Resolution 2728 (2024) has legally binding effect. Eran Sthoeger makes the claim that the Security Council’s resolution 2728, which demanded an immediate ceasefire in Gaza, is “not legally binding”. This conclusion, he asseverates, follows from “a sound understanding of the practice of the Security Council under the Charter”.
The true position is that the Council’s demands, in operative paragraph 1 of the resolution, are a binding decision under Article 25 of the Charter. Paragraph 1 of the resolution is in these terms:
The true position is that the Council’s demands, in operative paragraph 1 of the resolution, are a binding decision under Article 25 of the Charter. Paragraph 1 of the resolution is in these terms:
“Demands an immediate ceasefire for the month of Ramadan respected by all parties leading to a lasting sustainable ceasefire, and also demands the immediate and unconditional release of all hostages, as well as ensuring humanitarian access to address their medical and other humanitarian needs, and further demands that the parties comply with their obligations under international law in relation to all persons they detain”.
There are three reasons why Sthoeger’s analysis, like his conclusion, is wrong. The first concerns aspects of his analysis of previous Council practice under the Charter; the second, the Council’s use of the operative word “demands”; and the third, his analysis of the debates preceding and circumstances surrounding the adoption of the resolution.
First, I turn by way of introduction to Council practice under the Charter. Most will agree that it is the decisions of the Council, not resolutions as such, that have binding effect. Most will also agree that the Council can “adopt decisions intended to be binding in resolutions not under Chapter VII”; “t is not the reference to a particular chapter that is the ultimate arbiter of whether a resolution contains binding provisions” (Security Council Report (2008) p. 9). Sthoeger concedes that Members of the Council have considered binding resolutions that use the operative verb “demands”, which is what resolution 2728 does. His view is, however, that “virtually all of these resolutions can be traced back to Chapter VII, even if it is not explicitly invoked”. If, in other words, they were binding, it was not so much because of the use of the term “demands”, but owing to the fact that they were adopted under Chapter VII. He then proceeds to analyse certain resolutions which use “demands”, but which in his view differ from resolution 2728 because they were made under Chapter VII.
Sthoeger struggles in that regard with the example of resolution 1695 (2006). In that resolution, the Council, inter alia, demanded “that the DPRK suspend all activities related to its ballistic missile programme”. A reason Sthoeger claims that this resolution contained relevant indicators of binding effect not found in resolution 2728 is that “resolution 1695 reaffirmed that the proliferation of weapons of mass destruction (WMDs) ‘constitutes a threat to international peace and security’ in the language of Article 39”, which was not found in resolution 2728. Very few, however, will agree with his assessment: not even, it seems, the author himself when, two years ago, he was analysing the same practice (more coolly?) in a co-authored volume. In 2022 Wood and Sthoeger observed the following in connection with resolution 1695:
First, I turn by way of introduction to Council practice under the Charter. Most will agree that it is the decisions of the Council, not resolutions as such, that have binding effect. Most will also agree that the Council can “adopt decisions intended to be binding in resolutions not under Chapter VII”; “t is not the reference to a particular chapter that is the ultimate arbiter of whether a resolution contains binding provisions” (Security Council Report (2008) p. 9). Sthoeger concedes that Members of the Council have considered binding resolutions that use the operative verb “demands”, which is what resolution 2728 does. His view is, however, that “virtually all of these resolutions can be traced back to Chapter VII, even if it is not explicitly invoked”. If, in other words, they were binding, it was not so much because of the use of the term “demands”, but owing to the fact that they were adopted under Chapter VII. He then proceeds to analyse certain resolutions which use “demands”, but which in his view differ from resolution 2728 because they were made under Chapter VII.
Sthoeger struggles in that regard with the example of resolution 1695 (2006). In that resolution, the Council, inter alia, demanded “that the DPRK suspend all activities related to its ballistic missile programme”. A reason Sthoeger claims that this resolution contained relevant indicators of binding effect not found in resolution 2728 is that “resolution 1695 reaffirmed that the proliferation of weapons of mass destruction (WMDs) ‘constitutes a threat to international peace and security’ in the language of Article 39”, which was not found in resolution 2728. Very few, however, will agree with his assessment: not even, it seems, the author himself when, two years ago, he was analysing the same practice (more coolly?) in a co-authored volume. In 2022 Wood and Sthoeger observed the following in connection with resolution 1695:
“While it is understood that this resolution contains legally binding elements, it does not refer to Chapter VII or identify a threat to the peace; rather, it states that the Council is acting ‘under its special responsibility for the maintenance of international peace and security’, a reference to Article 24 of the Charter. But, regardless of how it relates to Chapter VII, it is considered binding” (p. 42).
Other authorities have reached the same conclusion, such as Marko Milanovic, who referred to resolution 1695 as not having made reference to “Chapter VII, nor made an Article 39 determination, but rather invoked the language of Article 24”. A study by Security Council Report similarly observed that 1695 was not, even by inference, a resolution under Chapter VII, but instead “uses the language in Article 24” (p. 10). The same study concluded that the intention, in resolution 1695, to adopt a resolution some provisions of which had binding effect should instead be inferred on a different basis: “the intention to adopt binding decisions seems to be clearly indicated in the use of operative paragraphs that ‘demand’ and ‘require’ certain outcomes” (ibidem).
That takes us to the second point: the use by the Council in resolution 2728 of the operative word “demands”. The International Court of Justice advised in Namibia that “[t]he language of a resolution of the Security Council should be carefully analysed before a conclusion can be made as to its binding effect” (p. 53, para. 114). It is generally agreed that “[a] key element in determining the binding nature of a provision is the operative word” (Wood & Sthoeger, p. 38). Thus Sievers & Daws observe that a first indication as to whether or not the Security Council intended its resolution to be mandatory can be “the operative word(s) used”; “[t]he terms ‘decides’, ‘requires’, ‘demands’, ‘requests’, and ‘authorizes’ are fairly definite and do not lead to many difficulties of interpretation” (p. 382). Indeed, the verb “demand” is defined by the Oxford English Dictionary as “to ask for (a thing) peremptorily”; the French equivalent in the practice of the Council, “exiger”, is defined by Dictionnaire Littré as “[r]éclamer quelque chose en vertu d’un droit”. The term is definite and should not lead to a difficulty of interpretation: it has a mandatory character.
As regards the effect of the use of the term “demands” by the Council, Wood & Sthoeger are of the view that “[t]he texts adopted by the Council indicate that a demand may be binding” (p. 40). In the same vein, the chapter on Article 39 in Cot, Pellet, and Forteau’s La Charte des Nations Unies: Commentaire article par article explains that the gamut of operative words, from the most likely to express binding effect to the least, begins with “decides” (“décide”), “demands” (“exige”), and then “orders” (“ordonne”) at the binding end, shading into “request” (“demande” or “prie”), “encourages” (“encourage”), “calls upon” (“appelle” or “exhorte”) towards the recommendatory end (d’Argent, p. 1167). As can be seen, “demands” is among the most mandatory operative words the Council uses. Wood & Sthoeger accordingly noted in their book that, when the Council, in resolution 662 (1990), used the wording “Demands that Iraq rescind its actions purporting to annex Kuwait”, that was an example of the “understanding that demands can be binding” (p. 40). And Sievers & Daws noted that resolution 2087 (2013), which used the wording “Demands that the DPRK not proceed with any further launches using ballistic missile technology”, and had no reference to Chapter VII or to a threat to international peace and security, seemed to be intended to be mandatory (p. 389).
But, and this takes us to the third point, if careful analysis of the language of Council resolutions is necessary, so too is analysis of “statements by representatives of members of the Security Council made at the time of their adoption” and also the subsequent practice of “States affected” by the resolutions (Kosovo, p. 442, para. 94). Such recourse is, as the Court observed in Namibia, part of the category “all circumstances that might assist in determining the legal consequences of the resolution of the Security Council” (p. 53, para. 114). For example, as Greenwood has observed, “t is impossible properly to understand the text of any Security Council resolution without reference to the debates which preceded it” (p. 73).
The circumstances of the adoption of resolution 2728 were that it was a so-called E-10 draft: the draft was produced not by one or more of the five permanent Members of the Council, but by its 10 elected Members, in consultation with the Arab Group. It was introduced on behalf of the E-10 by the group’s coordinator, Mozambique, which together with Slovenia had led the negotiations.
After the adoption, Mozambique expressed its view that the resolution was “binding and mandatory” (media stakeout by the E-10 on 25 March 2024). Slovenia similarly took the view that it was “binding for each and every State Member of the United Nations” (S/PV.9596, p. 8; see also S/PV.9586, p. 5). In total, a majority — 9 out of 15 — of the Members of the Council have expressed the view that resolution 2728 has binding effect. (Nine could incidentally be thought to be a salient number: it is the number of affirmative votes needed for the successful adoption of a draft resolution (on a non-procedural matter): Article 27(3) of the Charter.) Algeria said in connection with resolution 2728 that Member States were “bound, under Article 25 of the Charter of the United Nations, to carry out the decisions of the Security Council” (S/PV.9596, p. 6; see also S/PV.9588, p. 17). China expressed the view that it was “binding” (S/PV.9596, p. 12; see also S/PV.9588, p. 19). France expressed its view that resolution 2728 “must be implemented by everyone, as set out in Article 25 of the Charter” (S/PV.9588, p. 10), and that it was “contraignante”, i.e. binding (Point de presse live). Malta was of the view that the resolution was “binding and must be respected” (S/PV.9593, p. 11). The Russian Federation observed that the resolution was “a legally binding decision of the Council” (S/PV.9593, p. 3; see also S/PV.9588, p. 15). Sierra Leone observed that it was “binding” (S/PV.9596, p. 16; S/PV.9588, p. 14). Switzerland set out its position that the resolution was “binding” (9617th meeting, 02:36:24).
The United Kingdom, possibly on the basis that discretion is the better part of valour, limited itself to calling for “this resolution to be implemented immediately” (S/PV.9586, p. 11). Two Members of the Council committed themselves to the view that the resolution does not have binding effect: the United States expressed its view that the resolution was “non-binding” (S/PV.9586, p. 5); South Korea questioned the binding nature of the resolution in legal terms (media stakeout by the E-10 on 25 March 2024).
The Arab Group — in the words of Kosovo, “States affected” by the resolution — was like the majority of the Council membership of the view that resolution 2728 was “binding” (S/PV.9586, p. 15); the European Union, which may also be thought to represent States so “affected”, referred on 23 April 2024 to “the UN Security Council’s binding resolution 2728”.
In conclusion, resolution 2728 (2024) contains, in its paragraph 1, a decision in the sense of Article 25 of the Charter. It is binding. That is evident from the mandatory wording of the resolution and from the fact, expressed in the debates preceding the adoption and the circumstances surrounding it, that the collective body of the Council itself considers it so to be.
That takes us to the second point: the use by the Council in resolution 2728 of the operative word “demands”. The International Court of Justice advised in Namibia that “[t]he language of a resolution of the Security Council should be carefully analysed before a conclusion can be made as to its binding effect” (p. 53, para. 114). It is generally agreed that “[a] key element in determining the binding nature of a provision is the operative word” (Wood & Sthoeger, p. 38). Thus Sievers & Daws observe that a first indication as to whether or not the Security Council intended its resolution to be mandatory can be “the operative word(s) used”; “[t]he terms ‘decides’, ‘requires’, ‘demands’, ‘requests’, and ‘authorizes’ are fairly definite and do not lead to many difficulties of interpretation” (p. 382). Indeed, the verb “demand” is defined by the Oxford English Dictionary as “to ask for (a thing) peremptorily”; the French equivalent in the practice of the Council, “exiger”, is defined by Dictionnaire Littré as “[r]éclamer quelque chose en vertu d’un droit”. The term is definite and should not lead to a difficulty of interpretation: it has a mandatory character.
As regards the effect of the use of the term “demands” by the Council, Wood & Sthoeger are of the view that “[t]he texts adopted by the Council indicate that a demand may be binding” (p. 40). In the same vein, the chapter on Article 39 in Cot, Pellet, and Forteau’s La Charte des Nations Unies: Commentaire article par article explains that the gamut of operative words, from the most likely to express binding effect to the least, begins with “decides” (“décide”), “demands” (“exige”), and then “orders” (“ordonne”) at the binding end, shading into “request” (“demande” or “prie”), “encourages” (“encourage”), “calls upon” (“appelle” or “exhorte”) towards the recommendatory end (d’Argent, p. 1167). As can be seen, “demands” is among the most mandatory operative words the Council uses. Wood & Sthoeger accordingly noted in their book that, when the Council, in resolution 662 (1990), used the wording “Demands that Iraq rescind its actions purporting to annex Kuwait”, that was an example of the “understanding that demands can be binding” (p. 40). And Sievers & Daws noted that resolution 2087 (2013), which used the wording “Demands that the DPRK not proceed with any further launches using ballistic missile technology”, and had no reference to Chapter VII or to a threat to international peace and security, seemed to be intended to be mandatory (p. 389).
But, and this takes us to the third point, if careful analysis of the language of Council resolutions is necessary, so too is analysis of “statements by representatives of members of the Security Council made at the time of their adoption” and also the subsequent practice of “States affected” by the resolutions (Kosovo, p. 442, para. 94). Such recourse is, as the Court observed in Namibia, part of the category “all circumstances that might assist in determining the legal consequences of the resolution of the Security Council” (p. 53, para. 114). For example, as Greenwood has observed, “t is impossible properly to understand the text of any Security Council resolution without reference to the debates which preceded it” (p. 73).
The circumstances of the adoption of resolution 2728 were that it was a so-called E-10 draft: the draft was produced not by one or more of the five permanent Members of the Council, but by its 10 elected Members, in consultation with the Arab Group. It was introduced on behalf of the E-10 by the group’s coordinator, Mozambique, which together with Slovenia had led the negotiations.
After the adoption, Mozambique expressed its view that the resolution was “binding and mandatory” (media stakeout by the E-10 on 25 March 2024). Slovenia similarly took the view that it was “binding for each and every State Member of the United Nations” (S/PV.9596, p. 8; see also S/PV.9586, p. 5). In total, a majority — 9 out of 15 — of the Members of the Council have expressed the view that resolution 2728 has binding effect. (Nine could incidentally be thought to be a salient number: it is the number of affirmative votes needed for the successful adoption of a draft resolution (on a non-procedural matter): Article 27(3) of the Charter.) Algeria said in connection with resolution 2728 that Member States were “bound, under Article 25 of the Charter of the United Nations, to carry out the decisions of the Security Council” (S/PV.9596, p. 6; see also S/PV.9588, p. 17). China expressed the view that it was “binding” (S/PV.9596, p. 12; see also S/PV.9588, p. 19). France expressed its view that resolution 2728 “must be implemented by everyone, as set out in Article 25 of the Charter” (S/PV.9588, p. 10), and that it was “contraignante”, i.e. binding (Point de presse live). Malta was of the view that the resolution was “binding and must be respected” (S/PV.9593, p. 11). The Russian Federation observed that the resolution was “a legally binding decision of the Council” (S/PV.9593, p. 3; see also S/PV.9588, p. 15). Sierra Leone observed that it was “binding” (S/PV.9596, p. 16; S/PV.9588, p. 14). Switzerland set out its position that the resolution was “binding” (9617th meeting, 02:36:24).
The United Kingdom, possibly on the basis that discretion is the better part of valour, limited itself to calling for “this resolution to be implemented immediately” (S/PV.9586, p. 11). Two Members of the Council committed themselves to the view that the resolution does not have binding effect: the United States expressed its view that the resolution was “non-binding” (S/PV.9586, p. 5); South Korea questioned the binding nature of the resolution in legal terms (media stakeout by the E-10 on 25 March 2024).
The Arab Group — in the words of Kosovo, “States affected” by the resolution — was like the majority of the Council membership of the view that resolution 2728 was “binding” (S/PV.9586, p. 15); the European Union, which may also be thought to represent States so “affected”, referred on 23 April 2024 to “the UN Security Council’s binding resolution 2728”.
In conclusion, resolution 2728 (2024) contains, in its paragraph 1, a decision in the sense of Article 25 of the Charter. It is binding. That is evident from the mandatory wording of the resolution and from the fact, expressed in the debates preceding the adoption and the circumstances surrounding it, that the collective body of the Council itself considers it so to be.