Investment Treaties: The Reform Matrix

This essay proposes a matrix for understanding the dynamics of investment treaty reform. It tracks incremental, systemic, and paradigmatic reform options as applied across procedure, substance, and form. Although stylized and thus unable to capture all the nuances of individual positions, the reform matrix creates a framework for understanding some of the main debates about investment treaty reforms and offers a template for locating and comparing the approaches of key international actors, including the United States, the European Union, and Japan, together with Brazil, Russia, India, China, and South Africa (the BRICS).

The bilateral investment treaties (BITs) of major Western states can be roughly divided into two generations. First-generation treaties from the 1990s and earlier (BITs 1.0) typically included strong investor protections and binding investor-state arbitration. A series of cases against the states that entered these treaties forced them to recalibrate, leading to a second generation of treaties from the mid-2000s onward (BITs 2.0) that aim at striking a better balance between investor protection and state sovereignty, while retaining investor-state arbitration. However, the Western powers that have adopted these treaties are now split on various issues, most notably the merits of establishing a multilateral investment court.

In light of these divisions and the growing economic importance of the BRICS, it is worth asking where the BRICS fit in this story and which investment practices and reform proposals they support or are likely to support. As this symposium demonstrates, as regards their treaty practice trajectories and reform positions, the BRICS often differ significantly not only from major Western powers, but also from each other. Given the disaggregation of international power and the varying preferences of key actors both within and outside the West, the most likely outcome is a pluralist one encompassing the coexistence of multiple approaches to procedure, substance, and form.

Incremental, Systemic, and Paradigmatic Reform

I contend that it is best to conceptualize investment treaty reform options on three levels: incremental, systemic, and paradigmatic.¹ Incremental reforms involve adopting small-to-moderate improvements on the existing model. Systemic reforms work within the current system but require larger-scale, often-structural reforms. Paradigmatic reforms entail more extensive changes that go outside the existing system, often creating something new.

To operationalize these concepts, one needs to set a baseline. There is no neutral starting point, as the practices of states differ and whatever is chosen will effectively naturalize that position. Nonetheless, in view of the role of major Western powers in setting the terms of many reform debates, and given the general familiarity of many in the field with Western models, this section takes as its baseline for comparison the United States’ BIT 2.0 model (that is, the U.S. Model BITs of 2004 and 2012). Incremental reforms would thus seek small-scale changes in this model (say, BITs 2.1 to 2.2), whereas systemic reforms would involve large-scale changes (BITs 3.0). Rather than represent BITs 4.0, paradigmatic reforms would move outside the investment treaty paradigm altogether.

I use the terminology of incremental to paradigmatic reforms to capture the scale of the changes required by the reform option, not their merits. What resembles transformational reform or a revolutionary approach within one paradigm might look traditional within another. Some reform options, such as eliminating investor-state arbitration in favor of a return to state-to-state arbitration, seem transformational or revolutionary when viewed within the BIT paradigm but are quite traditionalist when viewed within a broader, public international law paradigm.

The range of incremental to systemic to paradigmatic reforms should not be understood as three distinct points on a spectrum. Actors do not need to try incremental reforms before moving on, or to transition through systemic options to adopt paradigmatically different approaches. Instead, these ideal types are better conceptualized as three points on a triangle. Actors can move directly from any one point on the triangle to any other point. States may adopt hybrid or intermediate positions between two or more of these ideal types. They may also adopt a combination of approaches, with respect either to different treaties or to different issues within a treaty.

Procedure, Substance, and Form

Although disaggregating these three levels of reform is useful, applying them alone to characterize the approaches of states to investment treaty reform is too blunt. Thus, it is helpful to consider how these three reform archetypes apply across three dimensions: procedure, substance, and form.

On procedure, BITs 2.0 permit investor-state arbitration. The Comprehensive and Progressive Agreement for Trans-Pacific Partnership exemplifies an incremental approach to procedural reform: it retains the core of investor-state arbitration but includes several reforms to address criticisms, such as the ability to strike out unmeritorious claims at an early stage and a code of ethics for arbitrators.² In the UN Commission on International Trade Law (UNCITRAL) reform process, the United States, Chile, Japan, and Russia were the most vocal advocates of incremental measures, though the position of the United States seems less clear more recently owing to political developments related to its North American Free Trade Agreement negotiations.³

Systemic reforms on procedure would still accept the basic premise that investors should be permitted to bring international claims directly against host states on the international plane but would change the nature of the dispute resolution from ad hoc arbitration to a permanent court with an appellate mechanism. The European Union, Canada, and Mauritius are the key advocates of this approach in the UNCITRAL reform process. States could also adopt intermediate positions, such as allowing for arbitration with incremental improvements and the systemic reform of an appellate mechanism, without accepting an investment court (semisystemic reform).

Paradigmatic reforms on procedure would challenge the assumption that investors should be able to bring international claims directly on the international level. For instance, South Africa terminated its investment treaties and, instead, under domestic legislation, permits foreign investors to bring direct claims against the government in its domestic courts. If a dispute persists, South Africa may later consent to state-to-state arbitration.⁴ Brazil has championed an approach whereby an ombudsman is empowered to try to resolve disputes involving foreign investors. But Brazil's advance consent to state-to-state arbitration is built into the process if the dispute persists.⁵ Although transformational or revolutionary within the BIT paradigm, both approaches are traditionalist when viewed in a broader historical context.

On substance, reforms are occurring on multiple levels and thus cannot easily be captured on a simple spectrum of incremental to paradigmatic reform. Two core characteristics of BITs 2.0 are that they (1) comprise a standard set of protections for foreign investors, including provisions on expropriation, fair and equitable treatment, national treatment, and most-favored-nation (MFN) treatment, and (2) do not impose enforceable obligations on foreign investors. Accordingly, they focus on strong investor protections but not investor obligations, and they grant protections to foreign investors, not domestic ones.

States adopting incremental reforms on substance accept the ongoing validity of the traditional protections for foreign investors but seek greater specifications and limitations. Examples include spelling out in detail and delimiting the application of controversial provisions like indirect expropriation, fair and equitable treatment, and the MFN clause.

Systemic reforms could involve taking a more selective approach to foreign investor protections, such as omitting protections for indirect expropriation and fair and equitable treatment, or trying to protect other values, such as environmental sustainability and fair labor standards. The softer and less enforceable the protections given to other values, the more likely they should be characterized as incremental rather than systemic reforms.

A handful of treaties or models envisage imposing enforceable obligations on foreign investors, which represents a more transformational reform as it shifts the focus of the paradigm from investor protection alone to investors’ rights and obligations. An example of this approach is the Southern African Development Community (SADC) Model BIT.⁶ Another paradigmatic reform could be the elimination of any distinction between the protections afforded to domestic and foreign investors.

On form, three main options can be detected. BITs 2.0 seek to ensure generalized protections for foreign investors through an inherently sovereign act undertaken on the international stage (a treaty commitment). Those adopting incremental reforms would accept that treaties remain the preferred form for investor protection. Systemic reforms, by contrast, would challenge at least one of the form characteristics of BITs. The best example is South Africa's decision to shift from treaty protection to domestic legislation. South Africa's approach still provides protections to foreign investors in general and is achieved by an inherently sovereign act: the passing of legislation. Yet a treaty constitutes an obligation undertaken on the international stage in a reciprocal agreement with another state, whereas legislation is enacted domestically. Such protections can therefore be unilaterally repealed by the host state and will often be subordinated to its constitution.

Paradigmatic reforms of form would involve a more fundamental shift beyond the BITs 2.0 model. Such reforms would be best illustrated by a move from generalized investor protections achieved through an inherently sovereign act (treaty or legislation) toward individual protection of investors based on an act undertaken by a state in its private capacity, that is, investment contracts. The commission tasked with reviewing Ecuador's investment treaty policy recommended this approach.⁷ Although transformational within the BIT paradigm, a contract-based approach is traditionalist when judged more broadly, as investment contracts preceded investment treaties and often exist alongside investment treaties. They exemplify a paradigmatic change because they transfer protection from a public law model to a private law one.

Different dimensions also interact to create stronger or weaker protections for investors and states. For example, if investment contracts are coupled with international arbitration, they may result in stronger limitations on states than if they are coupled with enforcement before domestic courts (interaction of procedure and form).

Locating States on the Reform Matrix

The above categories represent ideal types that embody a useful way of organizing the principal kinds of reform being discussed in international debates and the main dimensions along which reforms can operate. But not all states fit neatly in a single box and even their approaches to their own treaties often vary (see Table 1). One can thus speak only in broad terms when categorizing these approaches.

Table 1. States and Organizations on the Reform Matrix

The major Western powers are concentrated at the incremental end of each spectrum, with the exception of the European Union and Canada, which seek more systemic reform when it comes to procedure. This clustering is not surprising since these states played the dominant role in developing the BIT 2.0 model used as the baseline for the reform matrix. What is interesting, however, is the degree to which the BRICS are split across the reform matrix. Nothing close to a consensus position on investment treaty approaches and reforms prevails among the BRICS.

Brazil is an interesting outlier, as Henrique Choer Moraes and Felipe Hees explain.⁸ Alone among the major powers, Brazil has never been subject to a ratified BIT. This stance has given it a clean slate on which to forge a novel investment treaty policy. It has responded by adopting investment facilitation agreements and supporting debates on the issue at the multilateral level under the auspices of the World Trade Organization. Key to the investment facilitation agreements are a narrower range of investment protections (systemic reform on substance) and the use of domestic ombudsmen for foreign investors followed by state-to-state arbitration (paradigmatic reform on procedure). This approach is not revolutionary when judged from the perspective of Brazilian treaty practice or public international law more generally. But Brazil's practice holds out a model for paradigmatic reform when viewed from the perspective of other states within the system.

When it comes to the reform matrix, the Russian Federation is best understood as retrogressive rather than reformist, though this characterization also reflects the inherent biases built into using BITs 2.0 as the baseline for comparison. As a Communist state, the Soviet Union did not embrace investment treaties with strong investor protections or broad investor-state arbitration provisions. Following the USSR's dissolution, Russia signed investment treaties on the BIT 1.0 model. Even in its recent treaty practice, however, Russia has not embraced the movement toward the BIT 2.0 model.⁹ For this reason, Russia largely does not appear on the reform matrix. In UNCITRAL, it seemed to be strongly opposed to multilateral reform efforts on dispute resolution, repeatedly emphasizing the strengths of investor-state arbitration, the need for flexibility and confidentiality, and the importance of permitting states to decide individually or bilaterally on their treaty practices. Dmitry Labin and Alena Soloveva attribute this position partly to the categorization in Russia of investment treaties primarily as a form of private international law, rather than public international law or public law.¹⁰ Brazil and Russia present an interesting contrast on this point.

India is hard to classify under this schema because its approach is so eclectic.¹¹ It terminated its existing investment treaties and developed its own Model BIT in 2015.¹² India now accepts investor-state arbitration but conditions it on extensive resort to domestic remedies for up to five years.¹³ It limits investment protections in several ways, such as by removing the fair and equitable treatment clause, but does not clearly enable states to bring counterclaims against investors to enforce enumerated investor obligations. Yet it may be that requiring investors to comply with domestic laws to bring a claim will mean that noncompliance removes the jurisdiction of the tribunal rather than permit a counterclaim.

South Africa also terminated some treaties and allowed others to run out the clock.¹⁴ It has since charted its own course on investor protection, embedding such protections in domestic legislation that subjects them to dispute resolution before its domestic courts, and not providing for ex ante consent even to state-to-state dispute resolution. Nevertheless, it appears that South Africa has been more paradigmatic in reforming the procedure and form of its investment law than its substance because it did not go the extra step of imposing enforceable obligations on investors. In many ways, though, its legislation is aimed at putting domestic and foreign investors on an equal footing. In UNCITRAL, South Africa has argued that reforms that address procedure without addressing substance will not solve the real problems of the system.¹⁵

China's investment treaty practice has undergone an extraordinary transition from an extremely pro-state-sovereignty approach that allowed for limited investor protections and investor-state arbitration, to an approach that has much more in common with the BIT 2.0 standard.¹⁶ But its practice is also one of the most inconsistent of any major power, suggesting that it is often prepared to adopt different approaches in different negotiations. Although China has yet to state its preferences clearly at UNCITRAL, it has maintained that consistency and coherence are important and not adequately secured by the system's existing review mechanisms,¹⁷ which might imply that China could favor semisystemic reform (for example, adopting an appellate mechanism but not a court).

For now, it is safest to characterize China as having adopted a cautious, wait-and-see approach. China's strong interests as a capital exporter, particularly in light of its Belt and Road Initiative, make it highly unlikely to support paradigmatic reform, as Congyan Cai explains.¹⁸ China has announced its establishment of courts for disputes arising from commercial agreements along the Belt and Road,¹⁹ which Huiping Chen argues is not only about protecting Chinese interests but also about breaking the hegemony of Western institutions over the resolution of commercial disputes.²⁰ With respect to treaties, China seems likely ultimately to position itself between incremental and systemic reform because it has an interest in maintaining a workable system of investor protection in the future but, at present, it remains in an incrementalist stance.

BRICS and the Path to Pluralism

The next generation of investment treaties will not involve a singular gold standard. Instead, it will likely involve more options and less coherence, resulting in a more plural system.²¹ That should come as no surprise. The international system has moved past the era of Western dominance in international law in general, and in investment treaties in particular. On key points, the West is divided. The BRICS are asserting themselves on the international stage with respect to investment treaty policies and reforms, but their divergent approaches also increase the likelihood of a pluralist outcome. Although lawyers often clamor for coherence, the resulting diversity permits a sort of experimentation that is potentially very useful given the relative lack of knowledge policy-makers, academics, and practitioners still have about the utility of different approaches.