UN Security Council Structure Powers and the Veto

Of the six principal organs of the United Nations, only one can issue an order that every member state is legally bound to obey. The Security Council is where the Charter concentrates its coercive authority, and it is also where the Charter concedes most openly that law follows power. Understanding the Council means holding two facts together: it has the widest formal powers of any body in the history of international organisation, and it was deliberately built so that those powers can never be turned against the five states that wrote them.

The Bargain of 1944–45: Why the Council Has This Shape

The Council’s structure is not an accident of drafting or a compromise that time has spoiled. It is the price the wartime allies charged for joining, and it was paid consciously by everyone else.

The great powers’ terms

• The design came out of Dumbarton Oaks (August–October 1944), where the United States, Britain, the Soviet Union and China settled the outline of the new organisation before the smaller states were consulted at all.
• The unresolved question was voting in the Council, and it was settled by Roosevelt, Churchill and Stalin at Yalta in February 1945 — the arrangement known ever since as the Yalta formula.
  • Stalin’s opening position was absolute unanimity: no decision of any kind without the assent of every great power.
  • The Anglo-American counter-proposal was that a permanent member party to a dispute should abstain, which Stalin rejected outright.
  • The compromise limited obligatory abstention to pacific settlement — to Chapter VI matters, never to enforcement. That line survives verbatim in Article 27(3) today.
• The justification offered was that the organisation would enforce nothing without the great powers, so a decision they had not concurred in would be one nobody carried out.
• The League of Nations supplied the negative lesson — its collapse is treated in its own right elsewhere — and the drafters concluded it had failed partly by promising enforcement it could not deliver.

“The permanent members could not be expected, in the present condition of the world, to assume the obligation to act in so serious a matter as the maintenance of international peace and security in consequence of a decision in which they had not concurred.” — Statement of the Four Sponsoring Governments, San Francisco, 7 June 1945

The revolt at San Francisco and the ultimatum that ended it

• The smaller and middle powers fought the veto harder than any other provision of the Charter. They lost, and the manner of the loss is instructive.
• A questionnaire of twenty-three questions was put to the sponsoring powers on 22 May 1945, of which the decisive one asked what counted as a procedural matter and who would decide.
  • The answer, in the Statement of the Four Sponsoring Governments of 7 June 1945, was that deciding whether a question is procedural would itself be a substantive question — and so vetoable.
  • That single sentence created the double veto, and it did so through an interpretive statement rather than through the Charter’s text.
• Australia’s H. V. Evatt led the opposition, drawing the distinction that the small states could live with a veto on enforcement but not with a veto on debate.
• The United States closed the argument with an explicit threat. Senator Tom Connally told the delegates they could go home claiming to have defeated the veto — and equally that they had torn up the Charter — while shredding his own draft copy at the table.

“You may go home from San Francisco… and report that you have defeated the veto… but you can also say, ‘We tore up the Charter!'” — Tom Connally

• The veto was approved on 20 June 1945 by thirty votes to two, with fifteen abstentions — a margin recording resignation rather than assent.
• The choice was never veto or no veto; it was this organisation or none. Truman’s own later judgement was that no arrangement without the veto would have passed the United States Senate.

The Council was not built to be fair. It was built to be joined, and in 1945 those two objectives were understood to be incompatible.

Article 24: primary responsibility, conferred and delegated

• Article 24(1) confers on the Council “primary responsibility for the maintenance of international peace and security” — and does so in a specific way that repays attention.
• The responsibility is conferred by the membership, which agrees in the same article that in discharging it the Council “acts on their behalf”.
  • This answers the sovereign-equality objection — the fifteen act as agents of the 193 — and is equally the source of every demand for accountability to the wider membership, since an agent owes its principal an explanation.
• The word is “primary”, not “exclusive”. The Assembly keeps a real if recommendatory competence, constrained chiefly by Article 12(1), which bars it from recommending on a dispute while the Council is exercising its functions on it.
• Article 24(2) binds the Council to act in accordance with the Purposes and Principles of the United Nations — the textual hook for every argument that the Council’s discretion has legal limits at all.
• Article 24(3) requires annual reports to the Assembly — in practice a compilation rather than an account, and a standing grievance of the membership.

Article 25: the only organ whose decisions bind

• Article 25 is the operative clause of the whole system: members “agree to accept and carry out the decisions of the Security Council” in accordance with the Charter.
  • No other principal organ can do this. Assembly resolutions recommend; Council decisions oblige.
  • Article 103 completes the effect, giving Charter obligations priority over any other international agreement a member has entered.
• The long-running argument was whether Article 25 covers only Chapter VII decisions or extends to all decisions properly taken.
  • The narrow reading treats binding force as an incident of enforcement, reducing everything outside Chapter VII to advice.
  • The International Court of Justice rejected that reading in the Namibia advisory opinion (1971), holding Article 25 not confined to enforcement action but applicable to decisions of the Council adopted in accordance with the Charter.
• The practical test is therefore language, not chapter heading. Whether a resolution binds turns on whether the Council decided or merely recommended — which is why the verb in an operative paragraph is fought over so hard.
• A resolution can accordingly create obligations for states that never voted on it, including the 178 members outside the chamber — an effect with no parallel elsewhere in a consent-based legal order.

A body that never adjourns

• Article 28 requires the Council to be “so organized as to be able to function continuously”, with every member represented at the seat of the Organization at all times.
• This is a structural advantage over every other organ. The Assembly’s plenary work is seasonal; the Council can convene within hours, and has.
  • Members maintain resident permanent representatives whose entire brief is the Council file, and continuity of personnel produces continuity of practice — which is how so much of the Council’s law came to rest on precedent rather than text.
• Permanence also entrenches the permanent members’ advantage: a mission with decades of memory and its own intelligence feed does not compete on level terms with an elected member’s four-person team.

Composition: Five Fixed Seats and Ten Contested Ones

Article 23 does two very different things in one provision: it names five states in the text of a treaty, and it creates a competitive election for ten more. The tension between those mechanisms is most of what is politically alive about the Council today.

The five, and the two silent successions

• Article 23(1) names the permanent members: the Republic of China, France, the Union of Soviet Socialist Republics, the United Kingdom and the United States.
  • Two of those names no longer correspond to existing states, and both were replaced without a single word of the Charter being amended.
• China’s seat passed to the People’s Republic by General Assembly resolution 2758 (XXVI) of 25 October 1971, which “restored the lawful rights” of the People’s Republic and expelled the representatives of Chiang Kai-shek.
  • The Assembly, not the Council, decided, treating the matter as one of credentials and representation rather than of membership or amendment.
  • No amendment to Article 23 followed, and the Charter still names “the Republic of China” today.
• Russia took the Soviet seat in December 1991 with no vote of any organ at all. On 24 December 1991 Boris Yeltsin wrote to the Secretary-General that Soviet membership was “being continued by the Russian Federation”.
  • The letter was circulated, no member objected, the name-plate and flag were changed, and the Russian delegation occupied the chair.
  • The legal theory was continuation, not succession — awkward, since the Soviet Union had declared its own extinction three days earlier, but the alternative was a vacant permanent seat the Charter does not contemplate.
• What both precedents establish is not that the Charter is flexible but that it is flexible when the permanent members agree. The rigidity that has blocked expansion for sixty years dissolved twice, within days, when the five wanted it to.

Change to the Council	Instrument used	Charter amended?
Enlargement from eleven to fifteen members	Charter amendment to Article 23, adopted 17 December 1963	Yes — the only substantive amendment on the Council ever made
China’s seat to the People’s Republic	Assembly resolution 2758 (1971), a credentials decision	No
Russia in the Soviet seat	A letter of 24 December 1991, unopposed	No
Abstention treated as not blocking	Practice, confirmed in Namibia (1971)	No
Peacekeeping	Practice, from 1956 onward	No

The elected ten and the only expansion in eighty years

• The Council began with eleven members — five permanent and six elected — a size chosen at San Francisco to keep it small enough for “prompt and effective action”.
• Decolonisation destroyed the arithmetic. Membership rose from 51 to over a hundred in twenty years, and the new African and Asian states had almost no prospect of a seat.
• The 1963 amendment raised the elected members from six to ten and the Council from eleven to fifteen. It was adopted on 17 December 1963 and entered into force on 31 August 1965 — two distinct events, routinely conflated.
  • It also raised the majority required for a decision from seven votes to nine.
  • It remains the only substantive Charter amendment affecting the Council in eighty years, and it touched the elected category alone.
• The permanent membership has therefore never once been altered by amendment, which is the central fact of the Council’s constitutional history.

The regional formula

• General Assembly resolution 1991 A (XVIII), adopted alongside the 1963 amendment, fixed how the ten elected seats are distributed — a political agreement, not a Charter rule, and one that has held without deviation.

Regional group	Elected seats	Notes
African Group	3	The largest allocation; the A3 now coordinate positions as a bloc
Asia-Pacific Group	2	Serves the largest population of any group
Latin America and the Caribbean	2	The GRULAC seat has produced several of the longest deadlocks
Western European and Others	2	Includes Australia, Canada, New Zealand and, for electoral purposes, the United States
Eastern European Group	1	The smallest allocation, and the one most disrupted by the war in Ukraine

• A convention layered on the formula reserves one seat for an Arab state, rotating between the African and Asian allocations. Bahrain’s 2026–27 term occupies that rotation.
• Seats are renewed five each year, so the Council always holds a second-year cohort — the elected side’s only institutional memory.

How members are elected, and how elections go wrong

• Election is by the General Assembly and requires a two-thirds majority of members present and voting, under Article 18(2), which classes Council elections as an “important question”.
• Terms are two years, and Article 23(2) forbids immediate re-election: a state must sit out a full term before returning.
• Article 23(1) sets two criteria for choosing members — contribution to the maintenance of international peace and security, and equitable geographical distribution.
  • The first is honoured almost entirely in the breach. Nothing in practice tests a candidate’s contribution; the vote turns on reciprocal support, aid relationships and regional turn-taking.
  • The second has hardened into the 1963 formula and is the operative rule.
• The dominant practice is the clean slate — the group endorses exactly as many candidates as there are seats and the Assembly ratifies, though formal balloting is still required.
• Contested races are brutal, because the ballot rules make no provision for exhaustion.
  • The 1979 contest between Colombia and Cuba ran to 154 rounds before both withdrew and Mexico was elected on the 155th — the longest election in the Organization’s history.
  • In 2006 Guatemala and Venezuela deadlocked over 47 rounds before Panama came through as a compromise on the 48th.
  • In 2016 Italy and the Netherlands tied at 95 votes each and split the two-year term, one year apiece — the first split term since the 1960s, a device last used after 35 inconclusive rounds in 1955.
• Deadlock is resolved by attrition, never by rule, which is one measure of how genuinely competitive the elected category is and the permanent one is not.

The Council in 2026

• The permanent members are China, France, Russia, the United Kingdom and the United States.
• The elected members serving in 2026 are Denmark, Greece, Pakistan, Panama and Somalia (2025–26) and Bahrain, Colombia, the Democratic Republic of the Congo, Latvia and Liberia (2026–27).
• On 3 June 2026 the Assembly elected five members for 2027–28: Zimbabwe (182 votes), Trinidad and Tobago (181), Kyrgyzstan (141), Portugal (134) and Austria (131).
  • Kyrgyzstan takes a Council seat for the first time, having beaten the Philippines in a contested Asia-Pacific round — evidence that the clean-slate habit is not universal.
• The Council’s machinery has been visibly strained: allocation of chairs of subsidiary bodies deadlocked over the Iran sanctions “snapback”, producing the longest gap without allocated chairs since 1979.

The presidency and the participation of outsiders

• The presidency rotates monthly among all fifteen in the English alphabetical order of their names, so every member — permanent or elected — presides on identical terms.
• What the presidency controls is real but bounded. It sets the monthly programme of work, convenes and chairs meetings, represents the Council externally, and can stage signature events on themes of its choosing.
  • What it cannot control is outcomes. It cannot compel a vote, override a veto, or force a draft onto the table against the penholder’s wishes.
  • A presidency is therefore an agenda-setting month, not a month of power, which is why elected members invest so heavily in the events they can actually deliver.
• Article 31 lets any member outside the Council participate without a vote whenever the Council considers that member’s interests to be specially affected — with the Council deciding whether they are.
• Article 32 goes further for parties to a dispute: any member, and any non-member state, party to a dispute before the Council shall be invited to participate without a vote.
• Both are thinly used relative to their potential. Participation is granted freely for open debates and rationed carefully near decisions, and neither article opens the closed consultations where drafts are actually settled.

The Powers of the Council

The Charter gives the Council a graduated set of tools running from encouragement to armed force, and a discretion over which rung to stand on that is close to unreviewable. Chapters VI, VII and VIII carry the substance; Articles 94, 97 and 108 add powers over the rest of the system.

Pacific settlement under Chapter VI

• Chapter VI covers disputes “likely to endanger” the maintenance of international peace and security, and everything in it is recommendatory.
• Article 33 places the first obligation on the parties, who shall seek a solution by negotiation, enquiry, mediation, conciliation, arbitration, judicial settlement, resort to regional agencies or other peaceful means of their choice.
• Article 34 gives the Council its investigative power: it may investigate any dispute or situation which might lead to friction, in order to decide whether continuance is likely to endanger peace.
  • This is the gateway to seisin — investigation is how a matter formally becomes the Council’s business.
  • It is also rarely used in its own name; the Council prefers fact-finding by the Secretary-General, commissions of inquiry or panels of experts.
• Article 35 governs referral: any member may bring a dispute or situation to the Council, and a non-member state may bring a dispute to which it is a party, if it accepts the Charter’s pacific-settlement obligations for it.
• Articles 36 to 38 cover what the Council can then do.
  • Article 36 — recommend appropriate procedures or methods of adjustment, with a direction that legal disputes should as a general rule go to the International Court of Justice.
  • Article 37 — where the parties fail to settle, they shall refer it to the Council, which may recommend terms of settlement, not merely procedures.
  • Article 38 — with the consent of all parties, the Council may make recommendations on any dispute at all.
• The whole chapter therefore produces recommendations only. Nothing under Chapter VI binds anyone, which is why the Council’s centre of gravity moved to Chapter VII and stayed there.

Enforcement under Chapter VII

• Article 39 is the gateway to every coercive power the Council has. The Council “shall determine the existence of any threat to the peace, breach of the peace, or act of aggression” and then decide what measures to take.
  • The determination is the Council’s alone, and the Charter supplies no definition of any of the three terms.
  • Practice has made “threat to the peace” the workhorse; findings of a breach of the peace are rare and of an act of aggression rarer still.
  • Because no organ reviews an Article 39 finding, the Council’s jurisdiction over itself is self-conferred — the most consequential design choice in the Charter after the veto.
• Article 40 allows provisional measures — ceasefires, withdrawals, standstills — without prejudice to the parties’ rights, pending a decision on substance.
• Article 41 authorises measures not involving armed force: interruption of economic relations and of rail, sea, air, postal, telegraphic and radio communications, and severance of diplomatic relations.
  • The list is illustrative, not exhaustive, and the Council has read it very widely — widely enough to create criminal tribunals under it.
• Article 42 authorises action by air, sea or land forces where the Council considers Article 41 measures inadequate.
• Article 43 was meant to supply the forces, through special agreements under which members would place national contingents at the Council’s disposal.
  • Not one such agreement has ever been concluded. Negotiations broke down in 1947 and were never revived.
  • The Military Staff Committee of Article 47, composed of the permanent members’ chiefs of staff, therefore has nothing to command, and has met for eight decades without a function.
  • The Council therefore does not deploy force; it authorises willing states and coalitions to use it, which is a different constitutional animal from the one the Charter drew.
• Article 48 lets the Council decide that action be taken by all or some members, which is what makes selective implementation lawful; Article 49 obliges members to afford mutual assistance in carrying measures out, making sanctions-busting by a third state a Charter violation rather than an unfriendly act.
• The record of Chapter VII enforcement — Korea, the Gulf War, the cases where the Council did nothing — belongs to collective security as a subject in itself. What matters here is the legal architecture: the Council owns the determination, the measures and the authorisation, and it owns no forces.

Chapter	Trigger	What the Council may do	Legal effect
Chapter VI	A dispute likely to endanger peace	Call on parties to settle; investigate; recommend procedures or terms	Recommendatory
Chapter VII	An Article 39 determination	Provisional measures; sanctions; authorisation of force	Binding where the Council decides
Chapter VIII	A local dispute or regional enforcement	Encourage regional settlement; authorise regional enforcement	Regional action needs Council authorisation

Article 51 and the boundaries of self-defence

• Article 51 preserves “the inherent right of individual or collective self-defence if an armed attack occurs” against a member — the only unilateral exception to the Charter’s prohibition on force in Article 2(4).
• The right lasts only “until the Security Council has taken measures” necessary to maintain peace and security, which subordinates it in principle to collective decision.
• Measures taken in self-defence must be immediately reported to the Council, and the reporting requirement is the one procedural check the article contains.
  • Reports are filed; they are almost never adjudicated, and no organ rules on whether a claim was good.
• Three arguments have run continuously since 1945 and none is settled.
  • Anticipatory self-defence — whether an attack must have occurred, or whether an imminent one suffices, as the Caroline formula of necessity and proportionality suggested long before the Charter.
  • Pre-emptive or preventive force against a threat neither imminent nor certain — asserted in the United States’ 2002 security doctrine, rejected by most states and by the High-level Panel of 2004.
  • Self-defence against non-state actors on the territory of a state “unwilling or unable” to suppress them — relied on across Syria and the Sahel, and firmly rejected by many states.
• The stakes are institutional rather than academic. Every widening of Article 51 is a narrowing of the Council’s monopoly, because force justified as self-defence needs no authorisation at all.

Chapter VIII and the regional turn

• Article 52 encourages regional arrangements to settle local disputes before they reach the Council.
• Article 53 is the hard rule: no enforcement action shall be taken under regional arrangements without the authorisation of the Security Council.
  • It still carries the obsolete “enemy state” exception covering the Second World War’s Axis powers, which the Assembly resolved to delete in 1995 and never has, because deletion needs an amendment nobody will open.
• Practice has run well ahead of the text, and the Council has generally ratified it.
  • Retrospective authorisation and deliberately ambiguous mandates have blessed action already under way; Kosovo in 1999 was never authorised at all.
  • Regional and sub-regional bodies now do much of the work the Charter imagined for the Council — the African Union, ECOWAS, the European Union, NATO and the Collective Security Treaty Organization.
• Security Council resolution 2719 of 21 December 2023 is the most significant institutional change in this area in a decade.
  • It opens United Nations assessed contributions to African Union-led peace support operations authorised by the Council, funding up to 75 per cent of an operation’s annual budget, with the rest raised jointly from other sources.
  • Requests are decided case by case, and only operations under the African Union’s direct and effective command and control qualify.
  • It matters because it offers an alternative to blue-helmet peacekeeping at exactly the moment when peacekeeping is contracting — and because implementation has stalled on burden-sharing ever since.

The Council’s authority has migrated outward: it increasingly licenses others to act rather than acting itself, and each licence is negotiated rather than commanded.

Sanctions: the instrument the Council actually uses

Sanctions are where Article 41 has done nearly all its work. They are the Council’s default coercive response, imposed far more often than force is authorised, and their history is a case study in an institution learning from its own damage.

From comprehensive embargo to targeted measure

• Comprehensive sanctions were imposed only twice in the Council’s first forty-five years — on Southern Rhodesia from 1966 and on South Africa’s arms trade from 1977.
• Resolution 661 of 6 August 1990 imposed a near-total trade and financial embargo on Iraq, four days after the invasion of Kuwait, and it stayed in place until 2003.
  • The humanitarian consequences discredited the instrument. Child mortality rose sharply, public health and water infrastructure decayed, and estimates of excess deaths ran to hundreds of thousands.
  • The oil-for-food programme, created by resolution 986 of 14 April 1995, let Iraq sell oil for humanitarian goods under UN supervision. It relieved some of the damage and became a byword for corruption.
  • Denis Halliday, the UN humanitarian coordinator in Iraq, resigned in 1998 after thirty-four years of service, saying publicly that the sanctions regime was producing genocidal conditions.
  • His successor Hans von Sponeck resigned in February 2000 on the same grounds, as did the head of the World Food Programme in Iraq days later — three resignations over one sanctions regime, without parallel.
• The lesson drawn was that a whole-economy embargo punishes a population and insulates a leadership, since a regime controlling scarce goods often grows stronger under blockade.
• The response was the shift to targeted or “smart” sanctions developed through the Interlaken, Bonn–Berlin and Stockholm processes of the late 1990s and early 2000s.
  • The tools became asset freezes, travel bans, arms embargoes and commodity restrictions on diamonds, timber, charcoal and oil, and the unit of coercion became the named individual, not the state — with enormous consequences for the rights of the listed.
• Fifteen sanctions regimes are in force in 2026, covering Al-Shabaab, ISIL and Al-Qaida, Iraq, the Democratic Republic of the Congo, Sudan, Lebanon, the DPRK, Iran, Libya, the Taliban, Guinea-Bissau, Yemen, South Sudan, Haiti and Mali.
• Each regime has a sanctions committee of all fifteen Council members, deciding by consensus, so any one member — permanent or elected — can block a listing or a delisting.
  • Most committees are supported by a panel or group of experts, an independent monitoring team that investigates violations and reports publicly.
  • Their renewal is itself vetoable: Russia’s veto ended the DPRK Panel of Experts in 2024, removing the monitoring while leaving the sanctions in place.

The 1267 regime and the due-process problem

• Resolution 1267 of 15 October 1999 created the sanctions regime now covering ISIL and Al-Qaida, and it went further than any earlier regime.
  • It applies globally and indefinitely, tied to no territory, conflict or state, and listing imposes a worldwide asset freeze, travel ban and arms embargo on a named individual, binding every member state.
• The procedure was, for a decade, the sharpest human-rights problem the Council had created. A name could be added on one state’s proposal, on undisclosed evidence, by a committee sitting in secret.
  • There was no notice, no hearing, no reasons and no appeal, and delisting needed the same consensus that listing did — so the proposing state could block removal indefinitely.
• The challenge arrived through domestic and regional courts rather than international ones.
  • Yassin Abdullah Kadi, a Saudi businessman listed in 2001, challenged the European measures implementing his listing.
  • In Kadi I (2008) the European Court of Justice annulled the EU implementing regulation as a violation of fundamental rights, holding that it would review the EU measure even though it gave effect to a Council decision.
  • Kadi II (2013) confirmed the approach, and Swiss, Canadian and British courts reached comparable conclusions in Nada, Abdelrazik and Ahmed.
  • The consequence was direct: member states were being told by their own courts that they could not lawfully implement a binding Council decision.
• The Council’s answer was institutional, not judicial. Resolution 1904 of 17 December 2009 created the Office of the Ombudsperson to the 1267 Committee.
  • The Ombudsperson receives delisting petitions, gathers information, engages the petitioner and the designating states, and makes a reasoned recommendation to the Committee.
  • Resolution 1989 (2011) strengthened it decisively: a recommendation to delist now takes effect unless the Committee decides by consensus to keep the listing. This reverse-consensus rule shifted the burden onto those who want a listing kept, and delisting rates rose sharply.
• The reform is real and it is also narrow. The Ombudsperson covers the 1267 regime alone.
  • Every other regime offers only a Focal Point for De-listing, which transmits requests without any power to recommend.
  • There is still no independent tribunal, no disclosure of evidence and no hearing, so the mechanism remains administrative review rather than judicial remedy.

Compliance, and one piece of counter-design

• The Council imposes sanctions and has no capacity to enforce them. Implementation is wholly national, so compliance tracks capacity and political will — the DPRK regime has been eroded by ship-to-ship transfers and a shrinking appetite for enforcement among the Council’s own members.
• The Iran “snapback” of 2025 shows how far the design problem has been thought through in one case.
  • Resolution 2231 (2015), endorsing the nuclear deal, engineered a mechanism where reimposition of the earlier sanctions required a Council resolution to continue the suspension — which a supporter of Iran would have to veto.
  • France, Germany and the United Kingdom triggered it on 28 August 2025 and the previous UN sanctions on Iran returned on 28 September 2025.
  • The veto was structurally inverted: the same power that normally blocks action was made to compel it. It is the only significant instance of the drafters engineering around the veto from inside a resolution.

Creating institutions: tribunals, referrals and committees

The ad hoc criminal tribunals

• The Council created two criminal courts under Chapter VII, on the theory that prosecuting atrocity was an Article 41 measure to restore international peace and security.
  • The International Criminal Tribunal for the former Yugoslavia was created by resolution 827 of 25 May 1993.
  • The International Criminal Tribunal for Rwanda was created by resolution 955 of 8 November 1994, by 13 votes to 1 with 1 abstention — Rwanda itself voting against, objecting to the temporal scope, the shared appeals chamber and the exclusion of the death penalty; China abstaining.
• The legality of creating a court by executive fiat was challenged immediately, and the tribunal answered the challenge itself.
  • In Tadić (1995) the ICTY Appeals Chamber held that it had been lawfully established, that Article 41’s list is not exhaustive, and — significantly — that the Council’s powers are not unlimited and are bounded by the Charter.
  • That is the nearest thing to judicial review of the Council that exists, and it came from a body the Council had itself created.
• Both closed after building a substantial body of law on genocide, command responsibility and sexual violence in conflict — the ICTR in 2015, the ICTY in 2017.
• Resolution 1966 of 22 December 2010 created the International Residual Mechanism for Criminal Tribunals to carry the remaining trials, appeals, archives and witness protection — an inheritance that outlives both parents.

Referrals to the International Criminal Court

• The Rome Statute gives the Council two powers over the ICC that no other body holds.
  • Article 13(b) lets the Council refer a situation to the Prosecutor under Chapter VII, reaching states that are not parties to the Statute at all.
  • Article 16 lets it defer an investigation for twelve months, renewably — a power to stop the Court as well as start it.
• Two referrals have been made in twenty years.
  • Darfur, by resolution 1593 of 31 March 2005, adopted 11–0 with 4 abstentions (Algeria, Brazil, China and the United States) — the first referral ever made.
  • Libya, by resolution 1970 of 26 February 2011, adopted unanimously, which remains the only time all five permanent members supported a referral.
• Both referrals carried an exemption clause shielding nationals of non-party states from the Court’s jurisdiction, inserted at United States insistence — a carve-out that undercuts the equality claim of the whole exercise.
• Syria was not referred. A draft referral co-sponsored by 65 states was vetoed by Russia and China on 22 May 2014, and there has been no serious attempt since.
• The pattern is that referral is available only where no permanent member has a stake, which is the same limitation that governs everything else the Council does.

The legislating Council

• After 2001 the Council began adopting resolutions that impose general, abstract, open-ended obligations on all states, unconnected to any particular situation — a departure from anything it had done before.
• Resolution 1373 of 28 September 2001, adopted within seventeen days of the September 11 attacks, requires every state to criminalise terrorist financing, freeze terrorist assets, deny safe haven and tighten border controls.
  • It created the Counter-Terrorism Committee of all fifteen members to monitor implementation, with mandatory reporting by every member state.
  • Much of it was lifted from the 1999 Terrorist Financing Convention, then barely ratified — so the Council imposed by decision what states had declined to accept by treaty.
• Resolution 1540 of 28 April 2004 requires all states to prohibit non-state actors from acquiring nuclear, chemical and biological weapons and to enact export controls and physical protection measures, monitored by the 1540 Committee.
• The constitutional objection is serious and was made at the time by India, Pakistan, Brazil and others in the Assembly.
  • The Council was legislating: making rules of general application for the whole membership, indefinitely, rather than deciding on a situation.
  • Nothing in Chapter VII contemplates this. Article 39 presupposes a specific threat, and the treaty-making power belongs to states.
  • Stefan Talmon‘s characterisation of the Council as a world legislature framed the debate: fifteen states, five of them irremovable, making binding law for 193 without ratification, reservation or exit.
  • The defence is functional — the treaty route is too slow for a fast-moving threat and universal obligations need a universal instrument.

Peacekeeping: the power nobody wrote down

• Peacekeeping appears nowhere in the Charter. It was improvised in 1956 over Suez and built entirely on practice since.
• Hammarskjöld’s formulation of “Chapter Six and a Half” captures its position exactly: more than pacific settlement, less than enforcement, and authorised by neither chapter in terms.
• What is a Council power is the authorisation and the mandate: the Council creates each operation, sets its tasks and size, and renews or ends it, usually every six or twelve months.
• The doctrine, the generations and the contraction of peacekeeping are a subject of their own and are treated as such; the Council’s part in it is the mandate.

Institutional and appointment powers

• Admission of new members requires a Council recommendation before the Assembly can admit, under Article 4(2) — so the veto controls the door to the Organization itself. The application of Palestine was vetoed by the United States in April 2024.
• The Secretary-General is appointed by the Assembly on the recommendation of the Council, under Article 97, and the recommendation is a substantive decision subject to the veto.
  • The Council votes by straw poll, with “encourage”, “discourage” and “no opinion” ballots, and a candidate needs nine affirmative votes with no discourage from a permanent member.
  • A single permanent member can therefore end any candidacy without ever casting a formal veto — as the United States did to Boutros Boutros-Ghali in 1996.
• ICJ judges are elected by the Assembly and the Council voting separately and concurrently, an absolute majority required in each. The veto does not apply to these elections.
• Article 94(2) allows a party to an ICJ judgment that the other side will not perform to have recourse to the Council, which “may make recommendations or decide upon measures” to give effect to it.
  • The provision has been effectively dead since 1986, when the United States vetoed enforcement of the Nicaragua judgment against itself.
  • Any judgment against a permanent member or its close ally is therefore unenforceable by design, and the Court’s own remedies stop at the Council’s door.
• Charter amendment requires the Council’s concurrence twice over. Under Article 108 an amendment needs ratification by two-thirds of members including all five permanent members; under Article 109 the same applies to any revision by a General Conference.

How the Council Actually Works

Almost nothing about the Council’s daily operation is governed by the Charter, and very little of it is written down anywhere with binding force. The gap between the formal architecture and the working practice is where the elected members live.

Rules still provisional after eighty years

• Article 30 lets the Council adopt its own rules of procedure. It adopted Provisional Rules in 1946 and has never made them final.
  • They have been amended a handful of times, most recently in 1982 to add Arabic as an official language, and they remain formally provisional.
• Permanence would mean fixing obligations, and the permanent members prefer arrangements they can vary. Provisionality is not an oversight; it is a preference.
• The result is that most of the Council’s procedure lives in presidential notes, precedent and custom, none of which creates an entitlement anyone can invoke.

Where the work happens

• Formal meetings are public, minuted and webcast, and by the time one is convened the outcome is usually already settled.
  • The Council held 235 public meetings in 2025, against 20 private meetings and 115 closed consultations.
• Consultations of the whole are the real forum. They are held in a separate room off the chamber, are closed to non-members, and produce no verbatim record at all.
  • Drafts are negotiated line by line and disagreement aired in terms nobody would use on camera. The case for privacy is candour; the objection is that the Council’s real deliberation is unrecorded and unreviewable.
• Arria-formula meetings, named for Venezuela’s Ambassador Diego Arria who convened the first in 1992, are informal gatherings called by one or more members.
  • They let the Council hear non-governmental organisations, civil society and affected populations who could never be invited to a formal meeting, and they carry no formal status — which is exactly why they are possible.
• Horizon-scanning briefings let the Secretariat raise emerging crises before they become agenda items. They are held irregularly, and resisted whenever a member fears its own region appearing on the list.

Penholding and the distribution of drafting power

• A penholder is the member that drafts and leads negotiations on a file. The role appears in no rule and confers enormous influence.
  • The penholder decides when a draft is circulated, what it says and whether it is put to a vote at all.
• The P3 — the United States, the United Kingdom and France — hold most of the pens, an arrangement inherited from the 1990s and never formally agreed.
• The elected ten have pressed for wider distribution, with partial success. Co-penholding alongside a P3 state is now common, and China and Russia support broadening the system while rarely taking pens themselves — though China recently took the Afghanistan file.

Subsidiary bodies and the transparency argument

• Article 29 lets the Council establish subsidiary organs, and it has built an extensive machinery under it.
  • Thematic bodies including the Working Group on Children and Armed Conflict, the Informal Working Group on Documentation and Other Procedural Questions, and the Working Group on Peacekeeping Operations.
• Chairing a subsidiary body is the main substantive role available to an elected member, which is what made the 2026 deadlock over chair allocation so damaging.
• The Informal Working Group on Documentation is the vehicle for working-methods reform, and its product is the compendium known as Note 507.
  • First issued in 2006 and revised repeatedly, the current text is S/2024/507, adopted in December 2024.
• The transparency critique from the wider membership is long-standing and largely unanswered.
  • The Small Five (S5) — Costa Rica, Jordan, Liechtenstein, Singapore and Switzerland — pressed a working-methods resolution in the Assembly in 2012 and withdrew it under permanent-member pressure.
  • Its successor, the Accountability, Coherence and Transparency (ACT) group of some twenty-five states, has driven working-methods reform and the veto Code of Conduct ever since.
• What an elected seat actually buys is access, not influence — the consultations room, the briefings, the chairs, the drafting table.
  • They can convene, chair, co-penhold, block a committee consensus and use their presidency month, and increasingly act collectively — all ten co-sponsored the 2025 Gaza drafts.
  • They cannot overcome a veto, force a vote or carry a file their permanent colleagues have decided against, and their two years end before most files do.

The Veto

Article 27 is three sentences long and it determines almost everything about what the Council can and cannot do. Everything else in the Charter describes powers; this describes the condition on which they may be used.

Article 27 in its exact terms

• Article 27(1): each member of the Council has one vote. Formal equality inside the chamber is complete.
• Article 27(2): decisions on procedural matters need an affirmative vote of nine members — any nine.
• Article 27(3): decisions on all other matters are made by an affirmative vote of nine members including the concurring votes of the permanent members.
  • The word “veto” appears nowhere in the Charter. The power is the negative implication of a requirement for concurrence.
  • A permanent member’s negative vote defeats a draft even where the other fourteen are in favour, and no majority anywhere can overturn it.
  • The threshold is nine of fifteen, stated in the Charter as a number of members rather than as a fraction.
• Article 27(3) carries a proviso: in decisions under Chapter VI and under Article 52(3), a party to a dispute shall abstain from voting.
  • This is the obligatory abstention, and it is the Yalta compromise in its original form.
  • It applies only to pacific settlement, never to Chapter VII, so it has no effect where it would matter most.

The obligatory abstention nobody observes

• The proviso was taken seriously at first. There are eight documented instances of compliance between 1946 and 1951, four of them by India, three by the United Kingdom and one by Egypt.
• It then lapsed almost entirely. At least sixteen documented failures to abstain followed between 1952 and 1990, and the rule has been essentially dormant since.
  • Argentina abstained on the Eichmann resolution in June 1960 expressly citing Article 27(3) — among the last clear invocations.
• Three ambiguities are used to avoid it, and none has ever been resolved.
  • There is no agreed definition of “party to a dispute”, and a state can simply deny being one.
  • The Charter distinguishes a “dispute” from a “situation” without defining either, and the Council routinely calls matters situations.
  • Whether a draft rests on Chapter VI or Chapter VII is often left deliberately unstated.
• The result is that a permanent member votes on its own conduct as a matter of course, which is the single clearest example of the Charter’s text being overridden by convenience.
• The question has revived since 2022. The Pact for the Future of September 2024 endorses full adherence to Article 27(3), and Liechtenstein with thirteen other states circulated a draft resolution in March 2026 to define a party to a dispute.

The abstention convention and the empty chair

• The Charter requires the “concurring votes” of the permanent members. Read literally, an abstention is not a concurring vote and should block a decision.
• Practice decided otherwise almost immediately. From 1946 onward a voluntary abstention was treated as not preventing adoption, and the interpretation has never seriously been challenged since.
• The ICJ ratified the practice in the Namibia advisory opinion (1971), noting that abstention had for a long period been consistently interpreted as no bar to adoption, and that the procedure had been generally accepted by members.
  • This is amendment of a treaty by subsequent practice, achieved without a word of text changing — and it is the single most important such change in the Charter’s history.
  • It converted the veto from a requirement of positive assent into a requirement of non-objection, which is what makes abstention the workhorse of Council diplomacy.
• Absence was treated the same way, and the consequences were enormous.
  • The Soviet Union boycotted the Council from January to August 1950 in protest at the Republic of China’s continued occupation of the Chinese seat.
  • In its absence the Council adopted resolution 83 of 27 June 1950, recommending assistance to the Republic of Korea, and resolution 84 of 7 July 1950, establishing the unified command under the United States.
  • Nothing on that scale has been possible since, because no permanent member has ever again left the chamber empty.

The Statement of 7 June 1945 and the double veto

• The Charter says procedural matters cannot be vetoed but does not say who decides what is procedural, and the omission was noticed at San Francisco.
• The Statement of the Four Sponsoring Governments answered that the preliminary question is itself substantive, and therefore vetoable — creating the double veto in two steps.
  • Step one: a permanent member votes that the question of characterisation is substantive, and vetoes any finding that a matter is procedural.
  • Step two: having established that the matter is substantive, it vetoes the matter itself.
• The manoeuvre is powerful because it makes the procedural category collapsible at will: nothing is safely procedural if a permanent member says otherwise.
• It was used through the early Cold War and has not been used since 1959.
  • The permissibility question was never resolved: the Statement is an interpretive declaration, not part of the Charter, and was contested from the day it was issued.
  • The practice ended by informal agreement among the permanent members, not by any ruling — which means it remains available.

The chain veto and the hidden veto

• The chain veto is the use of successive vetoes to hold a position over time — most starkly in the eighteen drafts vetoed on Syria since 2011, all by Russia and ten of them jointly with China.
• The hidden or “pocket” veto is the more important phenomenon, and it is invisible in every published count.
  • A draft that will attract a veto is usually never tabled at all — withdrawn, watered down, turned into a presidential statement, or abandoned in consultations before it becomes a Council document.
  • The threat is communicated privately — in consultations, in bilaterals, sometimes only by implication — and leaves no paper trail, because written records exist mainly for drafts expected to pass.
• Council inaction on Rwanda, Sri Lanka, Chechnya, Sudan and Myanmar owes more to anticipated vetoes than to cast ones.
• Counting vetoes therefore understates the veto’s effect, probably by a wide margin. A cast veto is the residue of a bargaining process the veto has already shaped — it counts the times a sponsor forced the issue anyway.
• Diplomatic self-censorship is the veto’s largest output, and no reform grounded in transparency can reach it.

The vetoes that are counted are the failures of the veto system to work quietly; its successes leave no record at all.

The record, by member and by period

• Between 1946 and the end of 2025 the veto was cast on the order of 320 occasions against some 264 draft resolutions — roughly Russia and the Soviet Union 155, the United States 94, the United Kingdom 30, China 21 and France 18.

Period	Dominant user	What was being blocked
1946–1955	Soviet Union	Membership applications and Western-sponsored drafts; the USSR was in a standing minority of one
1956–1969	Soviet Union, declining	Cold War crises; the United States had no need to veto while it commanded the majority
1970–1990	United States	Israel and Palestine, and southern Africa alongside British and French vetoes on Rhodesia and Namibia
1991–2000	Almost nobody	Only about ten vetoed drafts in the whole decade — the quietest period in the Council’s history
2011–2022	Russia, with China	Syria above all, then Ukraine after 2014
2023–2026	Russia and the United States	Ukraine, and Gaza, where the United States has cast almost all its recent vetoes

• The first veto was Soviet, on 16 February 1946, over the withdrawal of foreign troops from Syria and Lebanon.
• The United States cast its first veto only on 17 March 1970, on Southern Rhodesia — a quarter of a century in which it never needed one.
• Neither the United Kingdom nor France has cast a veto since 1989, which is why both can advocate restraint at no cost.
• 2024 saw seven vetoed drafts and eight vetoes cast — the most since 1986 (Russia four, the United States three, China one).
• 2025 saw four: two United States vetoes on Gaza drafts tabled by all ten elected members, and two Russian vetoes on amendments to a Ukraine resolution.
• In 2026 a single veto had been cast by mid-year, when China and Russia blocked a Gulf-sponsored draft on shipping in the Strait of Hormuz on 7 April, eleven members voting in favour.

The case against the veto

• It disables the Council precisely where its powers matter most. The gravest crises are those in which a permanent member has a stake, and those are the cases the veto is designed to close.
• Permanent members veto criticism of their own conduct — Russia on Crimea in 2014 and Ukraine since 2022, the United States on Israel — and it shields clients as reliably as patrons, which is the substance of the Syrian record.
• It is a legal privilege attached to a 1945 distribution of power and to a Charter that names a state which no longer exists — indefensible on any principle of sovereign equality the Organization otherwise professes.
• It pushes states to act outside the Council, and every unauthorised use of force justified by Council paralysis weakens the prohibition in Article 2(4) that the Council exists to uphold.
• It empties other provisions of content. Article 94(2) cannot enforce a judgment against a permanent member; Article 27(3) is ignored; Article 4 admission is hostage to it.

The case for the veto, made seriously

• The Council cannot enforce against a permanent member, and a Council pretending otherwise would break. This is the argument the drafters actually made, and it is not obviously wrong.
  • Enforcement against a nuclear-armed great power means defiance or general war, and the second is what the Organization exists to prevent.
  • The League’s fate is the evidence: it promised collective enforcement, could not deliver against Japan, Italy or Germany, and lost its authority trying.
• The veto is what keeps the great powers inside the institution. An organisation that could bind them against their vital interests is one they would leave — and a UN without the major powers is the League again.
  • The privilege was tied from the outset to a matching responsibility for peace and security, which in 1945 only those states had the capacity to discharge.
• Gareth Evans’s formulation is the tightest statement of the case: the veto ensures the United Nations does not commit itself to things it could not follow through because of great-power opposition.
• Unanimity requirements are ordinary in consent-based legal orders, from treaty amendment to the EU’s own foreign-policy decisions. Singling out the Council’s version is a political, not a legal, judgement.
• It also restrains the powerful from each other: no permanent member can use the Council’s authority to legitimate its own project, which is a real constraint on unilateralism dressed as multilateralism.
• The honest weighing is this. The veto is defensible as a reason the Council cannot enforce against a great power; it is not defensible as a reason the Council cannot speak, investigate, refer or authorise humanitarian access.
  • Blocking enforcement against oneself is the bargain of 1945. Blocking a fact-finding mission, an aid corridor or a referral is an extension of that bargain far beyond what it was struck for, and it is on that extension that the case for the veto fails.

Whether the veto is amendable at all

• Articles 108 and 109 make the veto self-entrenching. Any amendment requires ratification by two-thirds of the membership including all five permanent members, so the veto protects itself.
• No power holding a veto has ever consented to its own dilution, and none has indicated it would.
• This is the hardest fact in the whole debate: the veto is not a rule that has resisted reform; it is a rule that cannot be reformed against the will of those who hold it.

The Council’s Record as an Institution

Output over time

• Council output tracks great-power relations almost exactly, which is the strongest single piece of evidence for what the institution is.
  • Through the Cold War the Council was largely inert, adopting a handful of resolutions a year and vetoed into silence on most crises.
  • The post-1990 surge brought new mandates, sanctions regimes and tribunals, and a decade in which vetoes almost ceased.
  • Output has since fallen back sharply. Resolutions ran to 57 in 2021, 54 in 2022, 50 in 2023, 46 in 2024 and 44 in 2025 — the lowest since 1991.
• The collapse of consensus is the sharper indicator. Between 2014 and 2023, about 84 per cent of resolutions were adopted unanimously; in 2025 the figure was 61 per cent, and in the first half of 2026, 57.
• Presidential statements and press statements are what a divided Council produces instead of resolutions, since both require consensus but neither binds.
  • Presidential statements fell to eight in 2025 and two in the first half of 2026; press statements to 34, against a decade average near 88. When even these fail, the Council’s output is nothing at all.

What counts as a threat to international peace and security

• Article 39 has no definition, and the Council has steadily widened what it will accept as a threat.
  • Internal conflict: resolution 794 of 3 December 1992 authorised force in Somalia to secure humanitarian relief, without a government to consent — the first such authorisation inside a state.
  • Terrorism as a standing threat, in 1373 and after.
  • Non-proliferation and non-state actors, in 1540.
  • Public health: resolution 1308 of 17 July 2000 on HIV/AIDS and peacekeeping; resolution 2532 (2020) demanding a general cessation of hostilities during the COVID-19 pandemic.
  • Sexual violence in conflict, children and armed conflict, and the women, peace and security agenda from resolution 1325 (2000).
• Climate change is the boundary the Council has not crossed. An Irish and Nigerien draft framing climate as a security concern was vetoed by Russia in December 2021, with India voting against and China abstaining, despite 113 co-sponsors.
• The objection is jurisdictional, and India has made it as consistently as anyone. Developmental, health and environmental questions belong to the Assembly, ECOSOC and the specialised treaty regimes, where all 193 members vote.
  • Mandate creep transfers subjects from universal bodies to a fifteen-member body with five permanent seats, expanding the reach of the veto rather than of the UN.
  • The counter-argument is that threats do not respect the Charter’s committee structure, and a body that cannot discuss what destabilises states is not doing its job.

An honest balance

• The Council is the sharpest available evidence that the United Nations reflects the distribution of power rather than transcending it. It works when the great powers agree, and eighty years of record show it doing exactly that and little else.
• The counter-case deserves stating, because a paralysed Council is not the same as a useless one.
  • It supplies a standing forum where adversaries meet continuously and in person, with value independent of outcomes.
  • It confers legitimacy states plainly want — which is why they lobby for authorisation they could act without.
  • Its panels of experts produce information available nowhere else, and its routine work continues through the paralysis — mandate renewals, sanctions administration, peace operations — on the many files where no permanent member has a stake.

India and the Council

• India has served eight elected terms: 1950–51, 1967–68, 1972–73, 1977–78, 1984–85, 1991–92, 2011–12 and 2021–22 — among the most of any member that has never held a permanent seat.
• India supplies the clearest historical instance of compliance with the obligatory abstention, abstaining under Article 27(3) four times in the Council’s first five years.
• The 2021–22 term shows what an elected member can and cannot do.
  • India chaired the Taliban Sanctions Committee and the Libya Sanctions Committee, and chaired the Counter-Terrorism Committee in 2022.
  • Its August 2021 presidency produced the Council’s first presidential statement on maritime security, at a debate chaired by the Prime Minister.
  • It drafted resolution 2593 of 30 August 2021 on Afghanistan, adopted with Russia and China abstaining.
  • It took the CTC to Delhi in October 2022, producing the Delhi Declaration on terrorist use of new technologies. What it could not do was move any file on which the permanent members were divided.
• India voted against the December 2021 climate-security draft, on the jurisdictional argument that climate belongs to the universal treaty process rather than to the Council.
• India launched its candidature for the 2027 election to the 2028–29 term on 13 July 2026, under the theme SHANTI, in what will be a contested Asia-Pacific race.
• India’s claim to a permanent seat is a different argument on a different footing, and belongs with the reform debate rather than with the Council’s present composition.

Where the Reform Argument Begins

This is the Council as it is: fifteen members, five of them irremovable, powers wider than any comparable institution has held, and a voting rule that makes those powers unusable against the states holding them. The case for changing it — the blocs and their models, the veto-restraint initiatives, the amendment barrier and India’s own claim — is a subject in its own right and is treated as one.