Getting ‘real’ about ISDS reform: a critical realist view of international investment law’s status quo

Abstract

Investor–State dispute settlement (ISDS) is the topic of a concerted reform effort by actors in International Investment Law (IIL). The proposals proffered by United Nations Commission on International Trade Law’s Working Group III have centred around rectifying procedural complaints like the inconsistency in arbitral case law. Little attention has been given to more paradigmatic ideas such as abandoning international arbitration altogether. This article argues that the absence of any meaningful consideration for a radical departure from the current status quo is largely the effect of epistemological privilege. I adopt the philosophical framework of Critical Realism to argue that much of the work towards reforming the regime is grounded in empirical epistemologies that do not consider the deeper, more complex issues that affect IIL’s operation. I make this argument through an examination of behavioural economics’ integration into scholarship and the positivist basis of the reform proposal for an advisory centre for international investment law.

INTRODUCTION

After years of contestation over its legitimacy, the regime for the promotion and protection of foreign investment, International Investment Law (IIL), is actively reflecting on how to best reform its operation.¹ For instance, the United Nations Commission on International Trade Law’s (UNCITRAL) Working Group III (WGIII) has convened since 2017 to focus on reforming IIL’s most contentious element, Investor–State Dispute Settlement (ISDS).² ISDS has been linked to issues such as: creating a democratic deficit within states by moving certain domestic policy questions of public interest to international tribunals, blocking the implementation of domestic environmental regulations and contributing to widening wealth inequality by enforcing awards that can amount to billions of US dollars on developing States.³ The WGIII has discussed various proposals at differing ‘incrementalist’ and ‘systemic’ levels of reform to attenuate some of these perceived faults.⁴ ‘Incrementalist’ proposals, such as a code of conduct, and ‘systematic’ proposals, such as the establishment of a multilateral investment court, have been proffered in an attempt to rectify procedural issues such as the costs and duration of arbitration as well as inconsistent decisions identified in the case law. However, ‘paradigmatic’ reform, the view that the regime is ‘irrevocably flawed’ as is and ‘in need of whole-sale replacement’, has been widely ignored at the policy level.⁵ This is despite this sentiment being evident in certain Global South States’ denouncement of central dispute forums and others’ discontinuation of their International Investment Agreements (IIA).⁶

This article argues that ‘paradigmatic’ reform like the abandonment of international arbitration is largely out of reach, both in practice and in theory, for most of IIL’s primary, immediate participants (States, practitioners, and academics). I premise this argument on the idea that, for a radical turn of events like ‘whole-sale replacement’ to occur, there must be a corresponding ability to envision said turn of events. The legal imagination to do so is largely missing in the conversation about IIL reform. This article contends that the lack of imagination is, at least in part, an epistemological problem. I adopt the framework of critical realism to argue that such an imaginary is missing in part due to the dominant epistemologies in the regime, ideas primarily based in contemporary economic theory.⁷ Critical realism is a field of philosophical enquiry concerned with the underlying causal structures and generative mechanisms of social ontology. It provides a critique of social methods that remain at the level of empiricism—a charge it makes against much of economic theory.⁸ I extrapolate critical realism’s critique of economic theory as a framework for analysing both IIL’s dominant empirical epistemologies and the WGIII’s reform efforts in the section ‘A critical realist framework for a critique of international investment law reform’.

In the section ‘Epistemological positivism in international investment law scholarship’, I focus this article’s epistemological critique on the integration of behavioural economics in IIL scholarship, particularly its concept of bounded rationality as it is understood by Lauge N Skovgaard Poulsen, and its reliance on empiricism, which has found wide appeal in IIL literature.⁹ Behavioural economics is a field of thought which posits that the rational decision-maker of traditional economic theory is constrained by its complex environment and own limited (often cognitive) capabilities.¹⁰ Behavioural economics and its corresponding methodologies can be understood as an evolution, perhaps even a corrective, of more traditional forms of economic and rational choice theory. This article focuses on behavioural economics because of the impact it has had on IIL scholarship, particularly regarding topics such as developing states’ engagement with the regime. While there is still mainstream IIL literature that adopts neoclassical theoretical bases about the regime’s operation, such analyses lost a certain amount of credibility within IIL scholarship once research showed the inconclusiveness of IIL’s impact on incentivising foreign direct investment.¹¹ These theories became further contested as the regime became increasingly susceptible to criticisms about ISDS’s asymmetrical treatment of developing states over the course of its history.¹² Behavioural economics gained scholarly attention because it offered an explanation as to the discrepancy within the regime between its rational analysis and developing states’ experience in ISDS based upon these certain actors’ cognitive limitations.

The framework of critical realism reveals how, despite its attempt to take into fuller account the different environmental and psychological factors involved in political, legal and economic decision-making within the operation of IIL, behavioural economic perspectives still accept the broader institutional parameters of the regime. That is, behavioural economic perspectives largely accept a globalized capitalist world economy in which States must rely on foreign capital to pursue their political economic objectives, and therefore, there is need for a legal regime tasked with maintaining accountability and efficiency between these foreign capitalists and capital-importing States. In this way, critical realism shows how IIL’s legitimacy problems are perhaps not really about cognitive limitations in states’ behaviour but deeper, less empirically verifiable, structural issues concerning ISDS’s place in the globalized capitalist system. Critical realism reveals how such an epistemological basis as provided by behavioural economics does not provide the imaginary space necessary for paradigmatic reform because such epistemologies are incapable of contemplating the empirically less verifiable social aspects involved in IIL.

In the section ‘The positivism of the proposed advisory centre’, I apply the critical realist critique made against behavioural economics in the section ‘Epistemological positivism in international investment law scholarship’ to the WGIII’s proposal for a World Trade Organization (WTO)-styled advisory centre for IIL. The advisory centre is meant to form part of a protocol or annex of the larger multilateral investment court—one of the WGIII’s major ‘systemic’ reform proposals.¹³ I focus on the proposed advisory centre because it is informed by and premised on the same epistemological grounds as behavioural economics: that the problems encountered by decision-makers (States), beyond the financial burden of arbitration, are largely a matter of missing information from complex environments and subjects’ limited capabilities.¹⁴ Therefore, by providing better institutional support in arbitration, the advisory centre is meant to attenuate some of the contentions developing States have with ISDS. However, by accepting the same epistemological basis as behavioural economics, the proponents of the advisory centre remain within the same theoretical and institutional parameters of the regime concerning the global political economy and international arbitration’s place therein. These epistemologies share an inability to conceive of a broader social, political, economic, and legal system without the need of international arbitration, or IIL more broadly. Consequently, paradigmatic reform remains inaccessible.

This article makes three interrelated contributions. First, it demonstrates how the discourse surrounding ISDS reform is constrained by its inability to envision further-reaching, paradigmatic reform. This is necessary because, though there has been much critical scholarship about the current status and likely future of the regime, this scholarship has largely taken a back seat to more mainstream literature which is preoccupied with providing technical/procedural fixes to IIL’s operation.¹⁵ The mainstream literature comes with its own assumptions about what justice entails in the relationship between sovereign States, foreign investors, and a globalized economy. Critical realism exposes how these notions of justice can be epistemologically limited in their breadth of analysis. It demonstrates how privileging certain epistemologies ultimately only enables incremental, possibly systemic, change while constraining a fuller conversation about the paradigmatic change that much of the more critical literature argues would be necessary for a more wholistic form of justice in the relationship between States, investors, and the global economy.

Secondly, this article makes this argument by introducing and explicating a novel ontological/epistemological theoretical basis for the critique of much of this mainstream scholarship in the form of critical realism.¹⁶ Critical realism serves as an alternative to the growing trend of international economic law literature utilizing empirical, largely economic, methodologies.¹⁷ The more quantitative methods in this scholarship have already been critiqued for how they limit consideration for less empirically verifiable aspects of the regime such as ‘relevance, validity, reliability, honesty and transparency’ in favour of scientific, empirically verifiable analyses.¹⁸ Critical realism builds upon such work by providing a critical theoretical basis which both critiques the epistemological dominance of empiricism and emphasizes the ‘complex, stratified, contingent and therefore open-ended’ ontology of the social world.¹⁹

Thirdly, critiquing the epistemological basis of the discourse about IIL’s reform provides more space for radical considerations about the future of the regime. Given IIL’s role in exacerbating the twin crises of global environmental degradation and wealth inequality, abandoning international arbitration may be a necessary step in recalibrating global governance over the distribution of resources and wealth.²⁰ Unfortunately, this article does not have the scope to provide an exposition of the broader social and structural factors that should be given more attention in mainstream IIL literature. I will therefore only be able to signpost to others’ work in these areas. Instead, this article’s main contribution is to expose and analyse the role epistemologies play in configuring the conversation about IIL reform.

A CRITICAL REALIST FRAMEWORK FOR A CRITIQUE OF INTERNATIONAL INVESTMENT LAW REFORM

Most participants in IIL (States, practitioners, scholars) recognize that there is much room for reform. Scholarship has identified multiple factors contributing to the dissatisfaction of ISDS including, to name only a few, its basis in commercial arbitration, tribunals’ history of inconsistent or even allegedly incorrect decisions, and the regulatory chill ISDS places on States’ policymaking.²¹ The difficulty, commentators like Puig and Shaffer aver, is in how to make institutional choices regarding reform.²² They concede that ‘all institutional processes are imperfect’, and therefore, what is needed is a framework for evaluating the trade-offs between the regime’s normative goals and its institutional alternatives.²³ Much of the scholarship taking up this task consists of analyses which attempt to parse out these trade-offs in an empirically verifiable manner through different methods such as behavioural economics or quantification.²⁴ The logic goes that, because empirical methods are often more concerned with providing an accurate description of the ‘hows’ and ‘whys’ of law than making normative assertions, they serve as a seemingly ideal, objective foundation for legal research to base its normative claims.²⁵ For instance, there is no denying a quantitative lack of diversity in arbitral tribunals and the normative proposal to reform the appointment procedures in international investment arbitration to provide a more diverse range of perspectives is certainly a viable and perhaps necessary action for this issue.²⁶ There are empirically valid underpinnings to many of the specific reform proposals, and, if IIL’s major features are to remain intact, systemic or incrementalist reforms such as amending the appointment procedures do likely offer up some form of redress to the challenges posed.

Be that as it may, this article is more interested in directing attention away from these immediate, practical concerns and focusing on the dominant bases of knowledge that precede normative goals and institutional choices. Specifically, I argue that certain dominant empirical epistemologies help skew such conversations towards the regime’s retrenchment and ‘maintaining its core attributes’.²⁷ The preoccupation with empirical verifiability and the privilege such methods receive effectively put blinders on the full scope of the broader political, historical, economic, and cultural—social/structural—analyses necessary for a normative evaluation of IIL’s operation.²⁸ To have a wholistic conversation about the normative goals and institutional choices of reform, these less empirically verifiable epistemologies should be given more prominence. For example, important critical work has been conducted about the regime’s colonialist origins and its scholarship’s racist tendencies.²⁹ However, such research is not as accessible to the same methods of verification as more empirically oriented methods. This should not preclude such scholarship, and the potentially more paradigmatic implications therein, from its inclusion in the deliberations concerning IIL reform.

To make this argument, I adopt critical realism’s framework for its critique of the ‘positivism’ it identifies in much of the social sciences.³⁰ Critical realism identifies positivism as a way of measuring the validity of logical and/or scientific postulations based upon the observation of empirical events that exist in causal relation to one another.³¹ In order to carry out this process, empirical methods must adopt systems of closure that define strict analytical variables and parameters. In doing so, empirical methods attempt to identify conjunctions of events (patterns) which they formulate into predictable laws,³² best reflected in the deductive explanation of events exemplified by the statement, ‘whenever event x then event y always follows’.³³ While this is a necessary process in making scientific claims in the physical sciences, when such methods are utilized in the social sciences, they limit the breadth of analyses to the level of empirical data. Economic theorists have criticized critical realism for relying too much on its critique of systems of closure and that the problems critical realism identifies about economic theory are due to a lack of understanding of economic theory’s aims.³⁴ However, such criticisms fail to acknowledge that critical realism is a philosophical critique that seeks to grapple with the more open, complex, and deeper levels of social reality that comprise our social world and that systems of closure which circumscribe social dynamics solely to the economic realm are unable to access this ‘open’ world on an ontological level.³⁵

The following argument’s focus on epistemology does not mean to reduce the material asymmetries visible in ISDS and IIL more broadly to a purely epistemological problem. Critical realism is fundamentally based upon a materialist ontology that posits there is materially generative mechanisms at work in both the physical and social realms. Indeed, IIL’s operation reflects the unequal material social structures of an asymmetric global division of labour in which certain states and corporations benefit from the economic exploitation of other states and local communities.³⁶ However, social structure’s social ontology gives it an incredibly complex relationship to epistemology. Social ontology is of a different nature than physical ontology, and, in its analysis, epistemologically, it must be treated as such. Namely, social structures are emergent from human activity and are therefore not purely reducible to naturalism in a similar way as physical structures.³⁷ Though, this does not make social structure any less objective in our experienced reality; like Bhaskar writes, ‘[these] social structures may be just as “coercive” as natural laws’.³⁸ While not meaning to elide the material wealth and power asymmetries underlying the operation of IIL, the focus on epistemology below is due to the largely epistemological character of legal argumentation and its important role in limiting the level of material engagement available to participants in the regime.

There are two related components to critical realism’s critique of positivism that I extrapolate in the following analysis: first, there is an ontological component, labelled empirical reality, in which empirical systems of closure presuppose a solely empirical ontology³⁹; and second, there is an epistemological component in which statements about the knowledge of reality get transposed into statements about reality itself, labelled the epistemic fallacy.⁴⁰

Empirical reality

To grasp the ontological component, it is first necessary to understand how critical realism stratifies reality into three distinct domains, or levels: the empirical, the actual, and the real.⁴¹ The empirical refers to the level of human’s atomistic sensory experience. It can be most easily understood as the sensory data that are collected during scientific observation. Empirical data can be the observed colour of an object, the stated sentiment of an interviewee, the collected statistical data of economic research and so on. The actual refers to the events that are apprehended through, though distinct from, experience or observation.⁴² Events are ‘instances’ which actually occur in the world, but their existence does not rely on the perception of an observing individual; they can exist independently from experience or observation.⁴³ One can think of the cliched thought experiment: if a tree falls in the forest and no one is around to hear it, does it make a sound? The answer critical realism would give is yes because the physical vibrations which cause sound would still be generated—the ‘event’ of those vibrations would occur regardless of anyone there to empirically observe them. The real is best understood as the level of reality in which causal forces or generative mechanisms are at work in both physical and social domains. The real is a philosophical category about these ‘generative mechanisms’ and therefore does not lend itself to an easy illustrative example. Rather, ‘generative mechanisms’ serve as critical realism’s theoretical catchall for the underlying physical and/or social processes that create our reality. Generative mechanisms give rise to both the actual and the empirical regardless of events’ actual manifestation or a subject’s empirical experience of them.⁴⁴ In other words, critical realism posits that there is a ‘real’ world of generative processes independent of humans’ perception/engagement; humans may perceive or experience the real, but its generative mechanisms do not rely on them doing so.

The critical realist approach to ontology contrasts with empiricism because, in empiricism, scientists must produce a closed analytical system (herein referred to interchangeably with system of closure) to collect their empirical data from events and identify generative mechanisms of nature. ‘Closed system’ refers to the artificially constructed (human-made) experimental parameters that exhibit event regularities and are necessary for the scientific method to occur.⁴⁵ An example of a system of closure from the physical sciences is the conditions of a vacuum when conducting gravitational experiments; in social or economic theory, a closed system could refer to something like rational choice models—a logical mental construct that represents a verifiable delimited parameter of analysis. In contrast to what is analysed in the empirical sciences, the real exists as an open system. Real generative mechanisms do not only exist within the closed system parameters that empirical experimentation creates to identify them. Rather, generative mechanisms are at work without them even being observed. This openness applies especially to the social realm.⁴⁶ Bhaskar states that ‘society … is necessarily “theoretical” [open] … it cannot be empirically identified independently of its effects; so that it can only be known, not shown, to exist’.⁴⁷ The limitation with much empirical work is that its analysis gets stuck in systems of closure—it accepts the empirical reality of its own analytical closed systems as reality itself.

Empirical methods, like those entering international legal scholarship through economic theory, run the risk of remaining in empirical reality when they confine themselves to the systems of closure in which they make their analysis. The institution–normative goal trade-off, which Puig and Shaffer speak, exhibits exactly this tendency. Institutions, such as legal regimes like IIL, multinational corporations, the capitalist economy, States, and more, are social constructs, meaning they are human artefacts.⁴⁸ Institutions are intrinsically artificial systems of closure. They may be knowable in their empirical effects, but they are not unalterable features of reality—institutions can change. To treat them as a fixed system of closure is falling into empirical reality because it is equating empirically manifest phenomena (human artefacts) with real generative mechanisms.

Certain aspects of reality can only be known by their effects, and these aspects’ generative social mechanisms cannot always be empirically verified. To access knowledge about these aspects of reality requires other, often more theoretical, engagement with their potential generative mechanisms. For example, the generative mechanisms of social aspects like hierarchy, power, and political economy are not fully accessible through closed system analysis, but they have been shown to play a significant role in the operation and evolution of international law. Certain scholars have linked how the foundational features of international law, such as the sovereignty of nations or the concept of civilization, have been used to limit certain States’ normative engagement with the system by creating social hierarchies of domination.⁴⁹ Similarly, others have demonstrated how international law itself is imbued with the social forms of capitalism and that we would not have the international law that we do today if it was not for this political economic struggle for power.⁵⁰ Such work all shares an open ontology in common—these authors treat institutions as historicized human artefacts. They acknowledge that the social mechanisms generating the dynamics of their analysis may not be empirically identifiable, but this does not stop such perspectives from attempting to grapple with the deeper, more complex factors giving rise to such social phenomena.

Epistemic fallacy

The second, epistemological, component of critical realism’s critique is concerned with how knowledge is treated regarding the different levels of ontology and reveals how empirical realism is largely the consequence of what it refers to as the ‘epistemic fallacy’.⁵¹ The epistemic fallacy is the mistake of transposing statements about our knowledge of reality into statements about reality itself. Again, critical realism distinguishes the real from both the actual and the empirical. If the empirical is based upon the experience of an observer, then it is a social domain between humans and its object of knowledge is transitive—it relies on the anthropocentric capacity of the observing subjects’ perception and is therefore dependent upon conceptualization and intrinsically social.⁵² In contrast, the real, as an object of knowledge, exists independent of human involvement—in this way, it is intransitive.⁵³ Generative mechanisms do not change according to human’s understanding of them but rather from their own generative processes independent of human engagement.

In the transitive work of scientific experimentation, scientists must create systems of closure to measure empirical atomistic events. Through closed system analysis, we may observe actual or even real events, but this observation should epistemologically only be treated as such—an empirical observation—because observation is intrinsically transitive. The epistemic fallacy transposes the transitive data found in empirical observation to the intransitive level of the real. It collapses the different ontological domains into one. Bhaskar writes:

As ontology cannot … be reduced to epistemology this mistake merely covers the generation of an implicit ontology based on the category of experience; and an implicit realism based on the presumed characteristics of the objects of experience, viz. atomistic events, and their relations, viz. constant conjunctions [the deductive equation].⁵⁴

The epistemic fallacy gives way to empirical realism.

The epistemic fallacy occurs when conclusions are made about a given (for our purposes, social) object or problem based upon only the empirical knowledge gained about said object or problem. Such conclusions are positivist because they give primacy to the empirical domain.⁵⁵ They do not take the complexity that exists in the social world into account. However, the real is not delimited to closed systems and atomistic events. The real is an open system—it contains generative mechanisms that transcend humans’ transitive ability to fully account for them with empirical methods. For instance, there are knowable effects of IIL which are not empirically verifiable. Authors from the Global South perspective have steadily produced work that examines IIL’s effects in this broader, often more critical albeit less empirically verifiable, social context.⁵⁶ For example, certain scholars have long demonstrated how IIL is one part of the international legal infrastructure that facilitates the global neoliberal project’s transfer of wealth from the Global South to multinational corporations in the Global North.⁵⁷ Others have demonstrated the significant and continuing influence the history of imperialism and colonialism have had on delineating the power relations between Global North and South States in the regime.⁵⁸ Still others argue that IIL creates a political economic dynamic in which States wishing to implement domestic regulatory changes do not do so out of a fear of the legal and economic repercussions from international arbitration.⁵⁹ These arguments have been proffered from a variety of different epistemological backgrounds, but they all share the view that the issues found within the operation of IIL can originate in extraordinarily complex social dynamics. While certain aspects of these dynamics like the differing levels of wealth between states or the environmental impact of industry may be empirically verifiable, aspects such as the structural motivations or causes behind these dynamics are very difficult, if not impossible, to empirically verify.

To reiterate, the epistemological problem critical realism identifies with empirical research is not about empirical methods per se or the scientific requirement of verification. Indeed, there is much to be gleaned from empirical exercises such as interviews or case studies. Moreover, certain critical authors often embrace empirical methods to strengthen their arguments.⁶⁰ Critical realism’s critique concerns the empirical research that falls into epistemological positivism. That is, critical realism critiques arguments that take empirical data (empirical reality) at face value—this is what is meant by epistemic fallacy. Actors fall into empirical reality when they mistake the empirical for the real. Recall that critical realism argues that there are real generative mechanisms that underly empirical manifestations. To acknowledge and analyse only the empirical manifestations of any given social phenomena is to disregard much of what may be giving rise to such empirical manifestations in the first place. This is the critique critical realism makes against positivist research. It is this critique that this article adopts to argue against much of the dominant, empirically focused discourse surrounding IIL’s legitimacy crisis and its corresponding conversation about reform.

Critical realism’s implications for IIL reform

Critical realism’s framework is highly abstract. Nonetheless, it has significant implications for the analysis of IIL reform. The following two sections explicate these implications from two different perspectives. The succeeding section demonstrates how scholarship can fall into empirical realism and the epistemic fallacy when it relies so prominently on knowledge about the empirical effects of the regime to make its prescriptions about reform. Certain scholarship can treat the knowledge gained from empirical experience as the ultimate version of reality rather than only an epistemological perspective of said experience.⁶¹ The next section serves as an illustration by examining certain uses of behavioural economics and its integration in IIL, especially the adoption of the concept of bounded rationality. These methods accept the closed institutional parameters of the regime without considering the broader social, historically contingent, political economic and legal context in which it operates.

Secondly, critical realism reveals how certain practice commits the epistemic fallacy and remains in empirical reality by addressing only the empirical effects of IIL.⁶² The subsequent section demonstrates this through an examination of the proposed advisory centre for IIL. Very basically: the advisory centre is predicated on the idea that the disparity between IIL’s Northern and Southern actors can be levelled by a body that provides additional personnel and expertise to States lacking in these aspects. I demonstrate that this perspective is premised on the same basic rationale that underlies behavioural economics—that subject’s actions are largely shaped and often constrained by ‘imperfect information and imperfect processing of information’.⁶³ Such a view ignores the complexity of the structural issues identified by some of the authors above.

The upshot from these two sections is that much of the reform effort by the WGIII remains at the epistemological level of empirical realism. Indeed, several of the WGIII’s academic participants are scholars actively engaged in empirical research and economic methodologies.⁶⁴ Only analysing the empirical effects of an institutional feature like international arbitration makes reform of such an institution look like a matter of technical, institutional choice—it calls for incremental, systemic at most, reform. Such a perspective does not bother with the structural, less empirically verifiable, aspects of the regime such as its colonialist past or neoliberal present—aspects which may call for more paradigmatic consideration.

EPISTEMOLOGICAL POSITIVISM IN INTERNATIONAL INVESTMENT LAW SCHOLARSHIP

International legal scholarship has adopted various concepts and methods of behavioural economics into the analysis of international law.⁶⁵ Specifically, its concept of bounded rationality has been extrapolated to examine both the perspectives of treaty design and the conditions by which States consent to treaties and international arbitration.⁶⁶ The concept has been applied rigorously to the analysis of developing States’ engagement with IIL in Lauge N Skovgaard Poulsen’s Bounded Rationality and Economic Diplomacy: The Politics of Investment Treaties in Developing Countries.⁶⁷ This section argues that the application of bounded rationality to developing States’ engagement with IIL commits the epistemic fallacy and falls into empirical realism because it both accepts the regime’s systems of closure and mistakes knowledge about IIL’s effects for the real causes of its contention. I make Bounded Rationality the focus of the following critique for purposes of space and to provide the necessary level of analysis which would not be available with a broader scope.⁶⁸ This critique should not be read as comprehensive of all behavioural economics’ integration into international law since the book does not represent all forms of the methodology. Rather, in line with the qualifications about empirical methods and positivism above, the following is a specific examination of behavioural economics’ potential shortcomings as they are found in one representative example.

In Bounded Rationality, Poulsen contests the narrative previously provided from the rationalist perspective about developing States’ participation in IIL. This narrative was largely driven by the idea that developing States’ participation was the result of a rationalist calculation for the competition of foreign capital.⁶⁹ Broadly understood, rational choice theory encompasses the analysis of the neo-classical economic concept ‘homo economicus’⁷⁰; that is, the abstract individual actor who contains perfect knowledge and foresight of its circumstances and is focused on the maximization of its own self-interested utility.⁷¹ Rational choice theory often attempts to explain homo economicus’s decision-making through predictive modelling about its choices in a way that empirically explains the (rational) actions that it (should) take to attain its given objectives.⁷² Accordingly, developing States’ engagement with IIL could be rationally understood as a move to maximize their ability to court foreign investment.⁷³

However, almost as early as rational choice theory began to develop, scholars recognized that humans rarely act as rationally as its models portray. Behavioural economics suggests that it is not always in the decision-maker’s immediate interests to devote the time and resources necessary to making an informed, rational decision.⁷⁴ Consequently, actors will often take the decision that may not lead to maximum utility but will suffice in meeting an optimal outcome. This notion is best exhibited by the concept of bounded rationality, an important feature in behavioural economics’ critique of rational choice theories.⁷⁵ Behavioural economics expands upon the notion that actors are influenced by more than just their own rational self-optimization by focusing on the ways actors’ psychology can be limited or complicated in decision-making.⁷⁶ The concept rejects rational choice theory’s traditional formulation of homo economicus and argues that actors’ rationality can be ‘bound’ by forces both internal (such as cognitive constraints) and external (such as environmental limiting factors) factors.⁷⁷ Bounded rationality posits that factors such as how actors frame their decisions, the default settings in a given decision-making context or scenario, or actors’ own biases shape to a large extent how they make decisions.⁷⁸

Poulsen extrapolates this concept to argue that it is due to bounded rationality that developing States have signed up to the treaty regime of IIL, despite its norms and procedures not always working in their best interest. Poulsen shows, through an empirical analysis of a number of case studies from Latin America, Africa, and Asia, that these States’ entry into investment treaties was often not the result of rational decision-making.⁷⁹ Rather, developing States commonly displayed a bias for default-type model Bilateral Investment Treaties (BITs) from Europe and North America and easy negotiating partners.⁸⁰ He argues that the concept of bounded rationality helps explain that these States apparently did not know what they were signing up to with IIL because, if they had, then they likely would not have ‘vastly overestimate[d] the economic benefits of BITs … [or] ignored their [sic] risks until hit by a claim themselves’.⁸¹ Accordingly, ‘[this] explains why so many were content to simply sign off on European models’.⁸²

The empirical reality of bounded rationality

Bounded Rationality remains embedded in empirical reality because it creates a closed system analysis on two levels: the theoretical and the institutional. First, Poulsen theoretically remains situated within the same rationalist system of closure that he is attempting to critique. He turns to the concept of bounded rationality because he argues that traditional rational choice models about developing States’ engagement with IIL ‘tells us little about how governments adopted the treaties’.⁸³ The concept of bounded rationality is meant to fill this gap by more accurately describing the ‘systematic information processing biases’ that cause developing States to fumble in the pursuit of their own preferences.⁸⁴ In traditional rational choice models, ‘preferences are independent from the circumstances in which they are revealed’, but this abstract treatment of choice does not account for the many different cognitive and environmental variables that can ‘bound’ what actors view as a preference.⁸⁵ The bounded rational model is meant to better reveal how the ‘framing’ around a decision or its relative reference point, in essence the information surrounding a decision about preferences, is never neutral.⁸⁶ Instead, how information is received and viewed by an actor makes all the difference in their decision-making.

However, in accounting for the cognitive impact of ‘framing’, this use of bounded rationality still relies on the original parameters of the pre-established preferences from its rational analysis. Poulsen explicitly states that ‘I follow standard applications in political science by using bounded rationality as a variant of rational choice theory studying preference-based and goal-oriented behaviour while catering for cognitive constraints’.⁸⁷ Specifically, he adopts two rationalist baselines: the ‘rational competition’ baseline, that developing States strategically sign up to BITs in competition for foreign capital; and the ‘rational learning’ baseline, that developing States are expected to learn about the risks of BITs based on observation and experience.⁸⁸ His bounded rational analysis does not remove the rationalist framework from its object of study; it only reorients its analysis to more comprehensively account for internal factors like cognition. In other words, the ‘framing’ around developing States’ decision-making in IIL is still within the overall rational theoretical framework of the regime. It remains within the closed theoretical system of rational choice.

Secondly, there is also a system of closure around the institutions in which developing States make bounded decisions in Poulsen’s analysis. He accepts and presupposes the closed historical, political, and economic systems in which States engage with IIL. For example, he holds up examples like Turkey’s entry into its first BITs to illustrate how developing States exhibited information biases in their initial engagement with the regime.⁸⁹ Poulsen highlights how Turkey’s first BIT was signed with the USA in 1985 and subsequently became its default treaty for the following negotiations with other States. He argues that Turkey displayed a motivational bias towards BITs’ ability to attract foreign investment in their continued commitment towards these treaties despite research indicating that this link was spurious. Poulsen claims that ‘The strong policy commitment to attract foreign capital meant Turkish policymakers wanted to believe BITs were crucial to attract investment, which appear to have impacted their information processing.’⁹⁰ However, while he does provide a cursory historical background of developing States like Turkey’s engagement with IIL and the regime’s emergence, he does not consider the possibility that they may have had little choice but to engage with the regime’s different institutional aspects like IIAs or international arbitration.⁹¹ By focusing only on Turkey’s own biases, he leaves the broader historical, political, and economic constraints in which the State made its bounded decisions underexamined. In doing so, he treats the emerging neoliberal doctrines of the 1980s, capital markets, free trade, and privatization, as an empirical reality.⁹²

Poulsen’s empirical realism is mistaken because it does not consider the open, historically contingent circumstances surrounding States like Turkey’s engagement with IIL.⁹³ Investment treaties, international arbitration, capital flows, and markets—all these institutional aspects of IIL are the result of historically contingent, often political, decision-making. Institutions are human artifacts; human constructs that are as transitive as the knowledge about them.⁹⁴ While States like Turkey may be subject to the rationalist variables of their specific institutional context and the corresponding cognitive constraints therein, in an open system, they are subject to far wider social structures that affect their decision-making. In other words, Turkey did not enter IIL in a political, economic, or legal vacuum which bound its rationality. Rather, Turkey’s integration into IIL coincided with an extensive international political economic transformation that drove it and similar States to transition from an economy based upon import substitution to neoliberal liberalization.⁹⁵

Specifically, Turkey’s first BITs followed a military coup and economic restructuring very similar to the ‘shock treatment’ that occurred in Chile only years prior.⁹⁶ The military took over Turkey after a failed attempt of its previous minority government to implement an IMF-backed economic package consisting of extensive trade liberalization, reduction on tariffs and implementation of export subsidies to incentivize foreign investment.⁹⁷ The military dictatorship forced through the economic restructuring and solidified its transformation by instating a puppet government to enact the reforms and implement a new constitution (while also outlawing union activity and imprisoning dissent).⁹⁸ The Turkish State was not only trying to maximize its foreign investment policies in a neutral (albeit bounded) political context. Turkey was ensconced in a far more open social system with pressures coming in the form of foreign relations, internal authoritarianism, and international economics. In an era which models like import substitution were becoming no longer politically and economically viable and powerful Global North States were projecting neoliberal policies around globalization, liberalization, and financial restructuring through international institutions such as the IMF, it is debatable if a State like Turkey had any choice in its foreign economic and legal relations at all, much less a bounded rational one.⁹⁹

The epistemic fallacy of bounded rationality

Poulsen’s work commits the epistemic fallacy because it mistakes the empirical effects of developing States decision-making for the reality of their circumstances. This is most visible in the prescriptions that he draws from his analysis. He argues that long-term strategies for developing states seeking better investment treaty strategies could include investing in ‘in-house’ IIL experts or coordinating with bureaucrats from other developing States to pool their collective experience.¹⁰⁰ In drawing such conclusions, he makes a causal inference between the empirical data he analyses in States’ actions and their causes. He commits the epistemic fallacy by equating transitive knowledge about developing States’ bounded circumstances in treaty design and negotiation to these States’ real agency in the social realm of global political economy and law.

It is not that demonstrating actors’ biases towards default positions in their decision-making is incorrect, or even that this is not a valuable insight. It is that this point does not account for the broader, more open, context in which actors take these default positions. In doing so, it obscures the open generative mechanisms which give rise to States like Turkey’s actions. Poulsen trades in a more open analysis about the structural underpinnings—the potential ‘why’s’—of the regime for one that is empirically verifiable. Consequently, his research demonstrates how Turkey was biased towards a default BIT and these legal instruments’ contested ability to attract foreign capital. However, in limiting the analysis to closed system variables and then proffering prescriptions based upon these parameters, the research is mistaking the transitive findings drawn from its examination as the reality of its object of analysis. The research does not provide the space for a deeper analysis, for example, of the historically contingent political economic circumstances (the open system factors) which might drive a State like Turkey to have these biases in the first place.

Poulsen does point out that his analysis is not concerned with whether States’ make the ‘right’ choices or if they should cancel their treaties.¹⁰¹ Rather, his objective is to demonstrate the constraints developing States experience in engaging with IIL. Nonetheless, even when Poulsen considers the implications of States’ withdrawal from the regime, it is only in terms of these States’ ability to force significant reform upon IIL, not whether the regime itself is part of the broader problem in developing States’ political, economic, or legal circumstances.¹⁰² Epistemologically, if prescriptions are going to be given about what States could do regarding their disadvantaged position in IIL, these prescriptions should come from an analysis that seeks to encompass the open social setting in which these States are disadvantaged in the first place rather than posit potential options based solely upon the findings from a closed system analysis around rationalist, though bounded, preferences. Consequently, Poulsen’s analysis immediately ensconces its critique within the same rationalist variables and institutional parameters that he is criticizing; he accepts that States’ preferences are first predicated on competition for foreign capital and that there is a rationalist learning curve to the political, economic, and legal variables that go into investment treaty design and negotiations. His framework of bounded rationality only provides a corrective to rationalist perspectives by providing a more nuanced empirical analysis about how states choose their preferences. He remains in a closed system analysis which does not account for the transitive, open nature of social practices such as law, politics, and the economy.¹⁰³

I argue that the incremental and systemic reform proposals proffered by the likes of the WGIII also remain ensconced within the institutional parameters and overall rationalist orientation of the regime. The next section demonstrates how this dynamic is at work in the reform proposal for an advisory centre for IIL. This reform proposal is predicated on the same, ultimately rationalist, formulation that developing States compete for capital and that there is a rationalist learning curve to their efficacy in doing so. This efficacy is only held back by a lack of resources and expertise. Accordingly, in line with Poulsen’s prescriptions, the proposal is for an institutional mechanism that assists such States in their engagement with the regime at both the negotiation and dispute settlement stage. As the critique of Poulsen above demonstrates, such a reform does not address the deeper, more complex issues identified by critical scholars such as the exploitative nature of globalized capitalism and the hierarchical residues of colonialism. Reforms like an advisory centre are incapable of recognizing such issues because they are only focused on addressing the empirical effects of IIL’s operation.

THE POSITIVISM OF THE PROPOSED ADVISORY CENTRE

This section argues that the WGIII’s reform proposal for a WTO-styled advisory centre falls into a similar empirical reality and commits a similar epistemic fallacy as that which was analysed in the previous section about behavioural economics’ integration into IIL scholarship. The proposal is based largely upon the idea that developing States’ negative experiences in international arbitration are the result of their having ‘neither the experienced personnel nor the financial resources to defend themselves adequately in international arbitral proceedings and prepare themselves properly in the crucial phase immediately ahead of such proceedings’.¹⁰⁴ Therefore, there is an ‘[in]equality of arms’ between developing States and international investors in arbitration.¹⁰⁵ The proposal for the advisory centre is geared towards providing technical assistance to developing States in a way that reflects Poulsen’s prescriptions of building technical capacity and information pooling.¹⁰⁶ I argue that this proposal, while maybe not misguided in its immediate concern of levelling the playing field between developing States and foreign investors in arbitration, suffers from the same epistemic fallacy as Poulsen above. The proposal’s prescriptions equate the empirical effects visible in developing States’ experience with international arbitration to a ‘real’ lack of capacity to appropriately engage with the regime. By focusing on the empirical level of developing States’ engagement with IIL, the advisory centre remains unable to engage with the deeper, more complex structural features of the regime that critical scholarship argues produce the ‘inequality of arms’ experienced by developing States. Consequently, it offers only a superficial technical fix to what are very complex, open problems, problems that at least warrant more consideration for paradigmatic reform than what is currently provided in the mainstream discourse.

As of February 2024, the WGIII have completed a draft statute for a WTO-styled advisory centre for international investment law that is meant to form either a protocol or annex to a broader multilateral instrument on ISDS reform.¹⁰⁷ The stated objectives of the WGIII’s draft statute are to:

• ‘provide training, support and assistance with regard to’ arbitration, and

• ‘enhance the capacity of States and regional economic integration organizations in handling international investment disputes, in particular, least developed countries and developing countries’.¹⁰⁸

The centre would focus primarily on assisting developing States. It would assist ex post by giving technical support in arbitration through activities such as providing preliminary assessments of cases, representing a Member through the duration of a case or facilitating the appointment/selection of legal counsel. The centre would assist ex ante by advising on issues pertaining to dispute prevention, functioning as a forum for information exchange or even potentially assisting developing States with ‘the review of, and potential amendment to, their international investment instruments’.¹⁰⁹ The last set of reforms to the draft statute of the centre are due to be presented to the Commission for its adoption in July 2024.¹¹⁰ The following analysis is based upon the working drafts made publicly available at the point of writing.

It could be argued by proponents of ISDS reform that singling out the advisory centre for analysis is mistaken because it forms only one part of a group of reform proposals. These include a code of conduct, a multilateral investment court, an appellate mechanism, different appointment procedures, and other ideas.¹¹¹ A comprehensive set of reforms, of which the advisory centre is only one part, could possibly answer many of the different factors of the dissatisfaction with ISDS. Alternatively, States could potentially choose a la carte from a ‘menu’ of reform options they see as best fitting their interests.¹¹² While there is something to be said about the likelihood of an over-complexification of ISDS from such an approach,¹¹³ the point of singling out the advisory centre is to expose how the WGIII’s overall epistemological approach is problematic. Critical realism’s framework reveals how the WGIII’s work ‘presupposes an ontology of closed systems’.¹¹⁴ It remains within the empirical reality already established by the regime and its corresponding scholarship.

The empirical reality of the advisory centre

This is most readily apparent in the ontological basis of the proposal for the advisory centre. The WGIII’s reforms do not account for the distinction between the transitive knowledge gained in empirical studies like that of behavioural economic methods and intransitive, open structures like that in which IIL operates.¹¹⁵ The proposal comes from a particular epistemological vantage point which affirms the overarching institutions, principles, and rationales of the regime. It affirms that ISDS is primarily designed for foreign investors’ claims against States, arbitration is an appropriate dispute resolution mechanism, that it should remain international and the regime’s substantive protections are desirable.¹¹⁶ In focusing on only what is empirically identifiable, reformers limit the breadth of their examination. Like Alvarez states, ‘Overly focused on plugging rule of law holes in investment arbitration, reformers are neither refuting nor responding to the fundamental criticisms of IIAs nor providing alternative reasons why such treaties are needed’.¹¹⁷ The generative mechanisms that give rise to IIL’s operation are found in the open social realm in which it exists and transcend the transitive ability of humans to fully account for them with empirical methods. Critical realism’s ontological framework emphasizes that ‘It’s not the cases, it’s the system’.¹¹⁸ Open generative mechanisms are not delimited to the closed parameters of study that are required for empirical methodologies like behavioural economics to carry out their analysis.

The advisory centre remains within the same rational baselines of competition and learning as Poulsen’s Bounded Rationality above. First, by making the assistance of developing States one of its core objectives, the advisory centre affirms the rational competition baseline. It presumes the necessity of international arbitration in making developing States an attractive investment destination for foreign capital.¹¹⁹ Such sentiment is reflected in commentators’ justification for the need for a multilateral investment court: ‘Private investment is indispensable to the world’s economic development and prosperity, and thus, it must come with means to be protected.’¹²⁰ This view assumes that it is not how private investment is protected, only that this protection needs to be further legitimized through the reforms offered by the WGIII. Secondly, the advisory centre’s procedural focus affirms the rational assumption that States’ problems with the regime are ultimately one of expertise and that more assistance will provide better learning. Such affirmation presumes that international arbitration is a desirable, perhaps even necessary, feature of IIL in the first place. This perspective is reflective of the narratives surrounding IIL and international arbitration’s role in providing a globalized rule of law or governance mechanism.¹²¹

The epistemic fallacy of the advisory centre

By remaining in an empirical reality which accepts the core attributes of IIL, its institutions and rationalist justifications, the reform efforts commit the same epistemic fallacy as Poulsen above. The proposal for the advisory centre mistakes the empirical effects of the regime for the real dissatisfaction apparent in its backlash.¹²² The succeeding paragraphs focus on two of the primary effects the proposal attempts to attenuate and suggest that the advisory centre would only engage with the empirical level of these problems. Because the advisory centre would remain at the empirical level, it is unlikely that its activities will engage with the deeper generative mechanisms that give rise to the issues it would attempt to attenuate.

The first empirical effect addressed by the advisory centre is relevant to its ex post assistance in international arbitration. Developing States have reported that funding for arbitration can be a problem because they do not always have the capacity for an in-house team dedicated to ISDS. ¹²³ Hiring outside counsel can create a trade-off between expenditures for domestic State functions or paying for outside counsel. Consequently, developing States are 22 per cent more likely to settle an investment claim than their developed counterparts.¹²⁴ The advisory centre would help attenuate this disparity by providing another means by which developing States can acquire technical expertise to defend against investor claims. Presumably, such expertise would have the knock-on effect of producing more favourable arbitral outcomes for developing States, which would ultimately lead to less dissatisfaction with the regime.

This solution commits the epistemic fallacy because it links the effect of developing States’ disparity in arbitration to a lack of expertise and personnel rather than engaging with the underlying economic or educational context that precedes such disparity. There has already been other initiatives meant to help bolster the technical capacity of developing States both ex ante in treaty negotiations and ex post in arbitration with limited impact.¹²⁵ For example, United Nations Trade and Development (UNCTAD) has hosted ‘facilitation rounds’ that brought officials from both developing and developed States together and provided additional expertise where there was a perceived lack of capacity in a party during negotiations.¹²⁶ The organization has also produced work on different policy options to develop a ‘new generation’ of investment treaties that were more oriented towards developing State needs.¹²⁷ Additionally, there are organizations such as Tradelab which trains students in the specific technical aspects of arbitration and disseminates its research to build legal capacity for States.¹²⁸

Perhaps these efforts’ inadequacy is a matter of the scope of the problem. But even if that were the case, these measures do not address the economic and educational factors that create developing States’ lack of expertise and capacity in the first place. Economic pressures simultaneously keep developing States in a state of dependency for foreign capital while creating the impetus of Global South citizens to seek opportunity elsewhere, a phenomenon commonly known as ‘brain drain’.¹²⁹ The point is not that technical assistance should not be provided to developing States, but that technical assistance only engages with the very surface level of the problem. Developing States’ lack of arbitral capacity is the result of a myriad of historical, political, and economic contingencies of which, like Alvarez points out, only ‘plugging’ the holes in technical expertise does not address.

The second empirical effect addressed by the advisory centre is relevant to its ex ante assistance in ISDS prevention. States have longstanding complaints about arbitral inconsistency and incorrectness.¹³⁰ Because IIAs are typically vaguely worded documents which leave much room for interpretation and previous arbitral decisions do not bind future tribunals, there have been controversial instances in which tribunals have reached inconsistent decisions.¹³¹ Accordingly, the advisory centre offers a solution by providing a potential assistance mechanism for the (re)negotiation or amendment phase of investment treaties.¹³² This solution would presumably help prevent tribunals from reaching inconsistent or incorrect arbitral decisions because they would have clearer, better defined treaty protections from which to interpret.

Like the above, this solution also commits the epistemic fallacy by linking developing States’ negative experiences in IIL to instances of inconsistent arbitral decisions rather than more deeply engaging with what leads to such dynamics in the first place. While more technical support in the treaty drafting or renegotiation phase could potentially provide more effective or efficient arbitral interpretation, this by no means necessarily entails a reduction of inconsistent or incorrect arbitral decisions. There are various structural factors that maintain tribunals’ autonomy to make such decisions including: public international law is not subject to third-party assessment for its correct interpretation; IIL is structured around broadly textured provisions that are highly flexible according to the circumstances of a case; Article 31 of the Vienna Convetion on the Law of Treaties (VCLT) requires various other factors of the case be considered beyond just the ordinary meaning of the text; there is no binding precedent set by arbitral provisions; there is a lack of consensus over customary rules; current review mechanisms limit themselves to procedural matters; and, finally, perceptions about incorrectness do not automatically mean the decision was in fact incorrect.¹³³

To reiterate Koskeniemmi’s point, ‘It’s not the cases, it’s the system’.¹³⁴ Tribunals do not make incorrect or inconsistent decisions because the provisions that they are interpreting have not been negotiated in a way that most accurately and fairly represents a developing State’s interests. They reach their decisions in the way they do because they are part of a legal regime that explicitly prioritizes the commercial interests and protections of foreign capital over other social or environmental interests.¹³⁵ Changing the setting of international arbitration would not automatically entail a re-envisioning of IIL’s raison d’etre, the protection and promotion of foreign investment. More progressive treaty formulations do not guarantee their intended outcomes.

Even if the advisory centre was only one part of a more comprehensive reform package that included a multilateral investment court with elected sitting judges, these judges would still be tasked with interpreting provisions that are oriented towards the commercial protections of foreign investment.¹³⁶ There is little indication either a centre linking experts with developing States or a court with fixed-term judges would depart from the already established institutional know-how concerning international arbitration or offer any novel insight about treaty negotiation to developing States.¹³⁷ Rather, there is every indication that such a panel of judges would likely comprise of individuals from the ‘revolving door’ which already comprises arbitral tribunals and who uphold the core attributes of the regime.¹³⁸ This is evident in the view of one interviewee in a scoping study about the advisory centre who ‘noted an automatic tendency of both private-sector as well as international organization advisors to include standard arbitration provisions in treaties without considering the state’s broader investment policy interests or priorities’ when assisting with treaty redrafting.¹³⁹

The problem with the proposal for an advisory centre for IIL is that it mistakes the empirical effects of inconsistent arbitral decisions or poorly drafted investment treaties as the reasons behind the dissatisfaction with IIL itself. The proposal remains within the empirical reality of IIL by only offering prescriptions from within the same institutions that comprise the regime. Procedural fixes like a body that streamlines the assistance provided to developing States in treaty negotiation may help ensure additional clarity in a particular treaty clause, but it does not address Indonesia’s call for a deeper consideration of some of IIL’s substantive norms.¹⁴⁰ Providing additional support to developing States in arbitration through expert council may give more balance to the inequality of arms in arbitration, but it does not address South Africa’s question of whether ‘ISDS mechanisms are desirable or necessary in the first place’.¹⁴¹

The questions raised in Indonesia and South Africa’s statements speak to the very paradigm of international arbitration and the current status quo of IIL’s operation. Critical realism’s ontological/epistemological framework exposes how the epistemological approach taken to such questions influences to a significant degree the actions that will be taken in response. The proposal for an advisory centre is not bad in and of itself. There are valid arguments for its place in the reform of ISDS. However, the problem with the proposal is that it comes out of the very same epistemologies and institutions that international arbitration and IIL come from. It originates out of an international regime of law which reduces nation states’ sovereign capacity to determine how foreign capital is regulated within their borders and enforces awards handed down by international tribunals unelected by the communities that will be most impacted by their decisions. The advisory centre originates from the same empirical reality that the problems it is meant to attenuate originate. It is a superficial ‘fix’ to problems that originate in incredibly complex historical, political, and economic circumstances.

Developing States may lack capacity in their engagement with international arbitration, but in only considering how to attenuate this empirical reality, States, practitioners, and scholars miss the complexity of the circumstances that underlay a developing State’s lack of capacity in international arbitration. It would therefore behove participants in IIL’s reform efforts to step back and consider the epistemological assumptions from which they approach these problems. Doing so may cause further consideration about what kind of paradigmatic changes might be possible or even necessary to reorient the relationship States have with foreign direct investment and their domestic exigencies.

CONCLUSION

Paradigmatic reform is difficult because it entails many unknowns. A post-international arbitration world would need to recalibrate how transnational investor–state disputes would be handled. Would Calvo’s ghost reappear on the horizon in the form of domestic litigation? Would a State–State dispute model be any less cumbersome for developing States? What might such rearrangement of the law for foreign direct investment mean for foreign investors and developing States economic relationships? What might it mean for the local communities who are directly affected by foreign investment? All these questions and many more would certainly require confronting. They are not easy. However, if there ever was a point where they should be considered, that time is now.¹⁴²

The time for such an extensive reflection is quickly passing. The deliberations of the WGIII are in their final stretch and it is becoming more and more apparent that what is being offered up are proposals for incremental and systemic change. This article turned to critical realism and its novel ontological and epistemological framework to demonstrate why this might be the case. The answer can, at least in part, be identified in the epistemological basis of those involved—States, academics, practitioners. Many of the more dominant participants in the reform process are stuck in an epistemology that is both institutionally and theoretically embedded in epistemological positivism. They view their agency through the empirical effects of the institutional and theoretical systems of closure in which they are encapsulated. They do not see that legal, economic, and political praxis takes place in an open social system. This open system is one of possibility.¹⁴³

This article has attempted to reveal this constraint in both Poulsen’s formulation of developing States’ bounded rationality and the WGIII’s proposal for an advisory centre. Both of these efforts are done with progressive change in mind. Bounded rationality is meant to be an epistemological corrective to traditional forms of rational choice. The advisory centre is meant to correct for the disadvantages developing States face in treaty negotiation and international arbitration. The problem critical realism exposes with these perspectives, however, is that they are still stuck within the overarching empirical reality of the current status quo. If we are to sincerely pursue justice in the contemporary circumstances in which international law finds itself, we must be willing to reflect upon its operation at a level deeper than its current institutions and empirical effects. The law can, and perhaps should, change paradigmatically.

I would like to thank Anil Yilmaz Vastardis, Piotr Wilinski, Federica Violi,  G.J. Meijer and the participants at the SAIL Conference 'Justice in a post-ISDS world' at Erasmus School of Law, Marina Lostal, Ryan Gunderson, Will Lewis, Meagan Wong and the Public International Law Cluster at University Essex Law School, who all provided thoughtful comments on the development of this article. I would also like to thank the anonymous reviewers for their comments. Any mistakes are my own.

Footnotes