UNCITRAL and ISDS reform: Lifelong learning

EJIL: Talk! | 23 November 2023

UNCITRAL and ISDS reform: Lifelong learning

by Anthea Roberts and Taylor St John

How many International Investment Disputes have there been since 1973? How many International Investment Disputes are ongoing now? We do not know. There is no authoritative count. How many treaty-based awards issued by arbitral tribunals since 2000 are not publicly available? Again, nothing authoritative can be said except that a sizeable percentage of such awards are not available. The important work of Investment Arbitration Reporter and others mean that we have informal knowledge about the existence of many cases. But our frequent inability to answer even basic questions about the full universe is a salutary reminder that we – as a scholarly community and as a policymaking community – know less than is ideal about investor-state dispute settlement (ISDS).

Starting with how little we know also reframes what is happening at UNCITRAL Working Group III, which has been discussing ISDS reform since 2017. Sure, the Working Group is developing and delivering reforms, as we discuss in yesterday’s companion blog. In addition, however, the Working Group is facilitating learning and generating authoritative information as states put their experiences with ISDS on the record. The process is a rare opportunity to learn about ISDS from states, particularly those states that we as scholars often hear from less or read about less. In this blog, we reflect on what we have learned about ISDS through listening to states and explore implications for future research.
Beyond Investment Treaty Arbitration

Questions of scope are discussed regularly by UNCITRAL Working Group III. One scope question in particular stood out during the discussions about a Code of Conduct for Arbitrators and a Code of Conduct for Judges: should the codes apply to all cases or only treaty-based cases? Eventually, after substantial debate, the Working Group agreed that the Code of Conduct for Arbitrators could apply to ‘international investment disputes’ defined as a dispute between a foreign investor and a state in which the instrument of consent is a treaty, legislation governing foreign investments, or a contract.

Similar scope questions look set to recur as each reform is discussed. During the Singapore Intersessional in September, for instance, attendees asked if an appellate mechanism should hear only treaty-based cases, or should it be able to hear contract-based and national law-based cases too? Should it be able to hear state-state cases? Views on these questions vary. Some states believe that an appellate mechanism should hear treaty-based cases only to better serve the aims of consistency and coherence. Other states envision an appellate mechanism that is able to hear contract, law, and treaty-based claims, as well as state-state disputes related to foreign investment.

These scope questions also emerge regarding an advisory centre: in drafts prepared for the October 2023 session, the Secretariat used the term investor-state dispute settlement rather than international investment dispute, to be more expansive. The Secretariat explained that ISDS was intended ‘to broadly refer to all types of dispute resolution mechanisms to resolve disputes between an investor and a State (including investment mediation and State-to-State dispute settlement) regardless of whether the legal basis for the dispute is an investment treaty, domestic legislation or an investment contract)’ (Draft provision 2, para 8, Working Paper 230).

This language is very broad, but accords with the desire of many states in the Working Group to define the term IID as more than investment treaty arbitration, and then define ISDS as even more inclusive. This desire to define ISDS broadly is not universally shared in the Working Group, the definition of ISDS has not been settled definitively, and even states that support a wide scope for some reforms believe that a narrower scope is preferable for other reforms. But listening to these recurring discussions about scope have led us to reflect on how what scholars often study as ‘investment treaty arbitration’ is different from what states see. Two questions stood out as we listened to the realities emphasized by states.

The first question is: how much of the contemporary ISDS universe is excluded if only treaty-based cases are included? From most academic literature, particularly quantitative literature, one gets a sense that most cases are treaty cases and therefore not much is excluded from focusing only on treaty-based cases. But listening to states during the Code of Conduct discussions, we heard African states insist that contract-based disputes be included in the definition of an international investment dispute, because this is where the majority of their cases come from.

Available evidence supports their point: investment treaties are the basis of 99% of ICSID cases with a European or North American respondent state, but only 26% of ICSID cases with an African respondent state and 64% of ICSID cases with an Asian respondent state, according to recent ICSID caseload reports. Last year, the PCA reported providing registry services to 200 investor-state cases, almost half of which (88 cases) were based on contracts. Other, traditionally more commercial secretariats are likely to have more contract-based cases (for instance, the main secretariat for African contract-based cases is likely to be the International Chamber of Commerce not ICSID or the PCA), so estimating that half or more of all investor-state arbitrations are contract-based is not unreasonable. Contract-based cases are not necessarily smaller or less important, as Nigeria reminded the Working Group in October 2023 by bringing up the contract-based case P&ID v Nigeria, in which the claimant was awarded USD$ 6.6 billion, one of the largest known awards.

Have we, as scholars, been studying less than half of the international investment disputes faced by states? It is a difficult question to answer, but an important one to ask. And it is an example of how researchers can use the points that states raise at Working Group III as inspiration to reflect on what aspects of ISDS they are able to see, and what aspects they are not able to see. We recognize that it is difficult to study awards that are not public or to count cases about which limited or no information is public, but we can still do better at discussing how these absences shape our perceptions or bias our conclusions.

A second question that stood out to us is: why are so many states keen on saving space for state-state mechanisms? For several years, we have heard Brazil intervene in the Working Group to request that any reforms developed would be compatible with, or able to accommodate, state-state dispute settlement mechanisms. Now we hear other states also speaking up to ensure that state-state dialogue or dispute settlement is accommodated. For instance, in the October 2023 discussions, Kenya reminded the room of a trend toward providing state-state dispute settlement in treaties, which the European Union then echoed, using Kenya’s point as an example of why it would be risky to exclude activities like state-state dispute settlement from the statute of the advisory centre.

Interventions to save space for state-state proceedings are not a death knell for investor-state proceedings. But for generations of legal practitioners and academics who were taught that providing investors with standing is an advance, it is an opportunity to reflect on why several states – including some often seen as less powerful, some seen as middle powers, and some seen as very powerful – are emphasizing the importance of state-state mechanisms. These states may have different reasons for seeing them as desirable, and may even be envisioning different types of state-state dialogues or mechanisms. Collectively, however, it signals a rethinking underway about the costs, politics or logic of how to resolve investment disputes, and a move away from the idea that providing investors with standing elevates disputes above politics and toward law.

The Importance of State Oversight

In earlier blogs, we observed how eager states were to share experiences and learn from each other and noted that the most important development occurring in the Working Group may be the emergence of a hub. Our sense that a hub is emerging has only grown stronger as the process has continued, and now, in addition to watching the Working Group serve as a temporary hub, we watch as states build a more permanent institutional hub. Or hubs, plural – it is an open question if states will create different bodies to oversee different reforms or one body with wider oversight. (Or a hybrid, with bodies that are different on paper, but meet physically in the same place with varying configurations of members; discussing the codes of conduct on Monday and the damages protocol on Tuesday, for instance, with only those states that had joined the code of conduct or damages protocols attending the Monday and Tuesday meetings respectively.)

At the moment, several possible hubs have emerged in connection with various reforms. The most recent advisory centre drafts include a forum for the exchange of information and the sharing of best practices among states (draft provision 6(1)(c) in Working Paper 230). Draft provisions for a standing first-instance tribunal envision a committee of the parties composed of representatives of all parties who have joined the tribunal (draft provision 3(a) in Working Paper 213). Draft provisions for the overarching framework agreement include a conference of the parties consisting of states that have joined the agreement (para 18 of Working Paper 221). This conference of the parties could have a mandate that includes discussing emerging issues, state practice, as well as possibly updating existing protocols or negotiating new ones, or other features that enable it to grow into a dynamic hub.

While many questions remain about how these hubs will function, what is clear is that states are building tools to monitor and exercise oversight over the institutional actors and instruments that constitute investment governance. The emergence of collective state oversight is already noticeable, especially in smaller, less contentious moments like the discussion of a Draft Legislative Guide on Dispute Prevention and Mediation (Working Paper 228) in March 2023. The discussion was brief and not characterized by any of the usual differences of view in the Working Group. Instead, there was unanimity – in a rare occurrence, every state and observer objected to the tone and substance of the draft that had been prepared by the UNCITRAL Secretariat.

Most of the objections raised by states circled around four main points. First, the document had a prescriptive tone. Early in the discussion, the draft was described as being ‘overly prescriptive’ and objections were raised to the imperative word ‘shall’ being used; these concerns were then echoed by speakers from a wide range of states. Second, the document overlapped in topics addressed with other agreements, notably the Joint Statement on Investment Facilitation at the WTO, but with stark differences in how it addressed these subjects. Third, the document did not reflect actual state practice. Several OECD states made clear that they did not do what the guide prescribed as good practice, and ‘we should avoid the impression that all states are expected or even required to establish legal frameworks based on this guide.’ Fourth, the document lacked sensitivity to differences among national systems. Canada stated that ‘we can’t use a one-size-fits-all approach for this’, while Chile stated that such a guide was not ‘attuned to the legislative reality of each state’, with their points being echoed by Colombia, India and others.

Since papers prepared by the Secretariat have otherwise been commended by states as excellent and responsive to state concerns, there seemed to be a sense of bewilderment in response to this one, vividly expressed by France:

We’re somewhat taken aback, as are other delegations. This document has descended upon us like a meteorite from heaven, it is not really what UNCITRAL usually does… we think that this format is not appropriate, not attuned to what we do here.

What happened? The UNCITRAL Secretariat co-authored the draft with the parts of the World Bank Group that provide technical assistance on investment climate issues to least developed states. While the UNCITRAL Secretariat is accustomed to having its drafts scrutinized and debated, these technical assistance providers are unaccustomed to this kind of scrutiny. This difference may explain many of the objections voiced by states. The disconnect with the Joint Statement on Investment Facilitation, for instance, is likely explained by the difference in state involvement: the Joint Statement was heavily scrutinized and debated, but there are not similar channels for formal, collective scrutiny of the advice given by technical assistance providers within the World Bank Group.

This is only one example of a larger reality faced by states: different technical assistance providers make different recommendations, and what some technical assistance providers describe as best practice and what Switzerland, for instance, actually does in practice are often profoundly different. This reality hit home in the Working Group as several states made clear that they want to learn from each other’s actual practice, but they do not want to receive prescriptive documents. India, for instance, stated, ‘we are open to learn from best practices of other states, but to have a one guide to fit all, cast in stone, has its own practical difficulties.’ Similarly, France said ‘what we would conceive of would be a collection or collation of good practices, a descriptive document’, observing that a descriptive document would be more helpful than a prescriptive one. States want to learn, but they want to learn from each other, from other officials in their shoes, who understand their constraints and know what it feels like to have skin in the game.

There were two insights we took away from the dispute prevention discussion. The first insight is how helpful it can be for scholars to say clearly where there is no evidence or the evidence is inconclusive, to say when we do not know things. There were two excellent examples during the discussions on dispute prevention. First, the observer from National University of Singapore stated during discussions:

It’s worth noting that although there are many ideas about dispute prevention and the approaches taken by individual states and international organisations like the World Bank are well noted, there is similarly very little empirical evidence about what works, what doesn’t work, and why.

The Academic Forum paper on this topic concluded similarly that there is little evidence of the extent to which dispute prevention approaches have been successful, noting ‘a dearth of studies to evaluate their day-to-day operation and practical effects’ and stating that there is insufficient evidence to support any claims of best practice. There are usually few academic incentives to register a lack of evidence or uncertainty, but doing so can help prevent inappropriate policies from proliferating, can help leave room for policy experimentation and learning, and can underline the need for more study and gathering of evidence. With regard to dispute prevention specifically, the Secretariat has undertaken a substantial redrafting to remove prescriptive words and reframe the paper as guidelines as to how states might organize themselves. The new document will be discussed in January; the Working Group may decide it is now a helpful document or may decide not to move forward with it, but it is a topic that an Advisory Centre could look into again in the future.

The second insight we took away was also about how little we know, this time about who is providing technical assistance. As we listened to observer interventions, it struck us how many actors were already providing technical assistance related to ISDS. The types of actors providing technical assistance include IOs, academics, arbitration secretariats, bar associations, law firms, and NGOs, among others. We know relatively little about what technical assistance these actors are providing and why they provide it. Studies from other areas show that while much technical assistance is helpful, other technical assistance can be harmful, and its influence should not be underestimated. In her study of science policy, Professor Martha Finnemore provided a memorable way to think about technical assistance by tracing the process whereby an IO ‘taught’ states how to build bureaucracies related to science policy: she argued that IOs can and do take on the role of teacher. With regard to ISDS, there seem to be many teachers, not just IOs. But who is teaching what, to whom? How is it funded? How is it connected to commercial activities or to policy views? We do not know the answers to these questions, but we think these kinds of questions bubble beneath the surface of some Working Group discussions, as states express a desire to learn from each other, as equals, rather than from a teacher.

Distinguishing Investors and Funders

Questions about money recur in many forms at Working Group III, and in these discussions, we increasingly hear about funders, and we hear states distinguishing between funders and investors. A recent example occurred in discussions about an advisory centre. Some observers and states would like small- and medium-sized enterprises to be able to access some services of the advisory centre; almost all developing states are opposed to this access for several reasons. What struck us, however, were interventions from developing states that there was no reason to include small and medium enterprises, because these enterprises can access funding if they want to bring a case but lack adequate resources.

These officials’ views are important indicators of the role funding may be playing in this system, given the scarcity of publicly available data regarding third-party funding in ISDS. How prevalent is third party funding? How common are contingency fee arrangements? How common is award purchase? None of these questions can be answered authoritatively. There is not even enough information from institutions or other actors to build estimates, unlike with caseload numbers. Even in-depth analysis of an individual award may not reveal if third-party funding was present, since awards do not always disclose if a litigant had third-party funding. Even the states that are involved in cases based on third-party funding might not know about the funding.

Given these challenges, the best available evidence on the prevalence of third-party funding is not numbers-based but interview-based. For a recent paper, Florence Dafe and Zoe Williams interviewed third-party funders about the prevalence of funding in ISDS:

I’m sure we’ve covered the entire market and I can say with some conviction that we are either funding [claims] ourselves, they are funded by somebody else, or they have actively considered funding and for whatever reason have chosen not to.

Answers like this suggest that there is a lot of third-party funding occurring that is not publicly known. It also hints at an information asymmetry; the business model of litigation finance firms requires that they have excellent information about (possible and current) international investment disputes, while it can be difficult for states to gather information about litigation finance, even in the claims against them.

More disclosure of third-party funding may emerge through this reform process or other processes, but in listening to states, we wondered if third-party funding is already shaping the ISDS realities they face in ways we, as scholars, have yet to appreciate fully. For instance, Dafe and Williams observe that states likely vary in their exposure to third-party funding; funders may be less likely to fund claims against states with the resources and experience to present a strong defense and also be less likely to fund claims against states known for not paying adverse ISDS awards. While this is not systematic evidence, it may not be a coincidence that many voices we hear that are concerned about third-party funding come from middle- and lower-middle income countries that have made efforts to pay adverse awards.

Third-party funding may also already be shaping behavior and perceptions regarding ISDS in ways that scholars can examine. For instance, do litigation funders behave differently from investors, for instance are funders on average less willing to settle? Are they more likely to fund when the investor made a sizable investment than when the investor is simply seeking to recover potential profits? There is still a lot that we do not know about the implications of third-party funding and many other topics within ISDS.


When we read the academic literature, we see a lot of criticism of the UNCITRAL Working Group for being too small and too incremental. When we read practitioner literature, we hear the opposite: the process is too big and too radical. But when we observe states in the process, we see that they are getting more engaged in the Working Group over time, not less. What is it that brings them back to the process, time and time again? As we observe states and listen to officials, we not only learn from them, we also observe them learning from each other. It is this learning process that seems to be a reason why states find the process useful and keep engaging with it. In a process of lifelong learning, we can learn a lot from listening to states in this process but also understanding why they need processes that help them listen to each other.

source: EJIL: Talk!