Scheduling Is the Ace in the Hole

by Carlo M. Cantore[1] & Petros C. Mavroidis[2]

 1. Consensus as Insurance Against the Waning of MFN

The WTO membership has been mulling over the legal status of completed JSIs (Joint Statement Initiatives). JSIs are not multilateral agreements. They could become plurilateral agreements, if the membership (including those who did not participate in their negotiation and are not willing to join) agrees by consensus to add them to the existing list (Article X:9 of the Agreement Establishing the WTO). This is unlikely. Take the Agreement on Investment Facilitation for Development (IFDA). Over 75% of the WTO Members are seeking to add it to the WTO rulebook as a plurilateral. They have presented a proposal in this sense at 11 consecutive meetings of the General Council only to see the adoption of a draft decision blocked by naysayers. What to do then? 

Legislative initiatives rest on two legs: intellectual- and democratic legitimacy. The two are not necessarily interconnected. Intellectually sound proposals risk being ditched in the name of selfish (and often unrelated to the issue) quid pro quos. The WTO regime reflects the apotheosis of democratic legitimacy: consensus. Is it sensible? Not necessarily. 

Pacta tertiis nec nocent nec prosunt is a fundamental legal maxim of international law meaning “treaties neither harm nor benefit third parties”. This principle is codified in Article 34 of the Vienna Convention on the Law of Treaties. How does this maxim work in the realm of multilateral agreements, like the WTO, the foundational principle of which is MFN? Think of an agreement between a sub-set of the membership. Think of a JSI, for example. Any agreement between a few WTO members only risk affecting rights of non-participants. In fact, even if advantages agreed between a sub-set of the membership have been extended on MFN-basis to non-participants, the same risk still exists: what good is it to non-participants to extend to them tariff reductions of classifications at the eight- or ten-digit level which reflect the comparative advantage of the negotiating parties? Or, what is the advantage of a commitment to trade only in paperless form for non-participants that cannot finance this transition? 

Similar considerations probably kept the consensus rule within the WTO legal edifice, even though Article II:3 of the WTO Agreement makes it clear that plurilaterals “do not create either obligations or rights for Members that have not accepted them”. Subjecting the addition of new plurilaterals to the rule of consensus functions, effectively, as insurance against deviating from MFN. 

2. How Should we Understand Consensus?

Consensus thus, in the WTO world, is not an end in itself. It is a means to protect MFN rights. But what exactly is the nature of the rights of WTO members? Is it to see the contract performed (“property” rules), or to be compensated in case of violation (“liability” rules)? Paradoxically, the WTO regime is not clear about it. 

Article XXVIII GATT allows a WTO Member to increase its level of duties in a given product-market, even in the absence of agreement between the Member modifying its concession and affected parties. The latter nonetheless, remain free to retaliate by increasing duties in equivalent manner. This provision implicitly endorses “liability” rules. On the other hand, Article 22.4 of the DSU clearly acknowledges the primacy of “property” rules. In a non-litigating context thus, “liability” rules are very much a possibility. This is a very important observation for the case we discuss later on, under Section 4. 

Before going any further though, we should note that the WTO membership has enjoyed a long experience of agreements between a subset of the membership. The WTO Secretariat issued in 2005 a paper entitled “Sector Specific Discussions and Negotiations on Goods in the GATT and the WTO”, which discusses all that since the advent of the GATT.[3] The one lesson emerging when reading the historical record, is that trading nations were not afraid to improvise in order to enforce agreements between a sub-set of the membership, while paying heed to rights of non-signatories. 

3. Scheduling vs. Consensus

Because scheduling is not harmonized across commitments in goods and services trade, we distinguish between the two. 

3.1 Trade in Goods

Hoekman & Mavroidis (2017)[4] explain under what conditions the requirement for consensus-based decisions could be circumvented through scheduling of commitments in the realm of trade in goods. In short, WTO Members who wanted to implement commitments additional to those embedded already in their schedules, could do so by incorporating agreements into Part III (entitled “Non-Tariff Concessions”) of their GATT schedules of concessions. Consensus is not required as long as deals were applied on a non-discriminatory basis and did not undercut existing rights of non-signatories. 

But what if there is disagreement regarding the non-discriminatory nature of commitments entered, or whether they undercut existing rights of non-signatories? In the goods-context, there is no definitive response in the body of Article XXVIII GATT. The Member(s) concerned can still go ahead and modify its/their concession, and those affected have the right to retaliate. It is unclear whether retaliation should observe MFN or not (even though better arguments side with the latter option). A comprehensive survey of practice prepared by Hoda (2001)[5] suggests that in practice trading nations did not reach the stage of retaliation and counter-retaliation in the realm of goods-trade. 

3.2 Trade in Services

Members wishing to include new commitments in their services schedule may do so following the rules laid down in a 2000 Decision by the Council for Trade in Services (CTS), which adds detail to Article XXI of GATS, the provision making modification of schedules agreed possible.[6] 

How does it work? A Member seeking to modify its existing GATS schedule by adding new commitments shall circulate its proposal to all Members. If no objection is raised by self-identified affected Members within 45 days, the Secretariat shall inform Members that the new schedule is certified. In the event of an objection, the modifying Member and the objecting Member shall enter into consultations. If these lead to an agreement and the objection is withdrawn within an additional 45-day term, the new schedule can be certified. 

If the objection persists, the procedures for the certification of improvements of schedules cease to apply, and the modifying Member may either give up, or follow the rules for the modification of schedules in Article XXI of the GATS, as complemented by a 1999 CTS Decision.[7] The GATS regime for the modification of schedules aims at keeping the originally negotiated balance, since concessions are reciprocal: following the modification(s) entered, trade between the modifying Member and the rest of the Membership should be no less favourable when compared to the situation that existed before the modification. In the event of persisting disagreement, the advantage of the GATS regime (over the GATT regime) is that it offers Members the possibility to resort to binding arbitration. Before the arbitration body, the burden rests on the objecting Member to show that its position is made worse off by the proposed modification.[8] 

Suppose a Member proposes to withdraw an existing commitment in a given sector. In that case, it would have to offer additional concessions in other sectors in order to keep the balance. The arbitration body will decide on the amount of compensation due at last resort. 

4. Applying the Analysis Above to the Ongoing JSI Discussions

Scheduling is the path already taken by participants to the JSI on Services Domestic Regulation (SDR). To recall, the outcome of the SDR negotiations is a reference paper codifying certain good regulatory practices in the fields of licensing, qualification requirements, and technical standards. SDR participants initially circulated proposals to update their services schedules, adding language in the column for “Additional Commitments” (Article XVIII of GATS) in order to incorporate the SDR disciplines by means of a simple cross-reference to the document symbol of the SDR reference paper. In most cases, based on a deal spearheaded by the European Union, on one side, and India and South Africa, on the other, SDR participants amended their initial proposal and decided to append the text of the SDR reference paper in its entirety to their schedules. They also inserted language in the accompanying communications indicating that the certification processes would not serve as precedent for the incorporation of other JSI outcomes in Members’ schedules.[9] Australia refused to join, and kept its proposed modification as a simple cross-reference to the SDR reference paper, with no additional language in the accompanying communication. Eventually, India requested arbitration. The arbitration body found that India had not demonstrated that Australia’s proposed modification altered the balance of its schedule in a manner that was less favourable to trade for India.[10] 

What matters to our discussion is that, in the realm of services, modification of schedules is effectively addressed through the arbitration process that has been put in place. Could then scheduling help in the context of the IFDA and the E-Commerce Agreement? Our answer is yes. Minor modifications would be required to accommodate the institutional and dispute settlement provisions of the IFDA and the E-Commerce Agreement. 

Regarding the IFDA, to the extent that the disciplines affect trade in goods and services, they could be transposed in, respectively, Part III of a Member’s GATT Schedule and as horizontal commitments in a Member’s GATS Schedule. Mamdouh (2021) warns that the IFDA has a broader scope compared to the GATT and the GATS, and therefore any new disciplines included in the GATT or GATS schedules would only be binding insofar as it concerns matters within the scope of application of those two agreements.[11] To the extent of course, that the IFDA includes issues beyond the scope of GATT and GATS, no questions regarding their consistency with WTO law can be raised. 

Turning to the E-Commerce Agreement, it also includes disciplines that affect trade in both goods and services. We will take the worst-case scenario here, and assume that the Agreement on E-Commerce includes provisions that do not observe MFN. Our discussion rests on two points. 

First, as we have explained above, liability rules apply here. As long as affected members have been compensated (assuming always that the modification of schedules through the incorporation of the Agreement on E-Commerce negatively affects the MFN rights of non-signatories), they can request no more. 

Second, what if compensation has not been agreed between signatories and non-signatories? As we explained above, while there remains some uncertainty regarding trade in goods, the path is clearer in the domain of services. Members who consider to be “affected” by the proposed new rules would have to demonstrate, in the context of an arbitration, how the level of commitments in the new schedule is less favourable to trade compared to the status quo ante before the modification. 

5. Conclusion

The inclusion of the IFDA and the E-Commerce Agreement in the WTO rulebook as plurilaterals under Annex IV would be the preferred solution. However, it is unlikely to happen anytime soon. It is therefore time for creative solutions, if Members want the agreed discipline to enter into force. We would like to emphasize that we are not suggesting a shortcut that would give carte blanche to a subgroup of Members to do as they please. Our solution observes WTO law as is. Schedules are integral part of, respectively, the GATT and the GATS. Disciplines embedded in the schedules need to be applied, in principle, on an MFN basis. In this respect, scheduling the outcomes of JSIs is superior to some of the alternatives currently being discussed. Schedules offer guarantees of publicity and the concessions and commitments embedded therein must be non-discriminatory, legally binding and enforceable. In the event that new disciplines were to undermine the value of existing concessions and/or commitments, compensation would be owed to affected Members, and rules on trade in services even establish an arbitration process should Members fail to agree on the amount of compensation that is due.[12] The same guarantees would not necessarily be present in the event that the Members concerned were to opt for soft law instruments (gentlepersons’ agreement) or full-fledged international agreements outside of the WTO framework.

[1] International trade lawyer.

[2] Edwin B. Parker Professor of Law, Columbia University in New York.

[3] WTO Doc. TN/MA/A/S/13 of 24 January 2005.

[4] Bernard M. Hoekman & Petros C. Mavroidis. 2017. MFN Clubs and Scheduling Additional Commitments in the GATT: Learning from GATS, European Journal of International Law, 28: 387-407.

[5] Anwarul Hoda. 2001. Tariff Negotiations and Renegotiations under the GATT and the WTO, Cambridge University Press: Cambridge, United Kingdom.

[6] WTO Doc. S/L/84 of 18 April 2000.

[7] WTO Doc. S/L/80 of 29 October 1999.

[8] WTO Doc. S/SECRET/13/ARB/F of 22 November 2024, para. 3.25.

[9] WTO Doc. S/WPDR/M/79 of 2 February 2024, paras. 1.33-1.36.

[10] WTO Doc. S/SECRET/13/ARB/F of 22 November 2024, para. 4.1.

[11] Hamid Mamdouh. 2021. Legal options for integrating a new Investment Facilitation Agreement into the WTO Structure, available at: https://www.intracen.org/file/iflegaloptionsmamdouhasofsept17finalpdf (last accessed, 22 December 2025).

[12] We should note that there is no appeal possible against the arbitration process provided for in Article XXI of GATS. We thus, do not run the risk of “appeals into the void” and the ensuing question marks regarding the legitimacy of the arbitral award.