by Chakravarthi Raghavan

Geneva, 12 May 2000 -- The ruling by a three-member panel in a dispute under a plurilateral agreement appears likely to open the way for WTO members, and developing countries in particular, to be dragged before panels through an expanded interpretation of the scope of 'non-violation' complaints at the World Trade Organization.

The ruling has been handed down in a dispute under the Government Procurement Agreement (GPA), a plurilateral agreement of which few developing countries are signatories, which has dispute settlement provisions on lines of those in GATT 1994. And the rules of the Dispute Settlement Understanding (DSU) apply to the settlement of the disputes.

The dispute arose over a complaint by the United States that in respect of the procurement for the Inchon International Airport building project, Korea had violated its obligations and commitments under the GPA.

The US claim was that Korea's Ministry of Construction and Transportation (MOCT), including the New Airport Development Group (NADG) under MOCT, the Korean Airport Authority (KAA), the Korean Airport Construction Authority (KOACA), and the Inchon International Airport Authority (IIAC), were all or have been in the past, Korean central government entities involved in the procurement for the IIA project, and covered under Korea's schedule of entities in Annex I of the Korean Appendix 1 of the GPA schedule of commitments. And, by imposing bid deadlines for receipt of tenders and other conditions, Korea had violated its obligations.

The panel held that on the basis of Korea's schedule, and a note explaining it, the Korean commitment to open procurement to outside bidding did not cover the Korean Airport Authority or its successors (including IIA).

The panel then dealt with the US non-violation complaint argument that the Korean measures in respect of the bidding and contracts for the IIA (inadequate bid deadlines, imposition of certain qualification requirements, and some domestic partnering requirements, and no domestic procedures for challenges) violated the "reasonable expectations" of the USA arising out of the GPA and Korea's accession to it.

In addressing this, the three-member panel (chaired by Michael D. Cartland, former Hong Kong Representative to GATT, and comprising Ms. Marie-Gabriella Ineichen-Fleisch and Mr. Peter Armin Trepte as members), has expanded the traditional, but seldom invoked or used, GATT concept of nullification or impairment of "reasonable expectations" of a benefit accruing from a negotiated concession and agreement (Art. XXIII.1(b) of GATT 1994).

In the 50-year history of the GATT and WTO, there have been only eight cases of non-violation disputes -- as the US-Japan (Kodak/Fuji) film panel (where the US unsuccessfully used the non-violation route to challenge Japan's private retail distribution arrangements and trade) had pointed out in its report

The Cartland panel has staked out the view that the non-violation remedy route can be used to claim impairment or nullification to "reasonable expectation" of benefit, not only for lack of good faith in implementation, but lack of good faith in negotiations or 'treaty error', and the DSU could be used to get around the problems (in view of the Marrakesh Final Act) of invalidating a part of the treaty, by a ruling, and DSB recommendation, to enable a party to withdraw reciprocal concessions.

Can the principle of good faith implementation of a treaty (Art 26 of the Vienna Law of Treaties, incorporating the principle of pacta sunt servanda) and the concept of non-violation arise in contexts other than the traditional approach represented by the pacta sunt servanda principle? Can, for instance, the question of error in treaty negotiation be addressed under Art.26 of the DSU and Art.XXII.2 of the GPA (the dispute settlement provisions using similar language to that of Art.XXIII.1.(b) of GATT 1994)?

"We see no reason why it cannot," says the Cartland panel (in para 7.100) in a sweeping assertion of the right to invoke 'non-violation' complaints. "Parties have an obligation to negotiate in good faith. It is clear to us (as discussed in para 7.110 and 7.121) that it is necessary that negotiations in the Agreement before us (the GPA) be conducted on a particularly open and forthcoming basis."

Though ultimately, the panel found against the US even on this issue, on the ground that the US had not exercised due care in the negotiations, the ruling appears likely to have some very serious consequences for developing countries across the whole range of WTO agreements and future non-violation disputes relating to them.

It will have implications for future disputes of 'non-violation" complaints -- Art.XXIII.1.(b) of GATT 1994, which hitherto was considered as encompassing only trade concessions exchanged, but whose effect is frustrated by other measures of the party making the concession -- and the relevant provisions of the GATS and TRIPS.

But even more, it is also likely to have serious implications if the EC-US drive for a 'transparency in government procurement agreement at the WTO' succeeds, beginning with a commitment of all countries to transparency in all aspects of government procurement. While the US and EC have been telling Third World negotiators that such a transparency accord does not mean market access, their own internal documents and briefings for their TNCs show that they are presenting the transparency in governnment procurement agreement at the WTO as a beginning for full market access for their TNCs in the lucrative Third World government procurement markets.

The US complaint against South Korea was that the South Korean commitments under the GPA extended to the IIA project, and that Korea violated its obligations under GPA in regard to the award of these contracts by the IIA.

While rejecting this US view that the IIA, as an entity, was covered by South Korea's commitments, under Appendix I of the GPA, the panel has taken up the 'non-violation' complaint, and has used some circuitous arguments to extend the scope of such complaints.

Article 3.2 of the DSU asks panels to clarify existing provisions of agreements in accordance with customary rules of interpretation of public international law. And the Appellate Body has said that this means the Vienna Law of Treaties -- according to which treaties have to be interpreted by using the ordinary meaning of the words used, and failing that, use of the contexts etc which may involve the negotiating history (approved before or after by all the parties).

But to interpret the scope of non-violation complaints, the Cartland panel has delved into the negotiating history, not of the WTO or its covered agreements, but the Vienna Law of Treaties itself!

The panel says (para 7.87): "A key difference between a traditional non-violation case and the present one would seem to be that, normally, the question of 'reasonable expectation' is whether or not it was reasonably to be expected that the benefit under an existing concession would be impaired by the measures. However here, if there is to be a non-violation case, the question is whether or not there was a reasonable expectation of an entitlement to a benefit that had accrued pursuant to the negotiation rather than pursuant to a concession."

The GATT jurisprudence in non-violation complaints (which has traditionally arisen in the context of actions which might undermine the value of negotiated tariff concessions), the panel argues, is on the basic premise that Members should not take actions, even those consistent with the letter of the treaty, which might serve to undermine the reasonable expectations of negotiating partners.

This, the panel says, is a further development of the principle of pacta sunt servanda... this principle is expressed in Article 26 of the Vienna Law of Treaties: "Every treaty in force is binding upon the parties to it and must be performed by them in good faith."

It seems clear, the Cartland panel adds, that good faith performance has been agreed to by WTO members to include subsequent actions which might nullify or impair benefits reasonably expected to accrue to other parties to the negotiations.

The panel then delves into the 'negotiating history' of the Vienna Law of Treaties, and cites from the report of the International Law Commission to the UN General Assembly: 'Some members felt that there would be advantage in also stating that a party must abstain from acts calculated to frustrate the object and purpose of the treaty. The Commission however considered that this was clearly implicit in the obligation to perform the treaty in good faith and preferred to state the pacta sunt servanda rule in as simple a form as possible.'

Since the wording, and thus the meaning, of Art. 26 of the Vienna Law of Treaties is clear and without any ambiguity, and 'performance' necessarily involves subsequent actions, why does the panel then have to delve into the negotiating history of that treaty to explain and justify this meaning?

The panel continues (para 7.96), "However, the relationship of the WTO agreements to customary international law is broader than this. Customary international law applies generally to the economic relations between WTO members. "Such international law applies to the extent that the WTO treaty agreements do not 'contract out' from it. To put it another way, to the extent there is no conflict or inconsistency, or an expression in a covered WTO agreement that implies differently, we are of the view that the customary rules of international law apply to the WTO treaties AND TO THE PROCESS OF TREATY FORMATION UNDER THE WTO." (emphasis added).

The Japan-US film panel had pointed out that both the GATT contracting parties and WTO Members "have approached the (non-violation) remedy with caution and, indeed, have treated it as an exceptional instrument of dispute settlement," but that remedy had an important role - that of protecting the reasonable expectations of competitive opportunities through negotiated concessions.

The Cartland panel says (in para 7.99) that in its view, these observations are entirely in line with the concept of pacta sunt servanda. It then adds: "the vast majority of actions taken by Members which are consistent with the letter of their treaty obligations will also be consistent with the spirit. However, upon occasion, it may be the case that some actions, while permissible under one set of rules (e.g. the Agreement on Subsidies and Countervailing Measures is a commonly referenced example of rules in this regard), are not consistent with the spirit of other commitments such as those in negotiated schedules... That is, such actions deny the competitive opportunities which are the reasonably expected effect of such commitments..."

Negotiated schedules (of concessions) are in country schedules annexed under Art.II of GATT 1994, and Art.III (for national treatment between imported and like domestic products) seeks to ensure the competitive opportunity of the negotiated concessions.

But any conflict between negotiated schedules and the Subsidies Agreement for e.g. was sought to be settled by the negotiators through the General Interpretative Note to Annex IA, specifically incorporated into the WTO by negotiators, to make clear that: "In the event of conflict between a provision of the General Agreement on Tariffs and Trade 1994 and a provision of another agreement in Annex 1A to the Agreement Establishing the World Trade Organization (referred to in the Agreements in Annex 1A as the 'WTO Agreement'), the provisions of the other agreement shall prevail to the extent of that conflict."

But the panels have been busy making this provision 'inutile' -- beginning with the Indonesia automobile case, where the conflicts between GATT 1994 and the subsidies agreement, was reconciled by reading the TRIMS reference to Article III to mean 'substance' of Article III -- and making the WTO agreement obligations cumulative.

The Cartland panel now says (para 7.100):

"One of the issues that arises in this dispute is whether the concept of non-violation can arise in contexts other than the traditional approach represented by pacta sunt servanda. Can, for instance the question of error in treaty negotiation be addressed under Article 26 of the DSU and Article XXII:2 of the GPA? We see no reason why it cannot. Parties have an obligation to negotiate in good faith just as they must implement the treaty in good faith. It is clear to us (as discussed in paragraphs 7.110 and 7.121 below) that it is necessary that negotiations in the Agreement before us (the GPA) be conducted on a particularly open and forthcoming basis."

The panel delves into the history of the 'negotiations' for exchange of concessions between the US and Korea for the latter's application for (a) accession to the Tokyo Round code and, (b) subsequently, to joining the GPA - the questions posed by the US and Korea's responses, and whether or not Korea gave full, and 'good faith', answers to the questions whether its offer would cover new airport constructions.

The panel then says (para 7.119) that Korea's answers to the US questions was 'insufficient', that Members have a right to expect full and forthright answers to their questions submitted during negotiations, particularly with respect to schedules of affirmative commitments such as those in the GPA.

"However, Members must protect their own interests as well and in this case, the United States did not do so. It had a sufficient amount of time to realize particularly in light of the wide knowledge of the KAA's role, that its understanding of Korean answer was not accurate. Therefore, we find that, even if the principle of a traditional non-violation case were applicable to this situation, the United States has failed to carry its burden of proof to establish that it had reasonable expectations that a benefit had accrued."

The panel did not stop with this finding, however.

In subsequent paragraphs (7.120 to 7.128) of its report, under the sub-heading "Error in Treaty Formation", it delves into the issues of bad faith in negotiations and consequences for a treaty, and tries to use it to stretch the concept beyond the actual wordings of Art. 48 of the Vienna Law of Treaties.

After outlining what it says are good faith obligations in negotiations in the GPA (paras 7.121-122), the panel says that the 'error in respect of a treaty' concept has developed in customary international law through the case law of the ICJ (rulings in 1933 and 1962, before the International Law Commission was asked by the Assembly to undertake the codification that resulted in the Vienna Law of Treaties).

The Cartland panel says that while these cases were concerned primarily with the question in which circumstances error cannot be advanced as a reason for invalidating a treaty, it is implicitly accepted that error can be a ground for invalidating (part) of a treaty. The panel acknowledges that these principles have been codified by the International Law Commission in what has become the Vienna Law of Treaties of 1969, and cites Art.48 of that treaty:

1. A State may invoke an error in a treaty as invalidating its consent to be bound by the treaty if the error related to a fact or situation which was assumed by that State to exist at the time when the treaty was concluded and formed an essential basis of its consent to be bound by the treaty.

2. Paragraph 1 shall not apply if the State in question contributed by its own conduct to the error or if the circumstances were such as to put that State on notice of a possible error.

Since this article is derived from case law of the International Court of Justice (and its predecessor at the Hague), this presently represents customary international law and "we will apply it to the facts of the case," says the Cartland panel.

It then says that though it has held that there was a duty (on the part of Korea) to demonstrate good as the "offering" party, it was for the partners to exercise due diligence to verify these offers and, the facts recounted by the panel (in paras 7.104 to 7.119) demonstrated that the US had not properly discharged this burden, but adds:

(para 7.125) "...We do not think the evidence at all supports a finding that the US has contributed by its own conduct to the error, but given the elements mentioned earlier (the 2-1/2 year interval between Korea's answer and the final offer, the actions by the EC - in excluding Korea's ability to bid on airport project contracts in the EU, because of the Korean offer not including IIA -- and the subsequent 4-month period for verification of offers), we conclude that the circumstances were such as to put the United States on notice of a possible error." The US error was hence not excusable, and only an excusable error could vitiate the consent to be bound by the agreement.

"For these reasons," adds the panel (para 7.126), "on balance we are of the view that the US has not demonstrated error successfully as a basis for a claim of non-violation nullification or impairment of benefits."

But tucked into a footnote (F.N.81), the panel says: "A finding of justifiable error in treaty formation might normally be expected to lead to the application of Article 65 of the Vienna Convention. However, Article 65 on the specific procedures for invoking validity of a treaty does not seem to belong to the provisions of the Vienna Convention which have become customary international law.... The Article on separability (Art.41) raises the possibility that provisions may be separated, such as e.g. separate reciprocal concessions in schedule, if they do not form an essential basis for the consent of the other party of the treaty as a whole..... We do not think that any of the provisions would be required with respect to non-violation under the WTO Agreements because Article 26 of the DSU clearly provides for the appropriate remedy."

Article 26.1 (b) makes clear that a measure found to nullify or impair, but without violating an agreement, cannot be asked to be withdrawn. The panel or Appellate Body could only recommend the Member concerned to make a satisfactory adjustment. And 26.1.(c) enables an arbitrator under 21.3 to fix the level of benefit nullified or impaired and suggest also, a non-binding, suggestion for mutually satisfactory adjustment. And 26.1 (c) enables compensation to be a final settlement of the dispute.

In theory, panel and Appellate Body rulings, and the DSB recommendations flowing from them, create no new rights or obligations. But they are increasingly cited by subsequent panels not only for reaching a similar decision (on the same or another issue), but to stretch it too. And Third World trade experts and observers have been noting that a tentative view advanced in one ruling, is then pushed further in subsequent rulings (all adopted by the Dispute Settlement Body under the negative consensus decision-making rule).

And while the panellists may be different, there is the common hand of the secretariat behind them.

The overall result is of increasing the obligations of developing countries, and tightening neo-liberal dogma for the benefit of transnational corporations.

This raises the question whether the dispute settlement system, set up to settle disputes arising out of the rules negotiated by Members, has now become a runaway system? And if WTO member governments don't or can't rein in the system, would it not result in civil society actions?

The explanation of some trade officials for the panels course is that in negotiating and concluding agreements, and in an effort to reach consensus, compromise language has been used without clear meaning provided, and hence the panels are forced to look outside to find meanings, interpret the rules and settle disputes.

International law has its sources in international treaties, and principles from the various systems of law in the world. While the Anglo-Saxon and Continental systems are prominent, there are also other sources. But even in the Anglo-Saxon system of law, where courts in some cases can strike down a law or a portion of it, the courts cannot legislate and at the most invite the legislature to exercise its prerogatives and legislate. But if judges overstep their limits, there are well-known remedies, not easy at the WTO.

And given the last sentence in DSU Art. 3.2 ('Recommendations and rulings of the DSB cannot add to or diminish the rights and obligations provided in the covered agreements'), the remedy for the panels, when there is ambiguity in language used, would seem to lie in the panel, referring the issue back to the WTO's legislative body, the General Council, and seeking an agreed interpretation and guidance to the panel. Instead the panels and the secretariat are writing law.

Without serious review of the DSU, and specific changes in the rules, trade diplomats may find it difficult to arrest this extension of jurisdiction and creation of new obligations by the dispute panels.

But at the minimum, developing countries need to pause and think before agreeing to negotiate and enter into new commitments and obligations, as Third World trade experts have been arguing. (SUNS4670)

The above article first appeared in the South-North Development Monitor (SUNS) of which Chakravarthi Raghavan is the Chief Editor.

[c] 2000, SUNS - All rights reserved. May not be reproduced, reprinted or posted to any system or service without specific permission from SUNS. This limitation includes incorporation into a database, distribution via Usenet News, bulletin board systems, mailing lists, print media or broadcast. For information about reproduction or multi-user subscriptions please contact < >