The role of precedent in the jurisprudence of the International Court of Justice: A constructive interpretation

Abstract

In cases that come before the International Court of Justice (‘the ICJ’, or ‘the Court’), its own jurisprudence looms large. This is despite the fact that no rule of stare decisis operates before the ICJ. The juxtaposition of the facts expressed in these two statements has generated significant doctrinal and theoretical debate. In this article I propose a novel theoretical framework for understanding the role that precedent has traditionally played in proceedings before the Court, and that which it should play in the future.

This theoretical framework rests upon a constructive interpretation of the doctrine of stare decisis. I examine the historical institutional practice of the Court in order to determine the range of possible versions of the doctrine which, in Dworkin’s words, ‘fit’ this practice. Noting that there is more than one potential version of stare decisis which fits the Court’s practice, I next consider which is justified as the version which best reflects the value of this doctrine. Having shown that the value that stare decisis strives to achieve is contextual justice, I argue that only a weak version of horizontal stare decisis can justify this doctrine. My ultimate claim is that the Court’s principal concern when considering whether to follow or depart from its own jurisprudence should be that of achieving justice in the context of that particular case, and not merely ensuring consistency, predictability or efficiency.

1. Introduction

In cases that come before the ICJ, its own jurisprudence looms large. ¹ This is despite the fact that no rule of stare decisis operates before the ICJ. ² The juxtaposition of the facts expressed in these two statements has generated significant doctrinal and theoretical debate over the years. ³ On top of that, in the last couple of years the former President of the Court has expressed a view of the role of precedent as binding that is inappropriate given the type of court the ICJ is. ⁴ In this article I show why such a view of the operation of stare decisis does not respect the ICJ’s institutional role, and argue that it is important to prevent such a view from becoming entrenched. In order to do so it is crucial that we understand the proper role of precedent in proceedings before the Court. To that end, I propose a novel theoretical framework for understanding the role that precedent has traditionally played in proceedings before the Court, and that which it should play in the future.

This theoretical framework rests upon a constructive interpretation of the doctrine of stare decisis. ⁵ Constructive interpretation ascribes a certain value to the object of interpretation, ⁶ and consequently interprets that object in such a way that reflects that value. ⁷ This is a two-step process. The first step is backwards looking, and requires an examination of the historical institutional practice of the Court in order to determine the range of possible versions of the doctrine which, in Dworkin’s words, ‘fit’ this practice. As such, the range of possible interpretations of the object (in our case the doctrine of stare decisis) identified at this first stage is circumscribed by the Court’s past practice. This ensures that the interpretive exercise produces an interpretation of the doctrine of stare decisis that is actually based in the Court’s practice and not some abstract invention.

The second step of the interpretive exercise is both normative and forward looking. Noting that the first step of the process reveals that there is more than one potential version of stare decisis which fits the Court’s practice, it becomes necessary to consider which is ‘justified’ in Dworkin’s words as the version which best reflects the value of this doctrine. ⁸ Having shown that the value that this doctrine strives to achieve is contextual justice, I argue that only a weak version of horizontal stare decisis can justify this doctrine. My ultimate claim is that the Court’s principal concern when considering whether to follow or depart from its own jurisprudence should be that of achieving justice in the context of that particular case, and not merely ensuring consistency, predictability or efficiency.

2. Constructive interpretation: Some preliminary methodological and terminological clarifications

I take law as a normative system to be an interpretive concept. ⁹ By this I mean to say that law does not depend on any kind of fixed criteria or procedure, but rather on the values and principles which best justify the practice related to that concept. Dworkin tells us that any part of the law or its practice that is disputed in a particular case necessarily requires the adoption of an interpretive attitude in order for us to arrive at the (or at least a) right answer. ¹⁰ This is true in all cases whether hard or easy, regulated or unregulated. ¹¹ In any case brought before the Court, the ICJ judge is bound to make legal determinations of the rights and obligations of the parties, and in doing so the judge will necessarily have to engage with and interpret international law. ¹² In relation to any legal question which the judge must address in order to dispose of the case – the meaning of ‘for scientific purposes’ in the context of a dispute over whaling, for example – the judge must adopt an interpretive attitude in order to arrive at the best answer. ¹³

This interpretive attitude emphasizes that any object of interpretation has a certain value. As such, it must be interpreted in a certain way in order to reflect the value that it is meant to serve. ¹⁴ In that sense, the interpretive attitude makes the content and operation of an object of interpretation dependent on that value. The goal of such interpretation is to give meaning to that which is being interpreted, with the object of interpretation itself being key, rather than any other factor such as the intention of the creator of the law. ¹⁵

In our case, the object of this interpretive exercise is the doctrine of stare decisis. The first step of this exercise, as stated above, involves examining the institutional practice of the ICJ in relation to this doctrine, ¹⁶ in order to determine the version of the doctrine that ‘fits’ this practice. ¹⁷ As such a word on what is meant by ‘practice’ in this regard may be helpful. In broad terms, practice includes all of what we may broadly consider to be international law including the international agreements, customary international law, the Court’s judgments, the submissions of parties and any other legally significant actions and practices of international actors.

For the purposes of our interpretive exercise, this practice can be subdivided into that practice relating specifically to the doctrine of stare decisis, the object of interpretation. In other words, what counts as practice for our purposes is the legally significant actions and practices of international actors relating to the doctrine of stare decisis before the Court. While practice understood in this sense may in principle encompass a wide range of actions and practices, in reality, the lack of a relevant rule in the Court’s constitutive instruments or in customary international law, and the lack of any centralized legislation on this judge-made doctrine, means that relevant legal practice in this context really only consists of the legally relevant actions and practices of the Court itself and parties to cases before it. As such, the first step of the interpretive exercise which examines the Court’s historical institutional practice is limited to such actions and practices of parties and the Court.

Five further preliminary remarks are necessary at this stage. First, interpretivism both as a theory of the nature of law and as a normative method is most commonly associated with the body of work of Ronald Dworkin, and his notions of ‘fit’ and ‘justification’ structure the present article. Nevertheless, it is important to acknowledge that Dworkin’s own ideas evolved over time. ¹⁸ I should like to make clear that the normative method as I see it, and deploy it in this article, is closer to Dworkin’s original presentation of his thoughts in this regard in ‘Hard Cases’. ¹⁹ By this, I mean to say that while I believe the claim that law and its practice are interpretive concepts, we can nevertheless separate the normative method into two stages, ‘fit’ which looks to institutional history, before considering value and justification in the second, normative step. In Law’s Empire, ²⁰ Dworkin collapsed these two steps together, placing the determination of value front and centre in the interpretive method. I believe that this move made the method less precise, which is why I advocate two distinct but related stages of fit and justification in this article. ²¹

Second, perhaps after the umpteenth mention of his name, it may be reassuring to some to learn that I do not believe that the claim that law is an interpretive concept means that one must necessarily accept any broader interpretivist claims regarding the nature of law. ²² Readers will naturally have different views on whether moral facts form part of the law itself, ²³ or whether they are somehow relevant in the interpretation of the law but do not form part of it. ²⁴ Nevertheless, whether one is a dyed-in-the-wool positivist or a diehard Dworkinian regarding the nature of law is immaterial for our present purposes. ²⁵

Third, in any case before the ICJ, parties’ submissions will cite the Court’s precedents in support of their claims. ²⁶ Since I am specifically interested here in the operation of the doctrine of stare decisis, I will assume that there is relevant precedent that brings this doctrine into play that is binding on the judge. Taking for granted that there is precedent that is ‘on point’, ²⁷ I will focus on the ICJ judge’s duty to interpret the doctrine of stare decisis itself.

Next, as mentioned above, in this article I refer to the respect that the Court shows its own precedents as amounting to the operation of a de facto doctrine of ‘stare decisis’. ²⁸ Accordingly, I am only concerned with just that: the Court’s own precedents, and do not make any claims relating to what others have called ‘external case law’. ²⁹

Finally, and perhaps most importantly, a word on terminology. One can safely predict that the mention of the term ‘stare decisis’ in the context of the ICJ (whose procedure is at best a mélange of the civil and common law traditions) ³⁰ will provoke in those lawyers formed in the civilian legal tradition a range of emotions including suspicion, doubt or rancour regarding the relevance of an alien common law legal doctrine associated with higher courts binding lower courts. ³¹ To be clear, my use of the term ‘stare decisis’ does not refer to the doctrine in the form that it operates in any one domestic legal order or tradition. ³² When I use the term I do not mean anything more or anything less than the (I believe uncontroversial) claim that we can observe a practice of ICJ judges taking the Court’s previous findings into account in the course of making legal determinations in a particular case. ³³

3.

The ‘fit’ bit: Horizontal stare decisis as the only possible version in the context of the ICJ

The range of possible versions of the doctrine of stare decisis that emerge from our interpretive exercise is not unlimited but is in fact circumscribed by past practice. ³⁴ There are a range of possible interpretations of the doctrine of stare decisis which include both vertical and (several forms of) horizontal stare decisis. But which best fits the institutional practice of the ICJ? In order to answer this question, it is necessary to consider the nature of stare decisis as a doctrine, before considering its historical application in the practice of the Court to date.

3.1 Rules, principles and the doctrine of stare decisis

The doctrine of stare decisis is one with which lawyers, at least those in the common law world, are familiar. Even to non-lawyers, the idea that lower courts should follow the decisions of higher courts in order to foster stability and predictability in a legal system makes instinctive sense. And yet, as soon as one starts to think a little more about the doctrine of stare decisis, one realizes that it is in fact fiendishly complex. ³⁵ Amy Coney Barrett has described it as a ‘shape-shifting’ doctrine which operates very differently depending on context. ³⁶ In order to understand this doctrine properly, what it is and how it operates, it is helpful to begin with the basics.

Both domestic and international law alike consist of general legal norms, ³⁷ namely legal rules and legal principles. ³⁸ The distinction between rules and principles is an important one. ³⁹ Rules are more conclusive than principles. If the conditions for the application of a rule are met, the vast majority of the time a specified legal result will follow (in the absence of another, more important, rule). ⁴⁰

Principles do not operate in the same way as rules. Principles are less conclusive in nature and do not necessarily set out conditions for when they apply. Rather, principles provide a reason for making a legal determination in a certain way, without requiring it. ⁴¹ As Raz has stated, ‘[r]ules prescribe relatively specific acts; principles prescribe highly unspecific actions’. ⁴² I believe that principles can be best explained as being requirements of fairness or some other dimension of morality, ⁴³ although not all commentators would necessarily agree. ⁴⁴ Fortunately, little turns on this point for our purposes, since what is important for us is to acknowledge that international law is made up of both rules and principles, and that each are binding on legal decision-makers.

Those principles which, along with rules, make up the law do not combine to produce a beautiful harmony in the sense of giving coherent reasons to reach the same legal outcome. In fact, principles can be, and often are, seemingly at odds with one another. ⁴⁵ For instance, the principles which make up the doctrine of stare decisis may militate in favour of adhering to precedent, while the principle of the proper administration of justice, depending on the particular case, may indicate to the judge that the best interpretation of the law is to depart from that precedent.

This cacophony of principles does not mean that either one is invalid. Rather, the legal decision-maker must decide which legal determination is required in the circumstances through the consideration of the relative weight of these principles. The question of weight (importance, in other words) is another significant difference between rules and principles. When two rules conflict, this conflict must be resolved head-on since rules provide near-conclusive reasons for a legal decision-maker to determine a certain result. ⁴⁶ This conflict can be managed through further rules that have been made specifically to deal with such conflicts, or through the use of relevant principles. ⁴⁷ When two principles are potentially applicable in a case, a judge must consider the relative weight of each principle, applying the weightier in the circumstances. ⁴⁸

Next, let us consider the doctrine of stare decisis more specifically, in light of what we have just considered regarding rules and principles. I have already repeatedly referred to stare decisis as a doctrine. Stare decisis is in fact a judge-made doctrine, developed to be applied to certain parts of the adjudicative process. A doctrine is a set of rules or principles which together make up a general legal proposition. ⁴⁹ The doctrine of stare decisis is no different; it is a collection of rules and principles. In order to elaborate, it is necessary to break down the doctrine of stare decisis into its constituent parts, namely vertical and horizontal stare decisis. ⁵⁰

Vertical stare decisis is generally considered as requiring that a lower court follow the legal determinations of hierarchically superior courts. ⁵¹ This is what most lawyers will naturally think of when stare decisis is mentioned. This form of stare decisis is in many ways less controversial – and its benefits for a system of courts in ensuring predictability, facilitating judicial co-ordination and, in theory, privileging the greater expertise of judges of the highest courts are not hard to discern. Vertical stare decisis – where it operates – is a rule due to its near-conclusive nature which requires a particular result in this case, adherence to, or distinguishing of, precedent. Lower courts bound by the rule of vertical stare decisis have no ability to depart from the determination of a higher court, even if they believe the determination to be incorrect. The most that they can seek to do is to somehow distinguish the present case from the previous one on the law or the facts, a task the difficulty of which depends greatly on the particular circumstances of the case. ⁵²

Horizontal stare decisis, on the other hand, refers to the obligation of a court to consider its own previous pronouncements. Horizontal stare decisis is more complex, with its operation dependent on both the type of court and the type of legal issue in question. Horizontal stare decisis is virtually non-existent in the practice of hierarchically inferior courts, but of much greater significance in courts of last resort, such as the ICJ. ⁵³ Horizontal stare decisis, is not a rule but a principle, ⁵⁴ since it does not obligate a judge to adhere to precedent, but rather creates a presumption of adhering to precedent, where appropriate. ⁵⁵ Crucially, however, the principle of horizontal stare decisis and the reasons it provides are just one of a range of principles that a legal decision-maker must consider. And in fact, these other principles may militate in favour of departing from precedent, for instance, if such precedent was incorrectly decided or is no longer legally desirable. It is the task of the legal decision-maker to consider which of these principles is weightier in the circumstances of the case at hand, and consequently whether a precedent should be adhered to, distinguished, or departed from. Again, greater detail on just how the weight of these particular principles should be determined in the case of the ICJ is provided in the following section.

In short, both the rule of vertical stare decisis and the principle of horizontal stare decisis together make up the doctrine of stare decisis. This clearly illustrates the point made at the beginning of this section: while a seemingly straightforward idea, in practice the operation of the doctrine of stare decisis is highly context-specific. But what role has this doctrine played in the practice of the ICJ to date?

3.2 Stare decisis before the ICJ: Practice

Empirical studies of the Court’s practice have shown that since the days of the PCIJ both parties and the Court have made reference to and sought to rely on the Court’s jurisprudence. ⁵⁶ This has, if anything, increased over the years, ⁵⁷ becoming ‘virtually ubiquitous in recent decades’. ⁵⁸ Such studies have prompted suggestions that the Court has become ‘increasingly common law-like’ ⁵⁹ and underline the importance of the present attempt to provide a theoretical account for the role that precedent plays. ⁶⁰

Any examination of this practice in an attempt to discern the Court’s attitude towards its ‘settled jurisprudence’ ⁶¹ in past decisions shows that, despite the absence of any relevant rule in its constitutive instruments, ⁶² the Court has been generally deferential in this regard.

As such, taking the practice of the Court in the round, I believe we can speak of a doctrine of horizontal stare decisis operating in proceedings before the Court. ⁶³ Like other courts of last resort, while the Court’s practice has been one of deference to its precedents, it has always left open the possibility of distinguishing or departing from such precedent if compelling reasons can be found. The Court’s most significant early pronouncement on the precedential status of its own judgments came in Mavrommatis. ⁶⁴ In this case the PCIJ stated that it found ‘… no reason to depart from a construction which clearly flows from the previous judgments the reasoning of which it still regards as sound’. ⁶⁵ This position was the Court’s lodestar in subsequent decades. In the absence of a compelling reason not to, the Court adhered to those precedents whose reasoning it still found persuasive. ⁶⁶ For instance, in the Croatian Genocide case in 2008 in which the Court stated that:

[t]o the extent that the decisions contain findings of law, the Court will treat them as it treats all previous decisions: that is to say that, while those decisions are in no way binding on the Court, it will not depart from its settled jurisprudence unless it finds very particular reasons to do so. ⁶⁷

As Hernandez has stated, the Court’s reference to the necessity of finding ‘very particular reasons’ to depart from an earlier body of past decisions suggests perhaps an even stricter approach than that adopted in Mavrommatis and subsequent cases. ⁶⁸ That having been said, the possibility that the Court may depart from its jurisprudence if it is presented with or finds a good enough reason to do so was explicitly mentioned by the Court.

The most recent significant, and somewhat puzzling, instance that is relevant for our purposes relates to the 2016 Marshall Islands cases in which the President of the Court suggested that certain judges may see the Court’s own judgments as more controlling than previously thought. ⁶⁹ President Abraham, who had the casting vote in light of a 7–7 split among the other judges, sought to justify his decision to uphold the preliminary objections by relying on two relatively recent cases, Georgia v. Russian Federation and Belgium v. Senegal, which he singled out as together amounting to a restrictive standard relating to the issue of the existence of a dispute. ⁷⁰ Intriguingly, the President stated clearly that despite the fact he was ‘not sure that the Court was right’ in either of these cases, the fact that the Court ruled in the way that it did in both cases whilst ‘adopting a clear and well-considered solution’ created an expectation that the Court should adhere to this position in the present case:

I nevertheless take the view that even if a judge has expressed reservations, or indeed disagreement, at the time the Court established its jurisprudence, once the Court has done so, he must consider himself to be bound by it thereafter (not legally, of course, but morally) just as much as if he had agreed with it. ⁷¹

This statement, and the commitments it contains, would have significant implications for the Court if a majority of its judges were to feel similarly, as it would effectively amount to the operation of a strong version of stare decisis of the kind to which only hierarchically inferior courts are usually subjected. ⁷² That said, we should remind ourselves that this is just one statement in one individual opinion in the context of preliminary objections in one dispute.

Taking the long view of the Court’s practice, it seems to me that a more accurate summary is that of Sir Elihu Lauterpacht, namely that precedent has played an ‘important, but not a controlling role’ ⁷³ in the Court’s jurisprudence.

3.3 Summary – precedent as important but not controlling – horizontal stare decisis as the only version of the doctrine which fits practice

The preceding analysis of stare decisis as a doctrine shows that there is more than one possible interpretation of stare decisis. The first possible interpretation is that stare decisis operates as a rule in the vertical form. In light of the exploration of the doctrine of stare decisis in the preceding section, President Abraham’s description in Marshall Islands of ICJ judges being bound to follow precedent irrespective of its merits, I would suggest is akin to vertical stare decisis, the rule which binds judges of lower courts to adhere to decisions of higher courts. However, this view of the operation of the doctrine of stare decisis before the Court is wrong for at least two reasons.

First, since the ICJ is a court of last resort, the rule of vertical stare decisis does not fit the ICJ’s institutional practice. ICJ judges are not bound by a rule which obligates them to follow precedent irrespective of its merits, since there is no relevant hierarchically superior court nor other rule of international law which makes this so. Second, in practice this is not the role that precedent has played in the Court’s determinations. While the Court has always been deferential to its own jurisprudence, it has often mentioned the possibility of departing from its previous findings, if it has compelling reasons to do so. And indeed, this is the case for other courts of last resort which apply a doctrine of horizontal, as opposed to vertical stare decisis – all have the possibility of departing even if their general practice is not to do so, except very rarely. ⁷⁴ Accordingly, stare decisis in its vertical form fails to capture how the doctrine of stare decisis both has traditionally operated, and how it should operate, before the ICJ. As such, we can exclude it from consideration in terms of being an appropriate fit.

The principle of horizontal stare decisis, on the other hand, is more appropriate. That is, the institutional history of the Court shows that its judges take the Court’s precedent into consideration, but are not bound to follow such precedent irrespective of its merit. However, this first step of assessing fit still leaves us with more than one potentially appropriate interpretation of the doctrine of stare decisis, given that we must also recognize that there are different forms of horizontal stare decisis. These include what are usually referred to as weak and strong versions, which both conceivably fit practice of a Court of last resort.

An example may be helpful in getting to grips with different versions of horizontal stare decisis. The Supreme Court of the United States (SCOTUS) recognizes three broad categories of cases in relation to which the strength of horizontal precedent will be more or less significant. For instance, so-called ‘statutory precedents’, in other words precedent of this court relating to an interpretation of legislation passed by the United States Congress, will have very strong effect, making it very difficult for the court to depart from such precedent. In contrast, ‘constitutional precedents’, namely those relating to SCOTUS’s previous interpretations of the US constitution are the easiest to depart from. ‘Common law precedents’, those precedents relating to precedent rooted not in legislation nor the constitution but rather in the common law, lie somewhere in the middle. ⁷⁵

This raises the question of what the rationale is for the doctrine of horizontal stare decisis applying in a stronger fashion in constitutional cases, making it harder for the Supreme Court to depart from such precedent. The answer to this important question lies in both and pragmatism the architecture of the US legal order itself. While the executive can ordinarily pass legislation to correct any judicial interpretation of its statutes if it wishes to do so, the executive’s ability to amend the constitution is often a much more onerous task. As such, were SCOTUS’ constitutional precedents not to be accorded less weight, the court would have an oversized role to play in the domestic legal system on constitutional matters. ⁷⁶

Could we ever envisage the operation of stronger and weaker versions of horizontal stare decisis depending on the specific type of case before the court (statutory, constitutional or common law) in the context of the ICJ? I would venture that the significant differences between the domestic and international legal orders most likely preclude a neat division of cases into statutory, common law and constitutional. Perhaps one could draw an analogy between treaties and statutes, but the absence of a centralized legislature stretches this analogy to (and most likely beyond) breaking point. Similarly, how are we to categorize customary international law? Perhaps as the common law, but again such an analogy is too tenuous to retain any real value.

Nevertheless, I would argue that we can in fact speak of different versions of horizontal stare decisis potentially available in the context of the ICJ. These versions simply entail that the ICJ’s precedents are accorded greater weight in accordance with a ‘strong’ version of stare decisis, whereas they are be accorded lesser weight in relation to a ‘weak’ version of horizontal stare decisis. This distinction is based not on the type of case, determined a priori against some sort of taxonomy, but rather depends on the ICJ judge’s own understanding of the weight that the Court’s own precedents must be accorded in the course of the adjudicative process.

Drawing on the exposition of the doctrine of stare decisis above, including the different notions of the weight that should be accorded to precedent between the Court’s statement in Mavrommatis and President Abraham’s more recent remarks in Marshall Islands, we note that there is more than one possible interpretation of horizontal stare decisis which potentially fits the Court’s institutional practice. Accordingly, it is necessary to turn to the second step of the interpretive method, namely justification. More specifically, we must consider the question of whether the justification of horizontal stare decisis entails that a strong or weak version of this doctrine in the context of the ICJ.

4.

The ‘justification’ bit: Considering stare decisis in its best light

Having identified a number of versions of stare decisis which fit the institutional history of that practice already identified, readers familiar with the interpretive method will no doubt expect that I will now advance one version which is justified as the version of the doctrine which best reflects the value of stare decisis. ⁷⁷ More precisely, the next step in this method would involve me showing how a weak version of horizontal stare decisis is justified as the version of the doctrine which best reflects the value of contextual justice, which I argue is the relevant normative yardstick.

However, it is here that I believe I have to (again) slightly modify this method in order to account for the fact that its straightforward application would lead to a rather abbreviated conclusion that would not allow us to consider many of the values commonly advanced as justifying a strong version of horizontal stare decisis that the reader no doubt expects us to consider having persevered this far. As such, I will first start by examining these values to show that they are not capable of providing the basis for justifying the practice of horizontal stare decisis in the context of the ICJ. It is only after doing so that I will then turn to arguing that the value that stare decisis strives to achieve is contextual justice, and that accordingly only a weak version of horizontal decisis can be said to both fit and justify this doctrine.

Two clarificatory remarks are required before delving into considering particular values. First of all, for the purposes of the present article and the normative method deployed, values are moral concepts that represent goals to be promoted. ⁷⁸ Second, it is necessary to recognize what can broadly be termed value scepticism, including the charge of legal realists that the values discussed below lend themselves to more than one legitimate interpretation or application, or are inherently indeterminate, or the critique of values offered by critical legal scholars. To elaborate, a typical legal realist argument may be that there is no necessary logical connection between abstract values and the institutional doctrine of horizontal stare decisis, that all justificatory arguments are dualist, and as such we cannot assume that the latter means the practical realization of the former. ⁷⁹ A more critical argument would build on the realist challenge, and also argue that the purportedly essentially contestable nature of values undermines the type of interpretive method employed in this article. ⁸⁰

Undoubtedly the most influential example of the critical challenge to international law is that of Martti Koskenniemi, who famously attacked international law’s claims to objectivity and normativity, or, more precisely, claims that these ‘cancel each other’. ⁸¹ Central to this argument is the proposition that international law derives its normativity from evaluative judgements, and that evaluative judgements are not capable of being objectively true or false. ⁸² It is not the aim of this article to contribute to the debate on value scepticism which has played out at length in philosophy and, to a lesser extent, international legal scholarship. For the purposes of this article it suffices to note that value scepticism has been met head on by philosophers such as Donald Davidson who have emphasized the importance of intersubjectivity in showing that truth-value can be attributed towards evaluative judgements. ⁸³ As Emmanuel Voyiakis has stated, value scepticism fails to account for the fact that ‘the concept of objective truth is essential to our ability to have evaluative thoughts and beliefs and to interpret the thoughts and beliefs of others’. ⁸⁴

Of course, recognizing the objective nature of values is not the same as saying that disagreements about values are easy to define or to resolve. ⁸⁵ What it does mean is that we must adopt an ‘open, critical and charitable conversational ethic’ when engaging with values. ⁸⁶ This is crucial, as Luke MacInness notes, since:

We develop a conception of a value by imputing other values and purposes to the practices of argument and thought in which the value under inspection figures. These other values give point and intelligibility to our use of the value under study and enable us to draw conclusions about its concrete requirements by identifying the best realization of those purposes in particular cases. ⁸⁷

It is for these reasons and in this spirit that I now turn to examining those specific values commonly cited as those which horizontal stare decisis strives to achieve. ⁸⁸ And, as Waldron notes, it is not difficult to find candidates:

Our jurisprudence is cluttered with a haphazard variety of considerations adduced to justify stare decisis. They include the importance of stability, respect for established expectations, decisional efficiency, the orderly development of the law, Burkean deference to ancestral wisdom, formal or comparative justice, fairness, community, integrity, the moral importance of treating like cases alike, and the political desirability of disciplining our judges and reducing any opportunity for judicial activism. The justification of stare decisis is a field to which many contributions have been made, but to which little system has been brought. ⁸⁹

Since space limits consideration of every conceivable value, for the purposes of the following section I will focus on those advanced most often in the literature: consistency, efficiency and predictability. Ultimately, while I acknowledge that such values are relevant to some extent in determining the gravitational pull of a particular precedent, none of these in and of themselves can be said to be the value that horizontal stare decisis strives to achieve in the context of the ICJ. Rather, I argue that it is another value, contextual justice, which is the relevant normative yardstick against which versions of this doctrine can be justified, a claim which I elaborate in the final section.

4.1 Consistency

Consistency is often cited as a value providing the basis for justifying the operation of the doctrine of stare decisis, ⁹⁰ including in the context of the ICJ. ⁹¹ To elaborate, the achievement of a consistent line of cases is portrayed as justifying an interpretation of this doctrine which entails adherence to precedent, irrespective of whether or not this achieves a just result in a particular case. This value is sometimes also discussed using the language of equality, in other words that treating two like cases differently would be to fail to treat the parties to these cases equally. ⁹² Consistency and equality will be considered together in what follows.

Perhaps the most famous appeal to consistency in this context is that of Dworkin in the form of law as integrity. ⁹³ According to Dworkin, integrity is the value of the practice of law which requires its interpretation in its most coherent light. ⁹⁴ According to him, it is the role of the judge to:

so far as this is possible … treat our present system of public standards as expressing and respecting a coherent set of principles, and, to that end … interpret these standards to find implicit standards between and beneath the explicit ones. ⁹⁵

It is these interpreted standards which the judge must follow in order to arrive at the ‘right answer’ in a particular case. ⁹⁶ Judicial decisions form part of the body of law that the judge discovers as part of their interpretive exercise, and the principles which are discerned from cases and the law more generally are then binding on the judge who must adhere to those principles in order to ensure consistency. This focus on consistency in legislative and judicial decision-making in Dworkin’s work is driven by the value of integrity that lies at the heart of Dworkin’s work. ⁹⁷

Accordingly, law as integrity is primarily concerned with the consistency of decision-making in accordance with the principles applicable within a particular legal context, ⁹⁸ rather than with the justice of the outcome of such decision-making. Law as integrity, as Dworkin sees it, is designed to ensure not only fairness in treating like cases alike, but also that government and its courts ‘speak with one voice’ in applying principles within a legal system. ⁹⁹

However, as Peters has shown, the evil that law as integrity was designed to address, namely inconsistency, does not transpose particularly well from the legislative to the adjudicative context. In short, while the example that Dworkin cites as motivating the drive for consistency, namely so-called checkerboard statutes (such as, for instance, a statute which would make abortion legal only for women born in even-numbered years), does strike us as problematic, the same is not necessarily true for case law. This is due to the fact legislatures and courts occupy significantly different positions within a legal order. We may reproach legislatures for acting inconsistently in accordance with principles, as legislatures are much less constrained in their ability to make law than courts are. ¹⁰⁰

Our expectations of courts are different from our expectations of legislatures, in that we are more ready to accept inconsistent decision-making from courts, so long as the reason for that inconsistency is an attempt by that court to reach a just result. It is more important for us that courts, especially courts of last resort, strive for justice rather than consistency. ¹⁰¹ For example, President Abraham in Marshall Islands appears to suggest that consistency with the Court’s previous decisions (of Georgia v. Russia and Belgium v. Senegal) should take priority over what the individual judge feels to be the better interpretation of the law (suggesting as he did that ‘even if a judge has expressed reservations, or indeed disagreement, at the time the Court established its jurisprudence, once the Court has done so, he must consider himself to be bound by it thereafter … just as much as if he had agreed with it’). ¹⁰²

However, as Judge Tanaka stated in Barcelona Traction, ‘the requirement of the consistency of jurisprudence is never absolute. It cannot be maintained at the sacrifice of the requirements of justice …’. ¹⁰³ Of course, it is important to stress that none of this is to say that a court should not strive for consistency in its legal decision-making. On the contrary, courts should take into account the values of efficiency, predictability, and so on, and together these may amount to a presumption of consistency with previous decisions. Crucially, however, we cannot say that consistency alone is the value capable of justifying horizontal stare decisis in the context of the ICJ. ¹⁰⁴

4.2 Efficiency

Efficiency is also cited as value capable of being the basis of a justification of horizontal stare decisis. It is a common refrain that without precedent being accorded significant (if not controlling) weight, adjudication would become much less efficient. ¹⁰⁵ According to such arguments, the version of the doctrine which facilitates efficient disposal of cases (most often the strong version) is that which is justified, regardless of whether this produces a just outcome in the case at hand.

Again, context is crucially important when considering such arguments which elevate efficiency to be the value that shows the doctrine in its best light. While efficiency may be the basis for a justification of a strong version of stare decisis in the context of other courts situated within a hierarchy beneath superior courts, or courts with very high caseloads, it cannot be the only value which justifies this doctrine in the context of the ICJ. Given the relatively modest number of cases that the ICJ considers each year, the argument that it is more efficient for the Court to adhere to its previous judgments without considering their merits is simply too hard to sustain.

Were the Court perhaps suddenly to find itself with a large number of cases involving one particular issue of international law, it may be that efficiency is the value which contributes to justifying a strong version of horizontal stare decisis, but the practice of the Court to date suggests that this is unlikely to happen. Even those land and maritime boundary delimitation cases that are the Court’s bread and butter do not so frequently come before the Court so as mean that efficiency is the value that justifies the doctrine of stare decisis. Given the modest number of cases the Court handles each year and the nature of its limited jurisdiction, it can in fact be decades before the Court is asked to deal with a substantive legal question that it has previously determined. ¹⁰⁶

Furthermore, stare decisis may conserve some judicial resources in ensuring that not every legal question has to be reopened and reconsidered on its merits in every single case, particularly in cases of vertical stare decisis. However, it is important to recognize that stare decisis also consumes judicial resources in a way that is particularly relevant in the context of the ICJ. How is this so? Well, precedent on a particular issue is routinely cited by parties in support of their submissions in the course of proceedings. As such, they also have to be considered by the Court, and either adhered to, distinguished or departed from. As Hershowitz has stated, ‘ … [a]n armchair investigation suggests that opinions do not grow shorter as the volume of precedent grows larger’ and this is certainly the case with the ICJ when one looks at the average length of judgments which has ballooned since its formation, due in no small part to the Court’s apparently compulsive need to cite relevant authority for each and every assertion it makes. ¹⁰⁷ In fact, the time and the energy parties and the court put into researching, arguing and distinguishing previous ICJ cases makes us wonder whether ‘all this energy might have been better devoted to considering the just or efficient settlement of the dispute … on its merits’, ¹⁰⁸ a point to which I will return momentarily.

But even if we were to accept that adhering to precedent would be an efficient thing for ICJ judges to do as it avoids judges continuously addressing questions already settled by the Court, the question remains as to whether efficiency as a value can justify adhering to precedent irrespective of its merit. To that end, it is important to note that ‘whether or not such a practice is efficient depends critically on the rules that the practice entrenches and on the particulars of the behaviour those rules govern’. ¹⁰⁹ For these reasons, I think it is clear that while efficiency may contribute to justifying a certain strong version of stare decisis, which accords precedents near-controlling effect in a different context, it is not a value that in and of itself justifies such a doctrine in the context of the ICJ.

4.3 Predictability

Predictability, as a candidate value capable of being the basis of a justification of a version of stare decisis, is inappropriate for many of the same reasons efficiency is. ¹¹⁰ A common argument in support of the doctrine of horizontal stare decisis is that, without it, judicial decision-making would be unpredictable or even arbitrary, undermining certainty and the utility of the law as a whole. ¹¹¹ Of course any judge would admit that they must take into account the value of predictability, seeing as all things considered it is a desirable thing that a court makes relatively predictable legal determinations in order that parties can attempt to plan their lives knowing more or less the extent of their legal rights and obligations.

However, again, predictability as a value cannot, in and of itself, justify horizontal stare decisis since, in the context of the ICJ, it is not the case that the value of predictability, or the expectations of the parties, are more important than whether or not the Court reaches a just result. ¹¹² Predictability is not an end in itself but rather that which is crucial in achieving an end. ¹¹³ In other words, ‘predictability needs to be weighed against the moral desirability of the law in question’, ¹¹⁴ and this is particularly the case when we speak of the ICJ. The ICJ is a court with an outsized level of importance within the international legal order, if only in the sense that its pronouncements on the law are perhaps the most authoritative statements available to us. I believe that it would simply not do to have the Court operate a doctrine which strives to achieve a value that privileges a predictable result as opposed to a just one.

5.

Justice according to law: A weak version of horizontal stare decisis as that which both fits and justifies the relevant practice

Our journey began with the banal fact that the Court pays great respect to its own jurisprudence and a degree of curiosity as to how to explain this fact in the absence of a rule of stare decisis in the Court’s constitutive instruments. In the intervening pages I have proposed a novel theoretical framework for understanding the role that precedent has traditionally played in proceedings before the Court, and that which it should play in the future. This theoretical framework rests on the commitment that the meaning of any object of interpretation is not determined by any fixed process or criteria, but rather on the values and principles which best justify the practice related to that concept. Adopting an interpretative attitude, I have shown that the best interpretation of stare decisis is that which both fits its past institutional practice, and that which is justified by the value that the doctrine strives to achieve. Noting that a number of different versions of stare decisis potentially fit the Court’s institutional practice, the best interpretation ultimately turns on that which is best justified by the value the doctrine strives to achieve. Accordingly, it became necessary to consider what exactly the value of the doctrine of stare decisis is.

Having considered a number of candidate values in turn, it was shown that each was inappropriate for a range of different reasons related to the Court’s institutional role. The Court is not so inundated with cases so as to adhere to its past judgments in the interests of efficiency. Nor, given the particular nature of the World Court as the principal judicial organ of the United Nations, a tribunal with general subject-matter jurisdiction, and one whose legal determinations are considered to be so authoritative by all other international legal actors, can we seriously say that it is more important that the Court is consistent or predictable rather than right, (or, in the famous words of Brandeis that ‘it is more important that the applicable rule of law be settled than that it be settled right’. ¹¹⁵ )

The only value, I argue, that we can say the doctrine of horizontal stare decisis strives for is what Koskenniemi has termed ‘contextual justice’; ¹¹⁶ arriving at the most just solution in the particular dispute all things considered. ¹¹⁷ In practical terms this means that it is for the Court’s judges to ‘isolate the issues which are significant in conflict, assess them with an impartial mind and offer a solution which seems best to fulfil the demands of the critical programme’. ¹¹⁸ With this value as the relevant normative yardstick, and a number of versions of the doctrine of horizontal stare decisis which ‘fit’ the Court’s institutional history, we must ask ourselves which of these versions is justified as the version which best reflects the value of justice. I argue that this is the weak version due to the fact that, in the context of the ICJ, this is the version of this doctrine which best enables ICJ judges to arrive at what they consider to be the just outcome all things considered. In this sense the ICJ’s jurisprudence has limited constraining effect, and instead represents an authoritative statement of the Court’s previous findings and a subsidiary means for the determination of rules of law in the sense of Article 38(1)(d) of the Court’s Statute.

The strong version, in contrast, whereby ICJ judges are bound to accord significant weight to its precedents better justifies stare decisis in contexts where this doctrine strives to achieve other values, such as efficiency or consistency. In the context of the ICJ, however, such values are less important than achieving a just result, and accordingly the version of stare decisis which best justifies this doctrine is that which gives ICJ judges the greatest discretion to achieve what it considers to be the best result – namely the weak version. It is this version of the doctrine which not only prevents the entrenching of any sort of version whereby the Court would be ‘bound’ by its previous decisions, but also allows the Court to draw on its own pronouncements as a repository or material source of rules of international law.

Such a conclusion may raise concerns regarding the Court’s judges being allowed undue discretion. However, I believe such fears can be allayed. The adoption of an interpretive attitude means that the range of possible interpretations of stare decisis, respecting past judgments in our case, is circumscribed by the Court’s past practice, and limited to the achievement of those values that are the point of the doctrine. In any case, it is not unusual for courts of last resort to be entrusted with a level of discretion. Courts are necessary precisely because law-makers, both domestic and international, cannot envisage every eventually to which law may be applied, and it is the highest courts to whom we accord the greatest degree of discretion in interpreting the law.

For instance, the 1966 Practice Statement of the then House of Lords (today the Supreme Court of the United Kingdom) set out that precedents are normally to be adhered to apart from ‘when it appears right to do so’. ¹¹⁹ This is an example of what Lewis has called a delegation of discretion to courts to ‘balance tailored justice with stability and reliability, often leaning towards the latter’. ¹²⁰ This is an important point. The UK Supreme Court’s theoretical ability to depart from its own jurisprudence when it thinks it right to do so has not translated into it often doing so. And likewise, the open admission that the value which stare decisis strives to achieve in the context of the ICJ is justice, and that the ICJ can depart from its own jurisprudence when it thinks it is right to do so, does not mean that the Court will become suddenly capricious, departing from its judgments willy-nilly.

A range of factors mitigate against this, in fact, perhaps most importantly the fact the Court operates under significant pressure to do justice according to the law, and in the international community ‘the watchdogs of the public interest are continually alert to yap at their heels if they appear to do any other thing’. ¹²¹ Yet one can imagine that regarding the prospect of a suddenly unpredictable ICJ the watchdogs will likely be often idle, as it is hard to imagine a court more acutely aware of the precariousness of its own position as the ICJ, having a century of experience of trying to do justice according to law, without alienating those states whose consent the Court’s own jurisdiction is ultimately based. These are reflections offered in the hope of calming the nerves of those who will inevitably worry that explicitly sanctioning judges, to use precedent in the way which best ensures contextual justice, is tantamount to condoning naked judicial discretion. I think that such fears are largely unfounded for the reasons set out above.

There is, however, one final constituency that I would like to address directly, namely those who may feel that the thrust of my argument is somehow not radical enough. Those of such a persuasion I would hope, however, can nevertheless recognize the value in showing how the Court’s well-known practice can be grounded in, and justified by, international law. Not only that, I hope that the present contribution will also be recognized as a plea to resist any suggestion that the Court’s own precedents should be seen as controlling, rather than merely authoritative. In that regard my argument seeks to act like a lighthouse lighting the Court’s path away from the rocks should it seem to be skirting too close to them.

Of course, one may disagree with the argument that contextual justice is the value which best justifies the doctrine under consideration, but if that is so, then it falls to that person to show why that is the case and advance a more convincing argument. Such authors will at least have the present novel theoretical framework with which to disagree or to improve upon. Even then we will have the advantage of meeting on a level argumentative playing field, to discuss the values that best justify stare decisis.