Jump to Content Jump to Main Navigation
Max Planck Encyclopedia of Public International Law [MPEPIL]

Treaties, Amendment and Revision

Jan Klabbers

From: Oxford Public International Law (http://opil.ouplaw.com). (c) Oxford University Press, 2015. All Rights Reserved.date: 22 July 2019

Subject(s):
Treaties, amendments and modification — Vienna Convention on the Law of Treaties — Treaties, conclusion — Treaties, entry into force — Object & purpose (treaty interpretation and) — Sovereignty — State practice — Unilateral acts — Customary international law — Soft law — UNCLOS (UN Convention on the Law of the Sea)

Published under the auspices of the Max Planck Foundation for International Peace and the Rule of Law under the direction of Rüdiger Wolfrum.

A.  Introduction

It is sometimes said that it follows from the pacta sunt servanda norm, perhaps in conjunction with basic considerations of sovereignty (see also States, Sovereign Equality), that revision of treaties requires the unanimous consent of all parties. This undoubtedly holds true with respect to bilateral treaties: a proposal to amend may well be regarded as a new offer, requiring acceptance by the treaty partner (see also Treaties, Conclusion and Entry into Force). Much the same may generally also hold true with respect to multilateral treaties which are open to only a limited number of parties. But with large-scale multilateral treaties, the unanimity rule is generally considered unworkable: it would stimulate stasis and make it difficult for treaties to be adapted to changed political circumstances and power configurations (for a slightly different perception see the entry by Brunnée Consent). Consequently, the Vienna Convention on the Law of Treaties (1969) (‘VCLT’) has come to recognize, in addition to unanimous amendment, the possibility of a modification of multilateral treaties between some of the parties, subject to certain conditions.

B.  Amendment Procedures

The main challenge for treaty makers when it comes to regulating change is somehow to strike a balance between flexibility and stability. In some cases, agreement on this balance remains out of reach; as a result, treaties are concluded without any provision relating to their possible amendment, or with a very general provision merely postponing any decision. Examples include the Convention on Elimination of All Forms of Discrimination against Women (1979) (Women, Non-Discrimination of) and other ‘single issue’ conventions concluded under the auspices of the United Nations, merely providing that the UN General Assembly shall decide what to do if a party proposes revision (United Nations, General Assembly). In the absence of any provisions, the rules of the VCLT (Arts 39–41 VCLT) come into play. Those rules are not mandatory however: they are often referred to as being ‘residual’ in nature, leaving States free to decide what rules for revision suit their treaty regime best.

Typically, amendment clauses will have to address who has the right to propose amendments; how amendments shall be adopted; how they shall be accepted; and upon whom they will be binding. The right of initiative normally rests with all parties to a treaty; if a treaty establishes an international organization (International Organizations or Institutions, General Aspects), one or more of its organs may also be granted a right of initiative (International Organizations or Institutions, Decision-Making Bodies; International Organizations or Institutions, Decision-Making Process; see also International Organizations or Institutions, Voting Rules and Procedures). For example, within the European Union the Commission has such a right of initiative (European [Economic] Community; see also European Community and Union Law and International Law; European Union, Historical Evolution), and while Art. 108 United Nations Charter does not specifically grant anyone a right of initiative, it would seem generally accepted that such a right has been bestowed upon the UN General Assembly and any of the other principal organs of the UN, in addition to the individual Member States.

In order to adopt amendment proposals, many multilateral treaties envisage the convening of a revision conference to which all parties will be invited; perhaps the most well-known example is contained in Art. 109 UN Charter (United Nations Charter, Amendment). Still, it is often also possible to adopt them by means of a simplified procedure. Usually, this will be a procedure by correspondence, with the treaty’s depositary communicating the proposed amendment to the parties on the understanding that the absence of a response within a specified period of time will qualify as support for the adoption of the proposal (Acquiescence; Tacit Consent/Opting Out Procedure). Usually, then, the proposal will stand as adopted if it has been accepted by a majority of the parties. Often a two-thirds majority is specified.

A tricky question here, as well as with acceptance or approval of amendments, may be who exactly the eligible parties are. Obviously, the circle of parties will include those States for whom the treaty was in force at the moment the proposal was made, but will it also include States for whom the treaty was not yet in force but entered into force in the meantime? There would be sound policy reasons to include these as well since they will be expected to act in accordance with the treaty, after all. But this may make it more difficult to get the proposal adopted (see also TO Elias The Modern Law of Treaties [Oceana Publications Dobbs Ferry 1974] 92–4). A more hypothetical issue would relate to parties who have given notice of withdrawal (see also Declaration; Unilateral Acts of States in International Law), and thus will no longer be parties at the moment the proposal amendment might become effective (see also Treaties, Termination; Treaties, Validity). It would seem justifiable to minimize their involvement, since their participation in the treaty as amended is no longer expected.

Once adopted, amendments will normally still have to be approved by the individual parties. As a result, many treaties specify that adopted amendments shall enter into force upon acceptance or approval by the parties, often again by a two-thirds majority. Famously, the UN Charter can only be amended with the consent of the five permanent members of the UN Security Council (United Nations, Security Council; see also Veto), a factor which, many think, has impeded UN reform for too long (United Nations Charter, Reform). In some cases, extraneous approval is also required. Thus, amendments to various human rights treaties concluded under UN auspices require the approval of the General Assembly.

The legal effect of adopted and accepted amendments is, quite possibly, the most difficult issue to address. It would seem to follow from general considerations of sovereignty and the pacta sunt servanda rule that amendments can only bind those States that have actually accepted them. Art. 40 (4) VCLT confirms this, by specifying that an ‘amending agreement does not bind any State already a party to the treaty which does not become a party to the amending agreement’. It further refers to the rules on treaty conflict (Treaties, Conflicts between), to the effect that between parties to the amending agreement, the amending agreement shall apply, whereas in relation to those who have not accepted the amendment, the original treaty shall apply (see generally G Binder, Treaty Conflict and Political contradiction [Praeger London 1988]; J Pauwelyn, Conflict of Norms in Public International Law [CUP Cambridge 2003]).

Obviously, this is not always workable; the solution of Art. 40 VCLT presupposes that all treaty provisions can be ‘split’ into bundles of bilateral rights and obligations. This may of course—depending on the nature of the treaty—be the case with all or some of its substantive provisions: a multilateral extradition treaty may well be regarded as a grouping of bilateral, respectively ‘bipolar’, relations. However, this is less likely to be the case with treaties creating interdependent regimes (typically, such treaties as human rights conventions or treaties for the protection of the environment; Environment, International Protection; Objective Regime; Regime Theory; Self-Contained Regime), and will also be difficult to envisage with treaties creating international organizations or establishing institutions or monitoring bodies: it is generally thought that one cannot have different decision-making procedures for different parties. And even with non-institutional, substantive, provisions it may be undesirable to have different regimes for different parties.

As a result, several other devices have been developed. One is that some treaties (most famously the UN Charter in its Art. 108) simply provide that an amendment, once accepted by a specified majority of the parties, enters into force for all parties. In such a case, Member States unwilling to accept the amendment are faced with the stark choice of either living with it regardless of their objections, or withdrawing from the treaty or organization. In the specific case of the UN, the latter is not an easy option, as the UN Charter does not provide for a withdrawal clause (see eg Karl and Mützelburg 1341–63; Cot and Pellet 2191–218).

10  A different option, pioneered by some of the technical international organizations—the International Telecommunication Union (ITU) being a prime example—is to specify that the amending agreement, once it enters into force upon acceptance by a majority of the States or parties, replaces the earlier agreement in its totality. As a result, the minority unwilling to accept the amending agreement automatically cease to be parties or members, as the instrument to which they were committed is no longer deemed in force. This leads to what are sometimes referred to as ‘serial treaties’, and pressures States into accepting amending agreements.

C.  Modification Inter Se

11  But perhaps the most elegant way out of the above problem is, as recognized in Art. 41 VCLT, to allow a modification between some of the parties only. This method was already available in the late 19th century (as exemplified by the Convention establishing the Union for the Protection of Intellectual Property [1883]; Intellectual Property, International Protection) but, in the VCLT’s formulation, comes with strict conditions. In essence, such a modification may ‘not affect the enjoyment by the other parties of their rights under the treaty or the performance of their obligations’ and, cumulatively, may not relate to a treaty provision ‘derogation from which is incompatible with the effective execution of the object and purpose of the treaty as a whole’ (Treaties, Object and Purpose). While practical examples are hard to come by, some discussion on the topic was inspired by plans within the EU, in the late 1990s, to de-activate the Convention relating to the Status of Refugees (1951) (‘Refugee Convention’) as between Member States of the EU (Refugees). Many observers felt that this would amount to a modification of the Refugee Convention between some of its parties that would be difficult to reconcile with Art. 41 (1) (b) VCLT as it would affect the very definition of refugee; many held it would be incompatible with the effective execution of the object and purpose of the Refugee Convention.

12  Difficult to classify is the technique used in 1994 to revise the 1982 UN Convention on the Law of the Sea before the latter had even entered into force (Law of the Sea; Law of the Sea, History of). After it had become clear that the 1982 UN Convention on the Law of the Sea would not attract universal participation, an agreement to ‘implement’ its most controversial aspect was concluded between a number of interested States, resulting in effect in a revision of the relevant 1982 provisions (Implementation Agreements).

D.  Other Methods of Change

13  There are various additional mechanisms by which treaties may change. Apart from regular amendment or modification, as recognized in Arts 40 and 41 VCLT, one may also think of replacing the entire treaty by a new one on the same subject-matter and comprising the same parties. While recognized in Art. 59 VCLT, it is doubtful whether this can still meaningfully be called ‘amendment’.

14  Treaties may also undergo substantive change by less directly visible means. The rules on interpretation of treaties (Interpretation in International Law) allow for the possibility of taking subsequent agreements between the parties, or their subsequent practice under the treaty (State Practice), into account when interpreting the treaty. As a result, a treaty may change beyond recognition due to the way its provisions are given effect, or due to an authoritative interpretation by its parties, often in the form of a recommendation (General Comments/Recommendations), resolution, or declaration. The International Court of Justice (ICJ), moreover, acknowledged this possibility in its advisory opinion (Advisory Opinions) on the legal effects of a UN Security Council resolution (International Organizations or Institutions, Secondary Law) on Namibia ([1971] ICJ Rep 16; South West Africa/Namibia [Advisory Opinions and Judgments]), holding in effect that the meaning of the word ‘concurrent’ in Art. 27 (3) UN Charter had been changed by the practice of the members of the Security Council and the UN at large.

15  In similar fashion, treaties establishing international organizations may undergo change, in effect, with the help of the doctrine of implied powers (Doctrines; International Organizations or Institutions, Implied Powers). Under this doctrine, an organization can be recognized as having powers which the Member States did not intentionally confer upon it. While a restrictive reading of the implied powers doctrine would suggest that such powers were always present and merely awaiting discovery, a broader reading (based, arguably, on the ICJ’s advisory opinion on Reparation for Injuries Suffered in the Service of the United Nations [Advisory Opinion] ([1949] ICJ Rep 173; Reparations) would suggest that constituent documents may actually be revised with the help of the implied powers doctrine.

16  It is also generally accepted that treaties may change under the influence of customary international law. Individual treaty provisions may be superseded by later rules of customary international law; and entire treaties may become obsolete due to the formation of later customary rules. Although extremely rare, the possibility cannot—and should not—be discounted altogether; it could be argued, for example, that this has happened to the third Hague Convention relative to the Opening of Hostilities (1907) (Hague Peace Conferences [1899 and 1907]; Humanitarian Law, International). After all, there is little point in insisting that old and outdated treaty rules remain adhered to until formally abrogated; indeed, the possibility of treaty termination by means of desuetudo points in the same direction, as does the general rejection in international law of the ‘acte contraire’ doctrine (according to which an instrument can only be changed or terminated by a later instrument of the same status) (see eg Vierdag, ‘Oorlogsverklaring’ [ie ‘declaration of war’] Inaugural Speech at the University of Amsterdam, 3 December 1992).

17  Precisely in order to accommodate the need for flexibility, relatively recent trends facilitating the revision of treaties include making a distinction between the basic principles of a regime, and the more technical and detailed rules. The basic principles are then subject to strict amendment procedures; the technical elaborations, however, whether laid down in annexes, in later protocols (Framework Agreements), or even in the treaty itself, would be subjected to less onerous and cumbersome procedures. In particular in fields of great technical complexity where the bases for rules may be subject to rapid change resulting from new scientific data or insights such as the field of environmental protection or disarmament, as exemplified by the opting-out procedure for annexes envisaged in the Convention on the Prohibition of the Development, Production, Stockpiling and Use of Chemical Weapons and on their Destruction (Chemical Weapons and Warfare), this has proved to be a useful device.

18  The regime on depletion of the ozone layer has, successfully it seems, experimented with an even more flexible procedure, allowing for some aspects of the regime to be changed by means of so-called ‘adjustments’, in effect simply majority decisions taken by the parties and binding upon all of them (Ozone Layer, International Protection). Also not unheard of is making a distinction between amendments based on their individual characteristics, and divided into those that create new rights or obligations for the parties, and those which do not; accordingly, proposals for amendment not creative of new rights or obligations may be subjected to a lighter procedure than those that alter the legal position of the Member States. As the dividing line between the two will by necessity be fluid, however, much political debate may be anticipated as to whether a proposal would be creative of new rights or obligations.

E.  Assessment

19  Treaty revision is a curiously under-analysed phenomenon in international law. While it is generally accepted that there are some legal rules governing the process, it is equally generally accepted that the process is, indeed, all these legal rules govern. Substantively, revision is often deemed to be a matter for politics and diplomacy, and as a result, no doubt, such clauses as do exist are widely diverging, revealing few generalities that may come to serve as normative guidance.

20  It is often held that treaties are difficult to amend and therefore instruments other than treaties (perhaps even outside the realm of international law altogether) would offer more flexibility to States. That is, at least to some extent, deceptive: it is up to the States themselves how difficult they wish to make it to revise their treaties. The underlying problem, more often than not, is that fundamental agreement on the desirability of change may be lacking; and it is this lack of agreement which makes it difficult to revise treaties. And to the extent that formal procedures are cumbersome, they are so for a reason: they protect the parties to the original treaty, and thus also the democratic structures within those parties.

21  Either way though, the perception of rigidity is difficult to overcome, and it is perhaps no coincidence that States often seek refuge in informal revision, by adopting resolutions or recommendations (General Comments/Recommendations) which may formally lack binding force and so are often not subject to domestic parliamentary approval but which are nevertheless expected to influence the parties’ future behaviour and thus, for all practical purposes, can be considered as tantamount to treaty revision (see also Soft Law). This may well respond to the perceived need for flexibility, but often at the expense of legitimate, democratic decision-making (Legitimacy). Whether that is a price worth paying may depend on the situation at hand.

Select Bibliography