle contains a number of constituent parts all of which are themselves very wide-ranging. Firstly, a not insignificant part of the principle concerns itself with the issue of risk. That is, very simplistically, the combination of the likelihood that a harm will occur and the magnitude with which the harm presents itself should it occur. While some interpretations of the principle qualify its application by reference to the potential harm being ‘significant’ or ‘irreversible’, the vast majority of definitions refrain from quantifying the threshold of harm.^{Footnote 17} Having said that, the application of the principle may, at times, be restricted by reference to the knowledge threshold of the relevant risk – that is, how indeterminate the threat is. In general, the primary threshold is said to rest where a particular risk is uncertain as opposed to certain (which in general terms would trigger the application of the principle of prevention) or residual (which tends to escape regulatory attention).^{Footnote 18} This wide spectrum of application still leaves significant scope for applying the principle, though it is generally assumed that the principle may be triggered where scientific knowledge is not conclusive but the risk is reasonably plausible.^{Footnote 19}

In many cases, however, the question of when to apply the principle cannot be answered in the abstract as such decisions will often be inherently political. This, in turn, gives rise to the argument that the real impact of the precautionary principle is procedural: that is, in the attempt to ensure that decisions on the application of the principle are made in an open, transparent and democratic manner, the primary obligations arising out of the principle centre on the need for sound administrative decision making.^{Footnote 20} Ideally, the role of the precautionary principle thus becomes facilitative in that it alerts us to the importance of a ‘carefully measured and targeted treatment of different states of knowledge’.^{Footnote 21} Therefore, procedural responses will necessarily have to take into account the potential breadth of the application of the principle and the consequences this has for the task of formulating generic directives on the application of the principle. Consider, for example, that the meaning of the term ‘scientific uncertainty’, which is so central to the principle, will vary significantly from one regulatory sphere to another, as will the manner in which regulators seek to eliminate such uncertainty.^{Footnote 22}

In respect of the environment, the precautionary principle is applied primarily in the context of risks arising from industrial and technological development or risks arising as a result of naturally occurring phenomena. Inherent in this characterization is often an assumption that the various risks are measurable, manageable and capable of being communicated and regulated. This understanding of risk is by many considered a defining feature of modern day regulation, so much so that risk is seen as ‘the organising concept that gives meaning and direction to environmental regulation’.^{Footnote 23} Whereas this conjecture gives the impression that the matters of risk assessment, management and communication are straightforward issues, the facts are rather different. These concepts have no meaning in their own right.^{Footnote 24} Risk and uncertainty, and by extension the precautionary principle, instead operate in a sphere that is highly contingent for reasons of methodology (how best to assess a risk), epistemology (what types of risk are we actually able to identify) and value (once identified, what risks are worth attending to?).^{Footnote 25} All these issues necessarily depend on, and are defined by, the context in which the specific risk arises.^{Footnote 26}

A particularly informative illustration of the complexity of risk is found in the work of Jasanoff, who alerts us to the highly contingent area in which these considerations play out. In her comparative work examining the approaches to risk and regulation in relation to biotechnology, Jasanoff highlights that these are socially construed in a manner which varies significantly from jurisdiction to jurisdiction, depending on political cultures and regulatory styles.^{Footnote 27} In the United States (US), the preferred approach is one which maintains a focus on scientific pluralism and openness. In the United Kingdom (UK) the decision-making process is one of informative cooperation in which a number of selected parties take part, whereas in Germany a culture of bureaucracy furnishes a much more programmatic response to the risks.^{Footnote 28} Altogether, this results in a situation where the approach to, and the management and regulation of risk is highly conditioned by social parameters.

The contingency is further highlighted if we consider the reference to cost effectiveness as this forms part of various incarnations of the principle (though not necessarily all of them). Principle 15 thus includes a reference to ‘cost-effective’ measures, as does the reference in the UNFCCC,^{Footnote 29} unlike, for example, the reference in the 1995 Fish Stocks Agreement.^{Footnote 30} The assessment of whether a regulatory response is warranted on grounds of cost effectiveness necessarily includes engaging in highly normative considerations of the value to be attached to a particular good. Nevertheless, whether cost–benefit analysis forms an explicit part of the principle may not seem all that important simply because any regulatory decision on whether or not to act in the face of uncertainty inevitably entails some basic weighing of advantages and benefits against disadvantages and costs.^{Footnote 31} It is against this background that the debate on the status of the precautionary principle takes place.

4. the role of customary international law and the two camps

We should start with a brief consideration of what may lie behind the desire to confirm the precautionary principle as customary international law. Presumably, the primary reason for invoking the customary status of the principle, or indeed any other legal norm, is that the principle is then considered to have legal force in its own right, in so far as customary international law is recognized as a separate source of international law.^{Footnote 32} For a practice to amount to a customary norm, it is generally accepted that the practice must be consistently adhered to among states (though not all states), that it must have endured for some time, and that the states acting in accordance with the practice do so out of a sense of legal obligation (known as opinio juris).^{Footnote 33} The real advantage of this is that a customary rule provides a peg on which to pin a legal claim in the absence of any explicit treaty provision. A customary rule thereby serves as a background against which other international obligations are to be interpreted. When faced with interpretative difficulties of, for example, a treaty provision, customary rules partly serve to fill the gap created by the treaty provision (though customary obligations may well develop in the absence of treaty obligations). A good example of this is found in the Legality of Nuclear Weapons opinion, in which the International Court of Justice (ICJ) identified the general obligation of states to ‘ensure that activities within their jurisdiction and control respect the environment of other States or of areas beyond national control’ as a customary rule (the so-called ‘no-harm rule’).^{Footnote 34} Similar declarations of customary international law pertaining to the environment include the principle of permanent sovereignty over natural resources, allowing a state to dispose of its wealth and natural resources, as identified in the Armed Activities Case.^{Footnote 35}

What this gives rise to is a situation where certain basic customary rules operate as a ‘constitutional’ foundation for each separate discipline within the broader framework of international law.^{Footnote 36} For international environmental law these rules are, as indicated, the permanent sovereignty of natural resources, the ‘no-harm rule’, and the obligation, under certain circumstances, to conduct environmental impact assessments (EIAs). The argument in favour of customary status advanced by the custom camp seeks to add the precautionary principle to this group of rules, thereby elevating the principle to a foundational rule in international environmental law – a rule against which other norms and directives are to be interpreted and by which all states are bound (except where they persistently object).

Support for the claim that the precautionary principle has achieved customary status is thus found primarily in the extensive use of the principle in international conventions, documents and agreements. This claim has intuitive appeal. The precautionary principle has made its way into numerous international environmental agreements covering such wide issues as marine pollution, biodiversity, ozone depletion, climate change, and waste; one count suggests that the principle is mentioned in over 50 multilateral instruments.^{Footnote 37} In addition to this, the principle has been relied upon by several international courts and tribunals. For example, in the Pulp Mills decision, the ICJ accepted that the ‘precautionary approach may be relevant in the interpretation and application of the provisions of the [disputed treaty]’.^{Footnote 38} Support for the principle was also advanced in the dissenting opinions of Judge Weeramantry and ad hoc Judge Palmer in the follow-up litigation to the Nuclear Tests case.^{Footnote 39} Unexpected support for the principle is also found in the case law from the European Court of Human Rights (ECtHR), where the Court in Tătar v. Romania observed that the principle has developed from a ‘principe d’une conception philosophique’ to a ‘norme juridique’.^{Footnote 40} More recently, the Seabed Dispute Chamber of ITLOS found that the precautionary approach has been ‘incorporated into a growing number of international treaties and other instruments [which in turn] has initiated a trend towards making this approach part of customary international law’.^{Footnote 41}

It is on the strength of this plethora of precautionary references that claims to customary status arise.^{Footnote 42} Pakistan made a similar claim in the Indus Waters arbitration dispute, as did the European Commission before the World Trade Organization (WTO) Dispute Settlement Appellate Body in the Beef Hormone dispute.^{Footnote 43} The argument is often accompanied by the claim that the justification for regulation of the principle in the face of uncertainty presupposes a high level of environmental protection, and that the principle requires a reversal of the burden of proof so that the onus is on the alleged polluter to show that the activity is not harmful.^{Footnote 44}

There are, of course, objections to this narrative. Foremost among these is the argument that the numerous mentions and references to the principle amount to no more than what one commentator calls ‘precaution spotting’.^{Footnote 45} This game of ‘I spy’ overlooks that underneath the multiple references to the principle are variable interpretations and applications of the principle. If we accept Jasanoff’s argument, precaution spotting operates on the mistaken ‘assumption that the principle is an autonomous transplantable rule’.^{Footnote 46} A reference to the precautionary principle in the Vienna Convention^{Footnote 47} does not necessarily mean or entail the same obligations as a similar reference in the OSPAR Convention,^{Footnote 48} and we ought to be careful when drawing general inferences from any such comparisons. Hence, the exercise of ‘precaution spotting’ fails to define ‘the context in which any particular example of the implementation of the precautionary principle is operating’.^{Footnote 49} What is more, the mere reference to several international formulations of the principle need not necessarily be an indication of the fact that a state considers itself legally bound by the principle in a general sense. As Dupuy has highlighted, a state may support the application of the principle in one context but oppose it in another, thus revealing a prima facie yet entirely rational inconsistency in the support for the principle.^{Footnote 50}

Even if the multiple references to the principle in international documents and court decisions are taken at face value, we have reason to question whether such a discursive endorsement is actually indicative of state practice.^{Footnote 51} As Bodansky reminds us, much of what passes for international environmental law ‘reflect[s] not how states regularly behave, but how states speak to one another’.^{Footnote 52} That is, the main sources used as a foundation upon which to stake the claim for customary status may, in the end, not be able to support it. Linked to this is the consideration of how best to assess whether a norm constitutes customary law. Traditionally, though not invariably, this assessment would demand a thorough empirical examination of ‘on the ground’ state behaviour – varyingly described as a ‘Herculean task’ and ‘always problematic to assess’^{Footnote 53} – that investigates domestic implementation, practice and enforcement, as well as the multiple intentions behind such activities. From this perspective, the ‘precaution spotting’ assertion that the principle amounts to a customary rule by reference to international declarations rings hollow, for there is very little to indicate that the proponents of the customary camp have, in fact, undertaken any such review. However, this may not necessarily inhibit the claim to customary law if we consider alternative methodologies of identifying customary law (see below, Part 5).

Against ‘the precautionary principle is plentiful’ claim to custom stands the argument that the principle is far too incoherent and vague to give rise to any such claim. It lacks a unitary logic form.^{Footnote 54} This lack of clarity affects questions such as what degree of risk will trigger the application of the principle, whether the principle aims for zero risk, whether costs may be considered, and what action is dictated by the principle. Thus, it is argued that many of the precautionary references which the custom camp has spotted are not examples of the principle but instead are ‘best understood as an elaboration of the obligation of due diligence … [rather than] some separate species of rule’.^{Footnote 55} A further argument of the no-custom camp relates to the contingency of the precautionary principle discussed above. The role of risk particularly is seen as limiting the international application of the principle. This is so if we, once more, reflect on the cultural contingency of risk. Again, Jasanoff’s work is illuminating:

Not surprisingly, the no-custom camp also has judicial support for its claim. In the Beef Hormones dispute, the Appellate Body observed with regard to the precautionary principle that ‘whether it has been widely accepted by Members as a principle of general or customary international law appears less than clear’.^{Footnote 57} In these circumstances, the no-custom camp argues, it is not possible to derive a clear-cut rule of law which is stable enough to form the basis of a customary norm.

Implicit in the arguments of the no-custom camp is the assumption that in order for a norm to amount to an international legally binding rule, it has to achieve a certain level of precision, be determinate and well-defined. This view finds apparent support in the Continental Shelf case, in which the ICJ observed that in order for a customary norm to emerge, a provision would have to ‘be of a fundamentally norm-creating character such as could be regarded as forming the basis of a general rule of law’.^{Footnote 58} This line of argument follows on neatly from domestic legal settings, where rules are often characterized as reflecting certainty, being applicable in an ‘all or nothing’ manner.^{Footnote 59} The standard narrative goes something like this: rules are the clearly defined commands which offer certainty and uniformity in the context provided for by the rule itself. Principles, on the other hand, are flexible commands which can point in several directions but add interpretative support to the rule (and are therefore not able to amount to a customary norm).^{Footnote 60} Arguably, however, the no-custom camp is guilty of exaggerating the difference between principles and rules. This is so for a number of reasons.

Firstly, the neat distinction between rules and principles alerts us to what Pierre Schlag terms the ‘rules v. standards’ dialectic, though we will modify it for our purposes to the ‘rule v. principle’ dialectic.^{Footnote 61} This dialectic plays out where controversy arises concerning the enactment of a legal directive or norm. The central argument in the dialectic is that this controversy is informed by an assumption that rules and principles respectively embody a number of distinct vices and virtues. Preference may be given to rules over principles, or vice versa, depending on which virtues and which vices cause the highest degree of concern. The virtues associated with rules are certainty, uniformity, stability and security, whereas the vices are intransigence, regimentation and rigidity.^{Footnote 62} The virtues of a principle are flexibility, individualization, open-endedness and dynamism, whereas the vices are manipulability, disintegration and indeterminacy.^{Footnote 63} Arguably the no-custom camp favours the virtues of the rule over those of the principle. Conversely, the no-custom camp heavily emphasizes the vices associated with the principle and argues that these effectively prevent the precautionary principle from achieving customary status.

However, it is important to bear in mind that the dialectic does not seek to argue that everything must either be a rule or a principle. The point of the dialectic is simply to highlight that when controversy arises as to the exact contours of a legal norm or directive, the question is often framed as a ‘rule v. principle’ dispute.^{Footnote 64} The reason why this controversy arises in the first place is often because the precise objectives of the legal norm (be it a rule or a principle) are ambivalent.^{Footnote 65} In the case of the precautionary principle this ambivalence is illustrated in simultaneous but somewhat conflicting emphasis on the prevention of environmental risks as well as on cost effectiveness.^{Footnote 66} Thus, the principle may be invoked on one occasion in support of measures to raise environmental protection, and on another in support of measures to ensure cost effectiveness. In this ambivalence the ‘rule v. principle’ dialectic rises to prominence.

The dialectic, moreover, alerts us to the claim that the ‘rule v. principle’ line of argument rests on a mistaken supposition – namely, that rules or principles are always either certain or flexible;^{Footnote 67} that is to say, that rules and principles define the context in which they operate. Challenging this assumption, Schlag points out that no rule or principle is ever able to ‘control or determine the context within and from which it is interpreted’, just as there is no secured or unequivocal background or context against which a rule or principle must be applied.^{Footnote 68} Attempts to discern the content of any principle ‘will always and necessarily proceed from [a particular] vantage point’.^{Footnote 69}

If this is correct, it gives rise to a situation where the above perceived difference between rules and principles breaks down. What does this mean for the precautionary principle? It seems that any claims for custom made by the custom camp run aground when countered with the argument that no principle is ever able to ‘form the basis of a general rule’ as no principle will ever specify its own context. Moreover, the breakdown of the ‘rule v. principle’ divide reverberates far beyond the precautionary principle. Established customary norms, including the ‘no-harm rule’, are no less unstable and hence constitute an equally unsuitable basis for claims to customary status.^{Footnote 70} Drawing the argument to its radical conclusion is likely to be met with consternation and head shaking. It therefore merits further scrutiny.

5. a third understanding and the implication for custom

The implication for customary international law that follows from the discussion above is that customary international law is a much more ambiguous concept than perhaps traditionally considered, and it seems that the identification of potential customary rules is a rather difficult task. Is it at all possible to identify customary rules? The answer may well be ‘no’. At the very least it seems that it may be impossible to identify one method in accordance with which we are able to define customary rules, and we may have to accept that there are different approaches towards identifying these rules. For example, we might adhere to a definition of customary law which lends strong emphasis to the element of state practice.^{Footnote 71} If this is the case, a claim to custom may be fatally undermined where no thorough examination of state practice has been undertaken, as a result of the empirical evidence being weak. However, the claim might still be rescued on the basis of alternative methods and means by which to identify customary rules – methods and means that do not necessarily see state practice as a central element. These ‘non-empirical’ and ‘norms of the literature’ approaches (to use Kelly’s terminology) differ from traditional approaches to customary law in that they ‘are declared without either general, consistent practice or clear evidence that the vast majority of states have accepted the norm as a legal obligation’.^{Footnote 72} Instead, approaches to identifying customary rules lend emphasis to statements and declarations of rules and opinio juris.^{Footnote 73} Where such approaches are followed, new customary norms arguably emerge more readily than under the ‘traditional’ approach where the sum total of customary rules is comparatively small.^{Footnote 74}

Bearing this in mind and desirous for a constructive response to the challenge of identifying customary rules, perhaps we are best served by a pluralist account of the development of customary rules.^{Footnote 75} If customary norms are not static, then perhaps neither is there a fixed and unchanging approach towards their development and identification. This has obvious implications for the current debate between the two camps in so far as the custom camp relies on the ‘modern’ approach, whereas the no-custom camp favours the ‘traditional’ state practice-centred definition. One implication to follow from this relates to the types of argument that it is possible to advance in support of a claim to customary status.^{Footnote 76} Consider, for example, the argument advanced by the no-custom camp that actual state practice (empirically ascertained) does not in reality support any claim to customary status of the precautionary principle. Such a process-based argument follows naturally from the strong emphasis on the ‘traditional’ state practice-centred approach to customary law. On the other hand, the ‘precaution spotting’ argument advanced by the custom camp follows neatly from the ‘modern’ approach, which emphasizes declarations and statements at the expense of state practice. We are therefore forced to accept that customary rules may emerge through different methods at different points in time. As illustrated, different international courts and tribunals have reached different outcomes when faced with the question of what status the principle has in international law. This may seem disconcerting, but it is not entirely surprising. After all, different judges and thereby different tribunals will have different preferences for methods of establishing customary rules.^{Footnote 77}

Nevertheless, such reflections serve to highlight the contingency in which international legal rules and principles operate, including the precautionary principle. They also remind us that what today is generally recognized as customary law is not necessarily any less incoherent or vague than the precautionary principle. Therefore the no-custom camp’s claim arguably rests on a misunderstanding regarding the clarity and uniformity of existing norms of custom. That is to say, the no-custom camp is perhaps guilty of confusing the normative force and thus the status of a rule with its effects. In practice, the fact that the application of a norm varies and thereby gives rise to a range of less well-defined activities does not necessarily preclude it from achieving customary status.

A good example of this is found in the legal obligations relating to transboundary environmental impact assessment (EIA), which are generally assumed to have achieved customary level, notwithstanding that their exact content is, at best, rather vague. Consider, for example, the ICJ’s decision in the Pulp Mills case.^{Footnote 78} Here, the Court readily asserts that the treaty obligations in dispute are to be interpreted in light of the requirement to carry out an EIA: ‘It may now be considered a requirement under general international law to undertake an environmental impact assessment where there is a risk that the proposed industrial activity may have a significant adverse impact in a transboundary context’.^{Footnote 79} So far, so good. The Court then goes on acknowledge that ‘it is for each State to determine in its domestic legislation or in the authorization process for the project, the specific content of the EIA required in each case’,^{Footnote 80} thereby injecting indeterminacy into the EIA requirement immediately after its elevation to a principle of customary international law. The only conditions which the Court attaches to the requirement to carry out an EIA is that one such assessment must be carried out prior to the project being implemented and, where appropriate, post-implementation monitoring ought to take place (a requirement which does not form part of many of the domestic EIA regimes). More importantly for our purposes, the ICJ seemingly finds that the EIA obligations amount to a customary rule without having undertaken the examination of state practice, which is requisite if we apply the ‘traditional’ approach to customary law. In so doing, the Court arguably adds implied support for the no-custom camp’s emphasis on the ‘modern’ approach to defining customary law.

Consider further the example of the so-called ‘no-harm rule’, which is similarly generally seen as part of customary international law.^{Footnote 81} The central obligation arising out of the ‘no-harm rule’ is, famously, one of due diligence. It entails a responsibility on states to enact domestic policies and practices grounded in best practicable means and techniques and to act in consultation with affected states.^{Footnote 82} Having said that, the various incarnations of the rule vary. Article 3 of the International Law Commission’s (ILC) Draft Articles on Prevention of Transboundary Harm from Hazardous Activities^{Footnote 83} applies to ‘significant’ harm, as does Article 7 of the Convention on the Law of Non-Navigational Uses of International Watercourses,^{Footnote 84} whereas Principles 21 and 2 of the Stockholm^{Footnote 85} and Rio Declarations^{Footnote 86} make no such references. Coupled with the fact that the due diligence obligation is one of process rather than result,^{Footnote 87} it becomes evident that the application of the ‘no-harm rule’ does not necessarily give rise to the certainty with which rules are traditionally associated. The reasons for this are partly epistemological and partly methodological: it simply is not possible to draw up a definition of the rule which would cover all possible scenarios, and neither is it evident who should be in charge of such an assignment.

The point to remember is that, if these considerations of the ‘no-harm rule’ and the rules of transboundary EIA are broadly correct, they have a number of similarities with the precautionary principle in respect of our ability to define the exact content of the norms. The main difference is that, in the case of the ‘no-harm rule’ and the rules regarding transboundary EIA, definitional indeterminacy has not prevented them from achieving the status of customary norms. The same cannot be said of the precautionary principle (at least if the no-custom camp is to believed).^{Footnote 88} What thus emerges is a situation characterized by differences in degree rather than differences in essence. On the whole, it may therefore seem that the custom camp has a stronger case, given that there are juridical precedents of principles being recognized as customary norms without the requisite clarity or empirical back up. In reaching this conclusion, however, it should be emphasized once again that no rule, standard or principle is able to specify its own context. If we embrace such a permissive approach to customary law, we should bear in mind the potential risks, one of which may be that customary norms are perhaps too readily established.