UNCITRAL and ISDS Reform: What to Expect When You’re Expecting

Written by and

From September 5 – 16, the delegates of UNCITRAL Working Group III were back in Vienna to continue discussing ISDS reform. While it was possible for registered delegates to watch the proceedings online, it was only possible to make interventions in person. This approach, which we understand to be a general UN approach, decisively shifted the Working Group’s centre of gravity back to the physical room. The people – and the margins – were back.

The margins were busy as delegates saw each other again after two years online. Many delegates arrived early and left late to have more time with colleagues. The passage of time came up often in these conversations, and was palpable in the room and process more generally. Many new, often younger, faces were in the room this time alongside the familiar faces, as some officials who have been with the process rotate into new positions within government, get promoted, or move on. 

Delegates seemed eager to get back to work with a full six hours of meeting time each day and no technical delays. Being back in person meant delegates also reoriented themselves within the process, taking stock of what has been agreed and what remains to be done before the process ends. To us, it felt like a big question bubbled just under the surface for participants: what are we expecting to emerge from this process?

The Due Date

UNCITRAL Working Group III has a due date: 2026. Many in the process seem committed to trying to give birth to some set of reforms by then. And there seems to be a broad consensus that whatever emerges from the Working Group must be a living agreement, or, as one participant quipped privately, a “multilateral organism.”

What is less clear, however, is what form the organism will take. Should we expect a framework convention with six protocols? Twelve protocols? Will there be soft law instruments and guidelines for tribunals, in addition to or instead of the protocols? Will it be possible to join individual protocols without joining the framework?

A related question is how developed the organism will be by 2026, if it does emerge. Many living organisms emerge into the world with lots of features yet to develop. Babies enter the world after three trimesters because their heads are too large to fit through the birth canal after that, not because they are finished developing. Many international organizations, similarly, emerge when they do because political or resource constraints set a due date, not because they are perfectly formed at that time. Like babies, IOs continue to develop after they emerge; the history of the OECD or World Bank or UNCTAD is a story of an organism adapting to a changing environment, developing new features and shedding old ones.

What does this mean for participants at UNCITRAL? Many participants do not expect ISDS reform to end in 2026. Part of the Working Group’s discussion has been about how to create the means for updating or refining agreements or guidelines, as well as the means to continue developing and docking new reforms into the framework agreement, skeleton, or ‘Android operating system’ after 2026. The possibility for reforms to continue to emerge softens expectations – not everything has developed by 2026. And not everything necessarily has to emerge at UNCITRAL; an OECD process on substantive reform is scheduled to start soon, for instance, and outputs it generates could complement or fit into what emerges at UNCITRAL.

But there is an expectation that in 2026, officials take something home to their capitals and to the UNCITRAL Commission. What features need to be developed by then to meet the Working Group’s mandate?

What is Viable?

States provide different answers on the question of what reforms are needed to fulfil the mandate. Reforms that are essential to some states are not viable to others. 

We heard from a number of developing countries this week that inconsistency in how damages are calculated must be on the Working Group’s agenda in order to fulfil its mandate. After hearing some developed states suggest damages were outside the mandate because they were not procedural reforms, Sierra Leone derided the excuses being given for excluding damages, observing that for his delegation, and multiple others in the room, damages were of high priority and that there was no clear division that could be drawn between procedure and substance.

Sierra Leone also responded to states that suggest issues like damages have been dealt with in recent investment treaty practice. Developed states may have solved this problem because they have resources and bargaining power to do so, but UNCITRAL provides the opportunity to do that for everyone. Multilateralism matters – Sierra Leone and several other states feel they need to make reforms multilaterally in order to overcome obstacles to these reforms. They see the Working Group as an opportunity to solve problems, and feel the stakes are high because they don’t get this kind of opportunity often. For these states, the process needs to deliver meaningful results on issues like damages and third-party funding to meet its mandate.

Other states see the process and its mandate differently. In the eyes of the EU and its Member States, the Working Group needs to deliver structural reforms like an Appellate Mechanism and non-party appointment in order to fulfil its mandate. In their view, individual reforms like a Code of Conduct emerging in isolation, without the emergence of a broader structural reform package, are insufficient to solve the concerns identified by the Working Group in earlier stages of the process

In the eyes of some other states, however, a Code of Conduct emerging from the process but not much else would be an okay, or even desirable, outcome. Some states do not appear to be expecting or wishing for much to emerge from UNCITRAL; soft law instruments or guidelines to complement existing instruments might be helpful, but they are not crucial. These states are not necessarily opposed to sharing practices and learning multilaterally; although they believe states have an important role to play in managing their treaties, they often prefer to manage their own treaties bilaterally or unilaterally.

Feeling Overwhelmed

Differences of view about which ISDS reforms are desirable have long been a feature of this process. In response to these differences, the Working Group has adopted a plural approach and planned to develop many reforms, with the understanding that individual states will be able to choose which reforms they want to join later. While this approach made earlier stages of the process easier, it may make later stages harder.

One way that it may make later stages harder is logistical: developing many reforms simultaneously is difficult. This two-week period alone required participants to consider 10 working papers, listed in the table below, plus other preparatory materials, such as compilations of relevant provisions from select investment agreements or comments from experts collected by the Secretariat. The number and quality of these papers is a testament to the Secretariat and to the officials and experts who have given their time to the process. But the volume of proposals and materials can still create a sense of deluge among even well-prepared participants.

Table: Schedule of UNCITRAL Working Group III, 43rd Session, 5–16 September 2022

5 September

Standing multilateral mechanism: Selection and appointment of ISDS tribunal members and related matters

A/CN.9/WG.III/WP.213 – Standing multilateral mechanism: Selection and appointment of ISDS tribunal members and related matters


6 September

Advisory centre


A/CN.9/WG.III/WP.212 – Advisory Centre

A/CN.9/WG.III/WP.212/Add.1 – Advisory Centre

7 September

Multilateral instrument on ISDS reform

A/CN.9/WG.III/WP.221 – Multilateral instrument on ISDS reform 

A/CN.9/WG.III/WP.194 – Multilateral instrument on ISDS reform

8-9 September

Procedural rules and cross-cutting issues

·         Draft provisions on early dismissal, security for costs, allocation of costs, counterclaims, and third-party funding

·         Assessment of damages and compensation

A/CN.9/WG.III/WP.219 – Draft provisions on procedural reform  

A/CN.9/WG.III/WP.220 – Assessment of damages and compensation 

12-13 September


·         Draft provisions on mediation

·         Draft guidelines on mediation

A/CN.9/WG.III/WP.217 ­- Draft provisions on mediation

A/CN.9/WG.III/WP.218 ­- Draft guidelines on investment mediation

14-16 September

Code of Conduct

A/CN.9/WG.III/WP.216 – Draft Code of conduct

A second way in which the Working Group’s approach of developing many reforms may make later stages harder relates more to negotiating strategy and tactics. The Working Group’s time is not infinite. The due date is set and the workplan agreed upon last year outlined the way forward in terms of which reforms would be discussed when. The Working Group’s days remaining before 2026 are already numbered, and, as the Chair reminded delegates, they are all already spoken for. What happens if discussion of one reform spills over into another reform’s allocated time? 

It is possible that the Working Group will choose to prioritize certain reforms over others. But in a process that has already seen battles over time allocation, prioritization is unlikely to be easy. It is also possible that some delegates begin to feel that others are elongating discussions as a negotiating tactic to avoid or delay reforms they do not support.

To us it seems that actors have varying senses of urgency, depending on what they believe the Working Group needs to develop in order to meet its mandate. For actors that are not looking for a lot to come out of this process, there is no rush to produce outputs. Actors that would like more dramatic innovations or to bring more reforms to the UNCITRAL Commission in 2026 tend to feel a greater sense of urgency.

When Is It No Longer Premature?

Actors’ different senses of urgency appear in various discussions, but were especially apparent during debates on the multilateral instrument. At several points, the US stated that further development would be premature, while the EU pushed for the Secretariat to start drafting text, saying that there were models the Secretariat could look to for inspiration. For instance, paragraph 16 of Working Paper 221 discusses objectives for a multilateral instrument. Canada voiced its concern that setting out objectives without having landed content would be premature—a concern echoed by the US and Mexico. The EU then put forward its view that the Secretariat could surely start drafting text based on generally agreed upon objectives, such as the Sustainable Development Goals and concerns identified by the Working Group in earlier stages, such as consistency, coherence, and predictability.

In the context of the multilateral instrument, we heard interventions emphasizing the importance of states giving clear instructions to the secretariat at this point in the process. To give clear instructions sometimes requires making difficult decisions that states have so far put off. The Chair clearly has his eye on the clock. At one point, he cautioned that the group did not have much more time for study and that it was time to instruct the Secretariat on the way forward. The gestation period is fixed, at least for now, so the question is what reforms will actually develop during that time.

Learning to Compromise

As individual reforms enter their final developmental stages, it becomes more important for states to compromise. UNCITRAL Working Groups proceed iteratively; each session they proceed through papers or draft text and give the Secretariat instructions on what they want to see in the next round. Over a series of drafts, the text becomes more precise and more polished. Another result of this process, however, is that finding a compromise for particularly difficult issues is often left to the end.

Already in his welcome remarks to the Working Group, the Chair noted that he would be looking for delegations to find a compromise on difficult issues, such as Article four in the Code of Conduct. This Article concerns double hatting, or as the text refers to it, a limit on multiple roles. Last November, participants discussed a draft with three options: full prohibition, modified prohibition, and full disclosure. At the time, we observed that discussions on the double-hatting article were “the most passionate of the week” and that no definitive agreement was found but the Working Group agreed to move forward by requesting the Secretariat prepare a potential middle ground in the form of a partial prohibition coupled with a full disclosure requirement.

The Draft Code of Conduct prepared for this session developed this potential middle ground. Article 4, paragraph 1 was drafted:

Unless the disputing parties agree otherwise, an arbitrator shall not act concurrently [and within a period of three years following the conclusion of the IID proceeding] as a legal representative or an expert witness in another IID proceeding involving: (a) the same measure(s); (b) the same or related party (parties); or (c) the same provision(s) of the same treaty.

Paragraph 2 of the draft added that an arbitrator shall not act concurrently [and within a period of three years following the conclusion of the IID proceeding] as a legal representative or an expert witness in another IID proceeding “involving legal issues which are substantially so similar that accepting such a role would be in breach of Article 3.”

Discussion of this article kicked off with the UK arguing against any cooling off period and against the “substantially so similar” language, saying that the draft provision would create an overly strict and unnecessarily burdensome ban. Several other developed states then also voiced opposition to a cooling off period and to language so broad that it would create a de facto ban. Argentina intervened to stress that his country’s preference was for an outright ban and a ten-year cooling off period but that he was not pushing that precisely so that a compromise, middle ground approach could be found. The Chair reminded delegates of their previous discussions and compromise, adding that now the question was now how to operationalize that compromise. Nobody would get their exact preferences, but the room needed to move toward the middle in order to move forward.

Would there be consensus around particular solutions? And how is consensus operationalized? At one point, Iran objected to some language that was in square brackets being taken out of square brackets. If Iran objected, the delegate said, there was no consensus and the process could not move forward without a vote. The Chair responded that this was not how progress was made at UNCITRAL, noting that this type of blocking would only frustrate efforts and the weight of the room was to remove the brackets. Since there was no need to finalize this particular point, the exchange ended with the language staying in square brackets for now, but to us it underlined that softer or harder versions of consensus norms exist and that different understandings of how consensus operates may become contentious at some point.

At the end of another session, it is clear that the process is moving forward – the people and margins are back and reforms are gestating. But what should we expect when the Working Group is expecting? To us, it seems that finding ways to cope with a potentially overwhelming reforms-to-develop list and finding compromises will become increasingly important. How participants handle these challenges may shape whether a multilateral organism emerges and what form it takes. But it is too soon for prediction, there are still four long years of gestation left until the 2026 due date…

Print Friendly, PDF & Email

Leave a Comment

Comments for this post are closed