Max Planck Institute for Comparative Public Law and International Law Logo Max Planck Institute for Comparative Public Law and International Law

You are here: Publications Archive World Court Digest

World Court Digest

I. Substantive International Law - First Part
7.5. Application

¤ Gabcíkovo-Nagymaros Project
(Hungary / Slovakia)
Judgment of 25 September 1997
I.C.J. Reports 1997, p. 7

[pp. 51-53] 67. Czechoslovakia had maintained that proceeding to Variant C and putting it into operation did not constitute internationally wrongful acts; Slovakia adopted this argument. During the proceedings before the Court Slovakia contended that Hungary's decision to suspend and subsequently abandon the construction of works at Dunakiliti had made it impossible for Czechoslovakia to carry out the works as initially contemplated by the 1977 Treaty and that the latter was therefore entitled to proceed with a solution which was as close to the original Project as possible. Slovakia invoked what it described as a "principle of approximate application" to justify the construction and operation of Variant C. It explained that this was the only possibility remaining to it "of fulfil1ing not only the purposes of the 1977 Treaty, but the continuing obligation to implement it in good faith". ...

75. With a view to justifying those actions, Slovakia invoked what it described as "the principle of approximate application", expressed by Judge Sir Hersch Lauterpacht in the following terms:

"It is a sound principle of law that whenever a legal instrument of continuing validity cannot be applied literally owing to the conduct of one of the parties, it must, without allowing that party to take advantage of its own conduct, be applied in a way approximating most closely to its primary object. To do that is to interpret and to give effect to the instrument - not to change it." (Admissibility of Hearings of Petitioners by the Committee on South West Africa, separate opinion of Sir Hersch Lauterpacht, I.C.J. Reports 1956, p. 46.)

It claimed that this is a principle of international law and a general principle of law.

76. It is not necessary for the Court to determine whether there is a principle of international law or a general principle of law of "approximate application" because, even if such a principle existed, it could by definition only be employed within the limits of the treaty in question. In the view of the Court, Variant C does not meet that cardinal condition with regard to the 1977 Treaty.

[pp. 76-77] 132. In this regard it is of cardinal importance that the Court has found that the l977 Treaty is still in force and consequently governs the relationship between the Parties. That relationship is also determined by the rules of other relevant conventions to which the two States are party, by the rules of general international law and, in this particular case, by the rules of State responsibility; but it is governed, above all, by the applicable rules of the 1977 Treaty as a lex specialis.

133. The Court, however, cannot disregard the fact that the Treaty has not been fully implemented by either party for years, and indeed that their acts of commission and omission have contributed to creating the factual situation that now exists. Nor can it overlook that factual situation - or the practical possibilities and impossibilities to which it gives rise - when deciding on the legal requirements for the future conduct of the Parties.

This does not mean that facts - in this case facts which flow from wrongful conduct - determine the law. The principle ex injuria jus non oritur is sustained by the Court's finding that the legal relationship created by the l977 Treaty is preserved and cannot in this case be treated as voided by unlawful conduct.

What is essential, therefore, is that the factual situation as it has developed since l989 shall be placed within the context of the preserved and developing treaty relationship, in order to achieve its object and purpose in so far as that is feasible. For it is only then that the irregular state of affairs which exists as the result of the failure of both Parties to comply with their treaty obligations can be remedied.

134. What might have been a correct application of the law in 1989 or 1992, if the case had been before the Court then, could be a miscarriage of justice if prescribed in l997. The Court cannot ignore the fact that the Gabcikovo power plant has been in operation for nearly five years, that the bypass canal which feeds the plant receives its water from a significantly smaller reservoir formed by a dam which is built not at Dunakiliti but at Cunovo, and that the plant is operated in a run-of-the-river mode and not in a peak hour mode as originally foreseen. Equally, the Court cannot ignore the fact that, not only has Nagymaros not been built, but that, with the effective discarding by both Parties of peak power operation, there is no longer any point in building it.

135. As the Court has already had occasion to point out, the 1977 Treaty was not only a joint investment project for the production of energy, but it was designed to serve other objectives as well: the improvement of the navigability of the Danube, flood control and regulation of ice-discharge, and the protection of the natural environment. None of these objectives has been given absolute priority over the other, in spite of the emphasis which is given in the Treaty to the construction of a System of Locks for the production of energy. None of them has lost its importance. In order to achieve these objectives the parties accepted obligations of conduct, obligations of performance, and obligations of result.

136. It could be said that that part of the obligations of performance which related to the construction of the System of Locks in so far as they were not yet implemented before 1992 - have been overtaken by events. It would be an administration of the law altogether out of touch with reality if the Court were to order those obligations to be fully reinstated and the works at Cunovo to be demolished when the objectives of the Treaty can be adequately served by the existing structures.

[pp. 127-128 S.O. Bedjaoui] 30. The theory of "approximate application" or "close approximation relied on by Slovakia in order to justify the construction and commissioning of Variant C is unconvincing. There is no such theory in international law. The "precedents" advanced in favour of this theory are worthless. At least because of its dangers, this theory deserved wholehearted censure, which 1 find lacking in the Judgment.

31. Were this theory to be accepted, it would be to the detriment of legal certainty in relations between States and in particular of the certainty of treaties and of the integrity of the obligations properly entered into. The consolidation of this theory would virtually signal the end of the cardinal principle pacta sunt servanda, since a State which undertakes a specific obligation is left free to fulfil another, which it would be quite cunning to present as being very close to the first obligation. The State would only have to observe that its "approximate application " was allowed since, according to it, the conduct of the other party placed it in the impossibility of performing its obligations under the treaty and since it had no other remedy. All breaches of the obligations of the State would thus run the risk of being presented as an "approximate application".

The danger is all the greater in that this theory provides no reliable criterion for measuring the tolerable degree of "proximity" or "approximation ". The "distance" - or the "difference" - which a State would be authorized to take in relation to the purpose of a treaty when performing the obligation remains dangerously undefined and is still left to the subjective evaluation of the State.
But this is not all.
32. What the theory of "approximate application" lacks in order to be a valid "reinterpretation" of the treaty is quite obviously the basic condition of the consent of the other State. Indeed Slovakia is not wrong in stating that deviations from treaty norms in the application of the Treaty may be considered a "reinterpretation" of that Treaty. Yet this species of "mutation" or "novation" of the obligation in its performance is subject to the existence of an essential condition which has not been fulfilled in the present case at all. The "approximate application" may only be recognized as valid and may only constitute a "reinterpretation" if the other party to the Treaty has given its consent. The weakness of Slovakia's case is only too apparent.

Moreover Hungary's position is a most distinctive one since not only did it not give its consent to the "reinterpretation" of the Treaty, it also considers that there was neither an original interpretation nor a re-interpretation of the Treaty since for Hungary it ceased to exist even before the advent of Slovakia.

[pp. 148-149 S.O. Koroma] The unilateral suspension termination of the Treaty and the works for which Hungary was responsible under it had amounted not only to a repudiation of the Treaty; it frustrated the realization of the Project as a single and operational system of works, jointly owned and used for the benefit of the contracting parties in equal measure. As a result of Hungary's acts, the objective of the original Project could only have been achieved by Slovakia alone operating it; according to the material before the Court, Variant C constituted the minimum modification of the original Project necessary to enable the aim and objective of the original Project to be realized. It should be recalled that but for the suspension and abandonment of the works, there would have been no Variant C, and without Variant C, the objective of the act of Hungary which the Court has qualified as unlawful would have been realized thus defeating the object and purpose of the Treaty. In my view Variant C was therefore a genuine application of the Treaty and it was indispensable for the realization of its object and purpose. If it had not proceeded to its construction, according to the material before the Court, Czechoslovakia would have been stranded with a largely finished but inoperative system, which had been very expensive both in terms of cost of construction and in terms of acquiring the necessary land. The environmental benefits in terms of flood control, which was a primary object and purpose of the Treaty, would not have been attained. Additionally, the unfinished state of the constructions would have exposed them to further deterioration through continued inoperation.

[p. 174 D.O. Ranjeva] On consideration, and contrary to the analysis in the Judgment, the unlawfulness of Czechoslovakia's conduct cannot be limited to the mere putting into operation of the "provisional solution" because of the status of the Danube in international law. I cannot subscribe to the idea that territorial sovereignty confers on a State the faculty of altering unilaterally the use of an international watercourse whose legal régime has formed the subject-matter of an international treaty. In these circumstances, it is not the construction or the non-construction of works on the territory o one or the other Party per se nor solely the diversion of the course of the Danube which constitute the only breaches of the obligations under the 1977 Treaty. The fact of substituting and implementing a national project in place of a joint international project is a serious contravention of the provisions of the Treaty of Budapest. Limiting the sanction for unlawfulness to the factual consequences of the breach of international obligations but not to the breach itself represents "a precedent with disturbing implications for treaty relations and the integrity of the rule pacta sunt servanda" (see Judgment, para. 114). These considerations explain the validity of the proposition in Article 25, paragraph 1, of the International Law Commission Draft Articles on State Responsibility, on unlawfulness of a continuing character:

"The breach of an international obligation by an act of the State having a continuing character occurs at the moment when that act begins. Nevertheless, the time of commission of the breach extends over the entire period during which the act continues and remains not in conformity with the international obligation."

[pp. 195-196 D.O. Herczegh] Since Variant C, as such, constituted a breach of the 1977 Treaty, the unlawful conduct of Czechoslovakia began when it proceeded to the construction of those works necessary for the unilateral diversion of the Danube waters. It is completely arbitrary and inconsistent to separate that conduct of Czechoslovakia - unlawful in my opinion - from its result - unlawful according to the Court.

Accordingly, I conclude that Czechoslovakia acted unlawfully when, in November 1991, it embarked on the provisional solution. In other words, it was no more entitled to do so than to commission it in October 1992.

[p. 206 D.O. Fleischhauer] According to Article 18 of the Convention:

"A State is obliged to refrain from acts which would defeat the object and purpose of a treaty when:

(a) it has signed the treaty or has exchanged instruments constituting the treaty subject to ratification, acceptance or approval, until it shall have made its intention clear not to become a party to the treaty; or
(b) it has expressed its consent to be bound by the treaty, pending the entry into force of the treaty and provided that such entry into force is not unduly delayed."

I do not want to go into the question as to whether the whole of Article 18 corresponds actually to general international law. However, as the International Law Commission remarked in its Commentary on Article 15 (which became Article 18 in the text of the Convention as adopted) - with a reference to the Permanent Court's decision in the case concerning Certain German Interests in Polish Upper Silesia (Merits, Judgment No. 7, P.C.I.J, Series A, No. 7, p. 30):

"That an obligation of good faith to refrain from acts calculated to frustrate the object of the treaty attaches to a State which has signed a treaty subject to ratification appears to be generally accepted." (Yearbook of the International Law Commission, 1966, Vol. 11, p. 202.)

A fortiori does that obligation apply to a treaty after its entry into force. It follows from there that a State party to a treaty in force is not free to engage in - even on its own territory as Czechoslovakia did as from November 1991 - construction works which are designed to frustrate the treaty's very object, i.e., in the present case the creation and the operation of the Joint Project. The question of a justification of Czechoslovakia's construction work as countermeasure does not arise, as the Court has - rightly - found that the diversion of the Danube carried out by Czechoslovakia - which is the central part of Variant C - was not a lawful countermeasure because it was not proportionate (para. 87).

[p. 234 D.O. Skubiszewski] 7. The withdrawal of Hungary from the Project left Czechoslovakia wit the possibility of doing on its territory what it was allowed to do by general law. In the circumstances of the dispute submitted to the Court action based on general law does not derogate from the binding force of the Treaty. The shift onto the plane of general law results from the Hungarian rejection of the Project. There was, actually, no "single and indivisible operational system of works" (Art. 1, para. 1, of the 1977 Treaty) in which first Czechoslovakia and subsequently Slovakia could participate. The conduct of Hungary led to a factual situation which, as long as it lasted, prevented the implementation of binding agreements. A full application of the Treaty required bilateral action. Thus, for the time being, the treaty relationship of the two States found itself in a state of abeyance or inactivity. As the objectives of the Treaty did not disappear, a temporary solution would be based on general law and equity, until there was a return to the bilateral enforcement of the Treaty. That is the essence of the concept of the Czechoslovak "provisional solution", maintained by Slovakia.