The Convention on the Prohibition of the Development, Production and Stockpiling of Bacteriological (Biological) and Toxin Weapons and on their Destruction, known as the #Biological_Weapons_Convention, was opened for signature in 1972 and entered into force in 1975. It was the first treaty in history to ban an entire category of weapons of mass destruction. Half a century later, the treaty has 189 States Parties, a strong moral standing, and almost no machinery to check whether states are keeping their promises. This article examines that contradiction. It traces the origins of the Convention, analyses its fifteen articles, follows its institutional growth through nine Review Conferences, and studies the long failure to agree on a #verification system. It then looks at the current Working Group on the Strengthening of the Convention, established in 2022, and at the pressures created by rapid change in the life sciences, including gene editing, #synthetic_biology, and #artificial_intelligence. The study uses a qualitative, doctrinal method based on the treaty text, official conference documents, and peer reviewed literature published mainly in the last five years. It finds that the Convention survives less because of its legal design and more because of the strength of the #norm it carries. It also finds that the treaty's weakness is not only technical but political, and that the divide between industrialised states and the #Global_South over #Article_X shapes almost every negotiation. The article concludes with a set of realistic recommendations built around graduated transparency, an institutionalised science advisory function, stronger #national_implementation, and a modular approach to compliance assurance rather than a single grand bargain. Keywords: Biological Weapons Convention; biological disarmament; verification; confidence building measures; biosecurity; international law; dual use research; arms control 1. Introduction Few treaties carry as much moral weight and as little institutional power as the #Biological_Weapons_Convention. Signed in London, Moscow, and Washington on 10 April 1972, and in force since 26 March 1975, it prohibits the development, production, stockpiling, acquisition, and retention of biological agents and toxins that have no peaceful justification, together with the weapons and delivery systems designed to use them for hostile purposes. It was the first multilateral #disarmament treaty to outlaw a whole class of weapons of mass destruction, and it did so at a time when the Cold War made almost every arms control agreement a difficult and slow negotiation. The achievement is real. Today no state openly claims to hold biological weapons, and no government defends their use as legitimate. That is not a small thing. Yet the Convention that carries this prohibition is a short document of fifteen articles, with no international organisation of its own, no inspectors, no laboratories, and, until recently, a permanent staff of only a handful of people. The #Chemical_Weapons_Convention has a large implementing body with a mandate to inspect. The Treaty on the Non Proliferation of Nuclear Weapons relies on an international agency with safeguards agreements and technical monitoring. The BWC has none of this. It has annual reports that states may or may not send, a complaint procedure that leads to the Security Council, and a small support office in Geneva. This gap between the ambition of the prohibition and the modesty of the machinery is the central puzzle that this article addresses. The puzzle is not new. It has been discussed since the 1980s, and it produced one serious attempt at repair, the negotiations for a legally binding protocol that ran from 1995 to 2001 and collapsed when the United States rejected the composite text. Since then the treaty has moved forward in small steps, through Review Conferences, expert meetings, and voluntary information exchanges. In December 2022 the Ninth Review Conference created a Working Group on the Strengthening of the Convention with a mandate that covers international cooperation and assistance, science and technology, #confidence_building, compliance and #verification, national implementation, assistance and response, and the organisational and financial arrangements of the treaty. That Working Group has met repeatedly since 2023 and is preparing recommendations for the Tenth Review Conference, expected no later than 2027. The timing matters. The life sciences have changed more in the last two decades than in the previous fifty years. Gene editing tools are cheaper and more precise. Commercial DNA synthesis is a global service industry. Machine learning systems can now support biological design work that once required years of specialised training. The COVID-19 pandemic, whatever its origin, showed every government on earth what a fast spreading pathogen can do to health systems, economies, and social trust. At the same time, the political environment for arms control has become harsher. Accusations and counter accusations about biological laboratories, made during the war in Ukraine, showed how easily the treaty's consultative mechanisms can be used for political theatre rather than for genuine clarification. This article asks three questions. First, what exactly does the Convention prohibit, and how strong is that prohibition as a matter of #treaty_law? Second, why has the treaty failed for fifty years to build a #compliance mechanism, and are the reasons technical, political, or both? Third, what realistic options exist now, given the state of the Working Group, the pace of scientific change, and the distribution of interests among States Parties? The argument developed here has three parts. The first is that the legal core of the treaty, the general purpose criterion in Article I, is stronger and more flexible than it is usually given credit for, because it covers any biological agent regardless of when it was discovered or how it was made. The second is that the absence of #verification is best understood as a political choice repeated across decades rather than as a purely scientific impossibility, although the scientific difficulties are real and should not be dismissed. The third is that the most promising path forward is not a single comprehensive protocol but a layered system: better transparency, an institutional science advisory body, a functioning cooperation mechanism under #Article_X, stronger national law, and a set of compliance assurance tools that states can adopt in stages. The article is written for students of international law, security studies, public health, and the life sciences. It uses plain language, but it follows the structure of a research article: literature review, method, findings organised by theme, discussion, limitations, and conclusion. 2. Literature Review The scholarly literature on the #Biological_Weapons_Convention can be grouped into five streams. The legal and doctrinal stream. This work analyses the text of the treaty, the meaning of its provisions, and the additional understandings adopted by Review Conferences. Legal scholars have paid particular attention to the general purpose criterion in Article I, which prohibits agents and toxins of types and quantities that have no justification for prophylactic, protective, or other peaceful purposes. The phrasing is deliberately open. It does not list banned pathogens. It bans purpose. This design choice means that the treaty automatically covers agents that did not exist in 1972, including engineered organisms, which is why many writers describe it as technologically neutral and future proof in its scope, even when they criticise it in every other respect. Recent legal writing also examines the relationship between the Convention and the 1925 #Geneva_Protocol, which banned the use of biological and chemical weapons in war but left development and stockpiling untouched. The verification stream. This is the largest body of work. It runs from the technical studies of the VEREX group in the early 1990s to contemporary proposals for graduated or modular approaches. Drobysz (2020) reviews the relationship between verification and implementation and argues that the two cannot be separated, since a treaty cannot be verified if states have not first built the national systems that produce verifiable information. Cropper and colleagues (2023) propose a modular and incremental approach to compliance verification, arguing that no single activity can prove compliance and that the treaty should therefore combine several imperfect tools into a system that raises confidence in stages. Shearer and colleagues (2023) develop the concept of assurance rather than verification, on the view that states may accept measures that increase certainty about compliance even if they refuse a formal inspection regime. Millett and colleagues (2022) revisit the feasibility of on site activity, and Kim and colleagues (2025) explore whether distributed ledger technology could support tamper resistant records of laboratory activity. What these studies share is a move away from the idea of a single treaty wide inspection system and towards a set of layered tools. The transparency stream. This work studies the #confidence_building_measures introduced at the Second Review Conference in 1986 and revised several times since. Sharma and Hobson (2025) provide the most systematic recent analysis, using a mixed method design that combines quantitative analysis of participation rates with qualitative analysis of working papers and national statements. They show that participation is uneven, that many submissions are incomplete or out of date, and that the barriers are often administrative rather than political: small states lack the inter agency machinery to collect the information the forms demand. Krasny (2025) treats transparency and denial as complementary approaches and argues that states will only reveal sensitive information if they are also allowed to protect legitimate secrets, which suggests that any workable transparency system must be built with defined limits. The science and technology stream. Dando (2023) analyses the broader chemical and biological nonproliferation regime after the COVID-19 pandemic and argues that the scientific revolution in the life sciences has outpaced the governance structures built in the twentieth century. Shang, Sheff, and Dando (2023) argue that scientists themselves must be brought into the treaty's processes, not as occasional guests but as a permanent advisory function. Lentzos (2022) examines the intersection of machine learning and biological weapons risk. Bao and colleagues (2025) study how scientific communities and new actors, including researchers from countries with limited diplomatic capacity, could be linked to the Convention through bottom up networks. Attal-Juncqua and colleagues (2024) situate the debate in the wider bioeconomy, arguing that safety, security, sustainability, and social responsibility must be designed into biotechnology from the start rather than added later. The political and institutional stream. Yassif, Korol, and Kane (2023) argue that preventing state biological weapons programmes requires shaping intentions and not only capabilities, since the technical barriers to misuse continue to fall. Boland, du Pont, and Lim (2025) study open source information and ask how the wider ecosystem of civil society, journalists, and researchers can contribute to detection. Feakes and colleagues (2025), writing from the perspective of the Implementation Support Unit, describe the long and unfinished road to institutionalisation, which is a polite way of saying that the treaty still has no proper organisation. Duque Estrada Meyer and colleagues (2025) describe the Convention as a system that adapts by absorbing external pressure, surviving even when formal negotiations stall. The gap in the literature. Most of these studies are strong in their own area but narrow in scope. Legal analyses often stop at the text. Verification studies often assume that the political obstacles can be engineered away. Transparency studies often treat participation as an administrative problem. What is missing is an integrated account that connects the legal design of the treaty, the political history of the verification failure, the science and technology pressure, and the #Global_South agenda on #Article_X, and then draws practical conclusions from that whole picture. This article attempts that integration at a level suitable for advanced students and early career researchers. 3. Methodology This is a qualitative study that combines doctrinal legal analysis with policy and institutional analysis. It does not generate new empirical data. Its contribution is synthesis and interpretation. Sources. Three categories of material were used. The first is primary legal material: the text of the Convention itself, the 1925 Geneva Protocol, and the final documents and additional understandings adopted by the nine Review Conferences held between 1980 and 2022. The second is official documentation from the Meetings of States Parties and from the Working Group on the Strengthening of the Convention, including national and joint working papers submitted between 2023 and 2026. The third is peer reviewed scholarship, with a deliberate preference for work published within the last five years, supplemented by a small number of older studies where they remain the standard reference for a specific technical debate. Method of analysis. The doctrinal component reads the treaty in the manner used by international lawyers: the ordinary meaning of the terms, in their context, in the light of the object and purpose of the treaty, with later agreements between the parties treated as evidence of shared interpretation. The institutional component follows the treaty as an evolving political process. It asks what problems states have tried to solve, what solutions they have proposed, why those solutions have failed or partly succeeded, and what pattern the failures form. The thematic component groups the material under recurring themes: prohibition, #verification, transparency, cooperation, science, and national law. Boundaries of the study. This article does not discuss the technical characteristics of any biological agent, nor any method of production, weaponisation, or dispersal. Such material would be inappropriate in any publication and irrelevant to the argument. The study is concerned with law, institutions, and policy. Where historical programmes are mentioned, they are described only at the level of publicly established fact and only to explain the political history of the treaty. Limits of the method. Doctrinal and institutional analysis cannot measure compliance, and this study makes no compliance judgement about any named state beyond what is already established in the public record and in the cited literature. Nor can it predict the outcome of the Tenth Review Conference. What it can do is explain the structure of the problem clearly enough that readers can follow the negotiations for themselves. 4. Historical Background: How the Convention Came to Exist 4.1 From the Geneva Protocol to the 1960s The modern legal treatment of biological weapons begins with the #Geneva_Protocol of 1925, adopted after the chemical warfare of the First World War. The Protocol banned the use in war of asphyxiating and poisonous gases and of bacteriological methods of warfare. It was an important step, but it had two serious weaknesses. It banned only use, not possession, so states could lawfully build and hold stockpiles. And many states ratified it with reservations that reduced it in practice to a no first use commitment, which meant they reserved the right to retaliate in kind. The result was a legal regime that discouraged first use while quietly encouraging states to keep a capability in reserve. Between the 1930s and the 1960s several major powers operated offensive biological weapons programmes. The details of these programmes are historically significant but they are not the subject of this article. What matters here is the political consequence: by the late 1960s, several governments had concluded that biological weapons were militarily unattractive. They were unpredictable, hard to control once released, slow to take effect compared with conventional or nuclear forces, and dangerous to the state that used them. This military judgement, more than moral revulsion alone, opened the door to a ban. 4.2 The negotiation In 1968 the United Kingdom proposed separating biological weapons from chemical weapons in disarmament negotiations. This was a tactical decision. Chemical weapons were seen as militarily useful by several states, and a joint negotiation would have stalled. Biological weapons, by contrast, were seen by many military planners as more trouble than they were worth. The separation made agreement possible, although it also created a long delay for chemical weapons, which were only banned in 1993. In 1969 the United States announced that it would end its offensive biological weapons programme unilaterally and destroy its stocks. This decision changed the negotiating landscape. The Soviet Union, which had initially resisted separating the two categories, eventually accepted the approach, and negotiations in the Conference of the Committee on Disarmament in Geneva produced a text in 1971. The Convention was opened for signature on 10 April 1972 and entered into force on 26 March 1975 after twenty two ratifications. 4.3 What was left out, and why it mattered The treaty that emerged was short. It contained no definitions section, no list of prohibited agents, no verification annex, no implementing organisation, and no dispute settlement machinery beyond a reference to the Security Council. Some of these omissions were deliberate simplifications. Others were the price of agreement. The most consequential omission was the absence of any independent means of checking compliance. Negotiators in the early 1970s accepted this partly because they believed that biological weapons were of limited military value, partly because verification was thought to be technically impossible in a field where the same equipment and knowledge serve both medicine and warfare, and partly because Cold War politics made intrusive inspection unacceptable to at least one major party. The consequences of that choice became clear within a decade, when it emerged that a State Party had built a large offensive programme in secret while remaining formally bound by the treaty. That episode shaped every later argument about #verification, and it is why the phrase "trust without verification" is still used as a criticism of the regime. 4.4 The Convention in its own terms It is worth pausing on what the drafters did achieve. They created a prohibition that was general rather than list based. They imposed obligations on development, production, stockpiling, acquisition, retention, and transfer, and they required destruction or peaceful diversion of existing holdings. They prohibited assistance and encouragement to others. They accepted, at least on paper, an obligation to share the benefits of peaceful biology. In 1972 this was a significant piece of legal engineering, and much of the criticism directed at the treaty since then is really criticism of what states refused to add later, not of what the drafters wrote at the time. 5. The Legal Architecture of the Convention The Convention has fifteen articles. Most students only ever read the first one. The others matter, and several of them explain why the treaty behaves the way it does in practice. 5.1 Article I: the general purpose criterion Article I is the heart of the treaty. States Parties undertake never in any circumstances to develop, produce, stockpile, or otherwise acquire or retain microbial or other biological agents, or toxins whatever their origin or method of production, of types and in quantities that have no justification for prophylactic, protective, or other peaceful purposes. The same article also prohibits weapons, equipment, and means of delivery designed to use such agents or toxins for hostile purposes or in armed conflict. Three features deserve attention. First, the prohibition is defined by purpose, not by biology. It does not name a single pathogen. This is the #general_purpose_criterion, and it is the treaty's most durable feature. An agent created by a technique invented in 2030 falls inside Article I on the day it exists, because the test is whether there is a peaceful justification for the type and quantity held, not whether the agent appears on a list drawn up in 1972. Lists age. Purpose does not. Second, the words about types and quantities that have no justification allow legitimate work to continue. A vaccine institute holding a dangerous pathogen for research is not violating the treaty. A public health laboratory maintaining reference strains is not violating the treaty. This flexibility is necessary, since a treaty that banned possession outright would ban medicine. But the same flexibility is the source of the #dual_use problem: the difference between a permitted defensive activity and a prohibited offensive one often lies in intent, scale, and context rather than in the equipment or the organism. Third, the phrase "never in any circumstances" is absolute. There is no exception for retaliation, no exception for war, and no exception for national emergency. In this respect the Convention is stronger than the Geneva Protocol, whose reservations left room for reprisal. Article I does not use the word "use". This omission is often raised by students. The better view, now widely accepted and reaffirmed by Review Conferences, is that use is prohibited in substance, because a state cannot use what it may not develop, produce, acquire, or retain, and because the Geneva Protocol independently bans use in war. The Fourth Review Conference in 1996 recorded the understanding that use is effectively covered by Article I. 5.2 Articles II and III: destruction and non transfer Article II requires States Parties to destroy or divert to peaceful purposes all agents, toxins, weapons, equipment, and means of delivery in their possession, as soon as possible and not later than nine months after entry into force. Article III prohibits transfer to any recipient, and prohibits assisting, encouraging, or inducing any state, group, or international organisation to acquire such items. Article III is the legal basis for export control cooperation and for the argument that a State Party cannot lawfully help a non party build a capability. 5.3 Article IV: national implementation Article IV obliges each State Party to take any necessary measures to prohibit and prevent the prohibited activities within its territory, under its jurisdiction, or under its control anywhere. This is the bridge between international law and domestic law. In practice it means criminal legislation, biosafety and #biosecurity regulation, export controls, and enforcement capacity. It is also the weakest link in the chain. Surveys of national legislation have repeatedly found gaps: states that have criminalised production but not development, states with no penalties for individuals, states with no export controls, and states whose laws exist on paper but have no enforcement machinery behind them. A prohibition that is not implemented nationally is a prohibition that cannot be enforced against a scientist, a company, or a terrorist group. 5.4 Articles V and VI: consultation and complaint Article V provides that States Parties undertake to consult and cooperate in solving problems that arise in relation to the objective of the Convention. This is the basis for the formal consultative meeting, a procedure used only twice in the treaty's history, most recently in 2022. Article VI allows a State Party that finds another in breach to lodge a complaint with the United Nations Security Council, and obliges parties to cooperate with any investigation the Council initiates. The design problem is obvious. The Security Council has five permanent members with a veto. A complaint against a permanent member, or against a close ally of one, will not produce an investigation. This is why Article VI has never been used successfully and why students should treat it as a formal rather than a functioning mechanism. In practice the more useful route has been the Secretary General's Mechanism, an investigative capacity based on earlier General Assembly and Security Council action, which allows the Secretary General to investigate alleged use of biological or chemical weapons. That mechanism sits outside the Convention, which is both its strength, because it avoids the treaty's consensus rule, and its weakness, because it is not owned by the States Parties. 5.5 Article VII: assistance Article VII commits States Parties to provide or support assistance to any party that requests it, if the Security Council decides that the party has been exposed to danger as a result of a violation. The condition is severe: assistance is triggered by a Security Council decision, which brings the veto back into the picture. In recent years, and especially after the COVID-19 pandemic exposed how badly the world responds to biological emergencies, several states have pushed to build a practical assistance and response capacity that does not depend entirely on the Council. Proposals include a database of available assistance, a voluntary fund, and pre agreed deployment procedures. This has become one of the more productive areas of the current Working Group, precisely because it is less politically loaded than #verification. 5.6 Articles VIII to XV: the surrounding structure Article VIII preserves the Geneva Protocol and makes clear that the Convention does not weaken it. Article IX records a commitment to negotiate a chemical weapons ban, a promise fulfilled in 1993. Article X, discussed in detail below, protects and promotes peaceful cooperation. Article XI allows amendment, but an amendment binds only those states that accept it, which makes amendment a poor tool for regime wide reform. Article XII requires review conferences, and this single provision has done more for the treaty's development than any other, because it forces states back to the table roughly every five years. Article XIII gives the treaty unlimited duration and allows withdrawal on three months notice in cases of extraordinary events, with notice to other parties and to the Security Council. Article XIV governs signature, ratification, and the three depositary governments. Article XV concerns authentic texts. The structural lesson is that the Convention was built as a statement of prohibition, not as an institution. It has an amendment procedure that is nearly unusable, an enforcement procedure that runs through the veto, and a review procedure that is its only real engine of change. 6. Institutional Evolution: Review Conferences, Confidence Building, and the Support Unit Because the treaty created no organisation, everything that now looks like institutional life has been built through #Review_Conference decisions. Nine such conferences have been held between 1980 and 2022. Each one added something, and several of them failed to add what was most needed. 6.1 The confidence building measures The Second Review Conference in 1986 agreed that States Parties would exchange information annually in order to reduce ambiguity, doubt, and suspicion. The Third Review Conference in 1991 expanded and formalised the scheme. The result is the set of #confidence_building_measures, politically binding but not legally binding declarations, submitted each year, covering research centres and laboratories, national biological defence programmes, outbreaks of infectious disease, publication of results, legislation and regulations, and vaccine production facilities. The CBMs are the only regular flow of compliance relevant information the treaty possesses. They are also chronically underused. Participation has improved over time and now covers roughly six in ten States Parties in a typical year, but that still leaves a large minority reporting nothing at all, and it leaves the quality of what is reported very uneven. Sharma and Hobson (2025) show that the causes are mixed. Some states withhold information deliberately. Many more simply lack the domestic machinery to gather it: the forms require input from health ministries, defence ministries, agriculture ministries, universities, and industry, and in a small administration nobody owns that task. Others submit the same document year after year with no updates. Some classify their submissions, which defeats the purpose of a #transparency instrument. The policy implication is important and often missed. If a large share of non participation is administrative rather than strategic, then targeted assistance, simplified forms, national focal points, and regional workshops will raise participation faster than political pressure will. That is a cheap and achievable reform, and it is one of the few areas where the Working Group has made visible progress. 6.2 The intersessional process The Fifth Review Conference, suspended in 2001 and resumed in 2002, was the low point of the treaty's diplomatic history. It ended without a substantive final document, in the wake of the protocol collapse. What it produced instead was a decision to hold annual meetings between Review Conferences. These intersessional meetings, expanded and reshaped in 2006 and 2011, allowed states to keep talking about biosafety, national implementation, science and technology, and assistance, even when they could not agree on anything binding. The intersessional process is often criticised as a talking shop. That criticism is fair in part. But it kept the treaty alive during a period when the alternative was silence, and it built the working relationships that made the 2022 Working Group possible. 6.3 The Implementation Support Unit The Sixth Review Conference in 2006 created the #Implementation_Support_Unit, a very small office within the United Nations Office for Disarmament Affairs in Geneva. Its functions are administrative support for meetings, support for national implementation, collection and distribution of the confidence building measures, and support for universalisation. It has no inspection function, no investigative power, and no authority to make findings about compliance. Its staff can be counted on one hand. Feakes and colleagues (2025), writing from inside that office, describe the treaty's institutional development as long and unfinished, and the description is accurate: the BWC is the only major weapons of mass destruction treaty without a proper implementing organisation. 6.4 Universalisation The Convention now has 189 States Parties. Four states have signed but not ratified, and four have neither signed nor acceded. Recent accessions include Tuvalu and the Federated States of Micronesia in 2024, and Comoros and Kiribati in 2025. This is close to universal, but it remains less universal than the Chemical Weapons Convention. #universalization work is slow, unglamorous, and effective: it consists largely of legal assistance to small states that have no objection to the treaty and simply need help completing the domestic steps. 7. The Verification Problem 7.1 Why verification is genuinely difficult The technical difficulty is real, and students should understand it before they read the political history. A nuclear weapons programme requires fissile material, and fissile material is hard to produce, easy to account for, and detectable. A chemical weapons programme requires specific precursor chemicals in militarily meaningful quantities, and those quantities are large enough to track. A biological programme has neither of these features. The same fermentation equipment used to brew beer, produce vaccines, or manufacture enzymes has many legitimate purposes and cannot be treated as evidence of anything on its own. Biological material multiplies, so accounting methods that work for chemicals and metals do not transfer. The relevant knowledge is published in open scientific literature. Facilities can look identical whether their purpose is defensive research or something else. And the line between permitted protective research and prohibited development is drawn by intent, which no instrument can measure directly. This is why #verification in the biological field cannot mean what it means at a nuclear reactor. It cannot deliver proof. At best it can deliver confidence, degrees of assurance, and a raised cost of concealment. Cropper and colleagues (2023) make exactly this point: no single verification activity can produce unambiguous evidence, so the design question is how to combine several imperfect tools so that the whole is more useful than any part. 7.2 VEREX and the Ad Hoc Group The Third Review Conference in 1991 created an ad hoc group of governmental experts, known as VEREX, to identify and examine potential verification measures from a scientific and technical standpoint. VEREX reported in 1993. It examined a long list of measures, both off site, such as declarations, information sharing, and remote monitoring, and on site, such as visits and inspections. Its conclusion was cautious but positive: no single measure could verify compliance, but a combination of measures could increase transparency and thereby contribute to confidence. A Special Conference in 1994 established an Ad Hoc Group with a mandate to negotiate a legally binding protocol. The Group worked from 1995 to 2001 and produced a composite text built around mandatory declarations of relevant facilities, routine visits to declared sites, challenge investigations of suspected violations, and an implementing organisation to run the system. In 2001 the United States rejected the composite text and stated that it would not accept a protocol of that kind. Its stated reasons were that the declarations and visits would not catch a determined violator, and that they would expose confidential business information belonging to its pharmaceutical and biotechnology industries as well as sensitive information about its own biodefence work. Other states, including some that supported the protocol, privately shared parts of that concern. Whatever the merits, the effect was decisive. The Fifth Review Conference collapsed, the Ad Hoc Group ceased to function, and #verification became a subject that could barely be raised in the treaty's meeting rooms for the next two decades. 7.3 The return of the question The 2022 Ninth Review Conference put compliance and verification back on the formal agenda for the first time in twenty years, as one of the seven topics assigned to the new #Working_Group. This is a genuine change, even if it produces no protocol. Three broad visions are now visible in the negotiations. The first is a return to the traditional model: a legally binding instrument with declarations, on site activity, investigations, and an international organisation. The second is a pragmatic approach that treats the objective, greater confidence in compliance, as fixed and the method as open, which may include legally binding elements but does not begin with them. The third is an incremental approach, which builds confidence through a series of national, bilateral, regional, and plurilateral steps that accumulate over time. The academic literature has moved decisively towards the second and third options. Shearer and colleagues (2023) argue for assurance rather than verification, a vocabulary shift with practical consequences: assurance measures are things a state does to demonstrate its own compliance, which is politically easier than submitting to a system designed to catch it out. Cropper and colleagues (2023) argue for modules that can be adopted separately, including voluntary transparency visits, peer review of biodefence programmes, standardised declarations, laboratory accreditation, and open source monitoring. Boland and colleagues (2025) examine the growing capacity of open source analysis, drawing on satellite imagery, procurement records, and patterns in the scientific literature, to raise the cost of concealment even without treaty based inspection. Kim and colleagues (2025) explore whether cryptographic record keeping could allow a laboratory to prove what it did without revealing proprietary details, which speaks directly to the confidentiality objection that helped defeat the 2001 protocol. None of these is a solution on its own. Together they describe a plausible system, and one that could be built in pieces. 7.4 Why the political obstacle is the real obstacle It is tempting to conclude that #verification failed for scientific reasons. That conclusion is too generous to the states involved. The scientific difficulties explain why verification cannot be perfect. They do not explain why it cannot exist at all. Chemical weapons verification is also imperfect, and it exists. Nuclear safeguards are also imperfect, and they exist. The honest explanation is that a verification system requires states to accept intrusion, and biological research sits at the intersection of three things governments guard most closely: national defence, commercial advantage in a very large industry, and public health infrastructure. Add a treaty with a consensus rule, in which any single state can block progress, and a security environment in which trust between major powers has fallen sharply, and the outcome is not surprising. Krasny (2024, 2025) frames the problem well: any workable system must be built to protect legitimate secrets, or states will not join it, and a system that protects too much will not reassure anyone. That is the needle the Working Group is trying to thread. 8. Compliance, Allegations, and the Politics of Suspicion 8.1 The historical shadow The treaty's compliance history is dominated by one fact. A founding State Party maintained a large offensive programme in secret for years after entry into force. The details of that programme were revealed gradually, through defector testimony and later official acknowledgement, and they are documented in the historical literature. The point that matters for this article is the lesson States Parties drew: a treaty with no means of checking anything can be violated on a very large scale without detection by its own machinery. Every argument for #verification since the 1990s rests on that lesson, and every argument against it has had to answer for it. A second historical case involved a state that concealed a weapons programme while a signatory, and whose programme was uncovered only through an intrusive inspection regime imposed after a war, not through the Convention itself. That case reinforced two conclusions at once. It showed that offensive programmes can be hidden. It also showed that even intense inspection produces contested findings, which is the argument sceptics use against building an inspectorate. 8.2 Allegations as a political instrument The Convention's consultative machinery has been used very rarely. Article V was invoked formally in 1997 and again in 2022, when the Russian Federation requested a consultative meeting over its claims about biological laboratories in Ukraine. That meeting was held in September 2022. No State Party supported the allegations. Russia then took the matter to the Security Council under Article VI, where the proposed investigation did not gain the required support. The episode is instructive for three reasons. First, it showed that the Convention's procedures can be triggered, which is better than nothing. Second, it showed that they produce no authoritative finding of fact, because there is no independent body to establish facts. The meeting can only record that states disagreed. Third, it showed how the absence of a fact finding mechanism creates space for disinformation. When there is no neutral referee, an accusation and a denial have equal formal status, and the audience is left to choose according to prior political loyalty. This is a structural weakness, not an accident, and it is one of the strongest practical arguments for building an independent technical capacity even if it never conducts a single inspection. 8.3 Biodefence and the transparency dilemma Many states run biological defence programmes. These are lawful under Article I, provided the types and quantities held have a protective justification. But defensive research can look very similar to prohibited development. Testing the effectiveness of a vaccine requires the pathogen. Understanding how an agent might be made more dangerous is sometimes argued to be necessary in order to defend against it. This is the sharpest edge of the #dual_use problem, and it is where suspicion grows fastest. The transparency dilemma is that the states most able to reassure others are also the states with the most to protect. Krasny (2025) argues that transparency and denial should be seen as complementary rather than opposed: a state can disclose the existence, budget, oversight structure, and legal framework of a programme while withholding operational detail. Systematic, comparable, and verified disclosure at that level would raise confidence considerably, and it would cost far less than a full inspection regime. 9. Science, Technology, and the Pressure on the Regime 9.1 The changing baseline When the Convention was drafted, the relevant capability sat in a small number of state laboratories. Today the tools of modern biology are distributed across thousands of universities and companies, in every region of the world. Gene editing is taught to undergraduates. Genetic sequences are digital information that can be transmitted instantly. Commercial synthesis firms will produce DNA to order. Automation reduces the tacit skill once needed for laboratory work. These developments are overwhelmingly positive: they underpin vaccines, cancer therapy, drought resistant crops, and rapid outbreak response. They also change the shape of the risk, because they lower the barriers that once did much of the work of nonproliferation without anyone noticing. 9.2 Artificial intelligence #artificial_intelligence has become the most discussed of these developments. Machine learning systems are now used to predict protein structure, design molecules, and plan experiments. Lentzos (2022) notes that the significance for biological weapons risk lies less in any single capability than in the general acceleration and democratisation of design work. The policy debate has focused on whether advanced models could provide meaningful assistance to actors who would otherwise lack the necessary expertise, and on what safeguards developers and governments should require. Within the BWC context, states have begun to discuss AI in working papers and side events, and the topic featured in the Working Group's discussions on science and technology review. Two cautions are appropriate. The first is that risk assessment in this area is young, and confident claims in either direction, that AI changes everything or that it changes nothing, are not well supported. The second is that the Convention is not the only, or even the main, place where this issue will be governed. Export controls, biosecurity screening of synthesis orders, publisher and funder policies, and model developer safeguards all matter, and none of them sits inside the treaty. The realistic role of the BWC is to hold the normative line, to provide a forum where states share assessments, and to ensure that the general purpose criterion is understood to cover whatever the technology produces. 9.3 The case for a standing science mechanism The Convention has no permanent scientific advisory body. This is unusual for a treaty whose subject matter changes every few years. Review Conferences consider science and technology every five years, which is far too slow. Shang, Sheff, and Dando (2023) argue that scientists must be given a formal, continuing role, and the International Science Council (2025) has produced expert guidance to the Working Group on precisely this point. Bao and colleagues (2025) add an important dimension: the science advisory function should not be dominated by researchers from a handful of wealthy states, because a mechanism that is seen as the property of the technologically advanced will not be trusted by everyone else. The Ninth Review Conference mandated the Working Group to recommend a mechanism to review and assess scientific and technological developments and to advise States Parties. The design questions are practical: who appoints the experts, how is regional balance ensured, does the body report to States Parties or to a chair, and does it have any power beyond advice. On the last question, the drafts under discussion are clear that the mechanism would be advisory and consultative, with no decision making authority. That is a modest outcome, but a standing advisory body would still be the most significant institutional addition to the Convention since the creation of the Implementation Support Unit in 2006. 10. Article X: Cooperation, Development, and the North South Divide 10.1 What Article X says #Article_X provides that States Parties undertake to facilitate, and have the right to participate in, the fullest possible exchange of equipment, materials, and scientific and technological information for peaceful uses of biological agents and toxins. It further provides that the Convention should be implemented in a way that avoids hampering the economic or technological development of States Parties or international cooperation in peaceful biological activities. 10.2 Why it dominates the negotiations To many students the article looks like a secondary provision. In the negotiating rooms it is anything but. For a large group of States Parties, mainly from the Non Aligned Movement and the wider #Global_South, Article X is the part of the bargain that they were promised and never received. Their argument runs as follows. They accepted a prohibition on a class of weapons they were unlikely to build anyway. In exchange they were promised full access to peaceful biotechnology. Instead, they say, export controls, informal supplier arrangements, and restrictive licensing practices have limited their access to equipment and materials, while the states imposing those controls have kept their own industries far ahead. The states that maintain export controls respond that controls are required by Article III, which prohibits transfers that would assist acquisition of prohibited items, and by Security Council obligations on nonproliferation, and that responsible controls protect everyone. Both positions have merit. The result is a chronic stalemate in which progress on compliance and #verification is held hostage to progress on cooperation, and vice versa. 10.3 The cooperation mechanism The Ninth Review Conference mandated the Working Group to recommend a mechanism to facilitate and support the full implementation of international cooperation and assistance under Article X. Drafts discussed in 2024 and 2025 include an implementing programme, an advisory group, a database matching offers and requests, a steering group, and a voluntary fund. The International Science Council (2025) has proposed criteria for identifying capacity needs, prioritising funding, and improving project oversight, with an emphasis on equitable access to knowledge and technology. The prospects here are better than in the compliance area, for a simple reason: nobody has to give up a secret. The obstacles are money and control. Donor states want oversight of how funds are used and resist anything that looks like an obligation to transfer technology. Recipient states want a mechanism with real resources and not another database that nobody funds. If the Working Group delivers anything concrete to the Tenth Review Conference, an Article X mechanism is the most likely candidate, and it would be worth more than it appears, because progress here buys political goodwill that the compliance discussion badly needs. 11. National Implementation and Non State Actors 11.1 The gap between ratification and law Article IV requires national implementation, but the treaty says nothing about what that implementation should contain. Databases of national implementing legislation maintained by research organisations have repeatedly found that many States Parties have incomplete legal frameworks. Common gaps include the absence of criminal penalties covering all prohibited activities, no coverage of nationals acting abroad, weak or absent biosecurity regulation for laboratories, no controls on the transfer of relevant materials and technology, and no designated national authority to coordinate the treaty's requirements. These gaps matter for three reasons. First, without national law, a state cannot prosecute an individual who tries to acquire or make a biological weapon, and the Convention itself imposes no duties on individuals. Second, without national systems, a state cannot produce the information required for #confidence_building_measures, which is part of why participation is low. Third, without domestic regulation, any future compliance mechanism would have nothing to inspect and nobody to talk to. 11.2 Non state actors The Convention was written for states. It says nothing directly about terrorist groups, criminal networks, or individual actors, except indirectly through Article III and Article IV. The gap was addressed outside the treaty by Security Council Resolution 1540 in 2004, which obliges all states to adopt and enforce measures to prevent non state actors from acquiring weapons of mass destruction, including biological weapons, and to establish appropriate controls over related materials. Resolution 1540 has become, in practice, part of the biological nonproliferation regime, and its committee has driven much of the national legislation drafting of the last two decades. The relationship between the two instruments is complementary but awkward. The Convention is a treaty negotiated by its parties; Resolution 1540 is an obligation imposed by the Security Council on all states. Some States Parties resent the second as an intrusion on treaty based multilateralism. The practical answer is that both are needed, and that a state which implements Article IV properly will also satisfy most of what Resolution 1540 requires. 11.3 Scientists, industry, and codes of conduct Yassif, Korol, and Kane (2023) argue that prevention must work on intentions, not only on capabilities, because the technical barriers keep falling. Part of that work is cultural. Education in dual use awareness, codes of conduct for life scientists, biosecurity review in funding decisions, and screening by DNA synthesis providers all shift the environment in which a would be programme or a would be terrorist has to operate. None of this is legally binding under the Convention. All of it is consistent with it, and several Review Conferences have encouraged such measures. Attal-Juncqua and colleagues (2024) make a related argument for the wider bioeconomy: safety, security, sustainability, and social responsibility must be designed into biotechnology development from the beginning rather than added as an afterthought. 12. Discussion 12.1 The norm is doing the work The most striking finding of this study is how much of the Convention's effectiveness rests on the #norm rather than on the law. No state today admits to having biological weapons. No state argues that their use would be legitimate. This is a remarkable normative achievement for a treaty with no inspectors, and it is reinforced by the widely held military judgement that such weapons are unreliable and dangerous to the user. But norms are not self sustaining. They are maintained by repetition, by institutions, and by the visible cost of violation. A norm with no fact finding capacity behind it is vulnerable to exactly the kind of accusation and counter accusation seen in 2022, which erodes the norm from the inside by making it appear to be merely one more contested political claim. 12.2 The failure is political, and therefore reversible The technical arguments against verification are real but they are limits, not prohibitions. The literature of the last five years converges on a design that accepts those limits: not proof, but assurance; not a single system, but modules; not full disclosure, but structured disclosure with protection for legitimate secrets. That design exists on paper. What is missing is the political decision to build it, and political decisions can change, particularly when the perceived cost of the status quo rises. 12.3 The bargain has two sides Any analysis that treats the BWC only as a compliance problem will misunderstand its politics. For roughly two thirds of the States Parties, the pressing question is not inspection but access. Until the #Article_X agenda is treated as a serious obligation rather than as a bargaining chip, consensus on anything else will be difficult. This is not merely a diplomatic observation. It is a practical one: a cooperation mechanism that builds laboratory capacity, trains scientists, and strengthens disease surveillance in states that currently have little of any of these things would also, over time, build the national infrastructure without which no compliance mechanism can function. 12.4 Institutionalisation is the quiet priority The Convention's deepest weakness may be the one least discussed in public: it has almost no institution. A treaty with a five person support office cannot maintain a science advisory function, run a cooperation programme, service a verification system, and support universalisation. Whatever the Tenth Review Conference decides on the substance, it will fail in implementation unless it also decides to build an organisation adequate to the tasks it assigns. 12.5 What the Working Group is likely to deliver Based on the pattern of the sessions held between 2023 and 2026, three outcomes look plausible. A science and technology review mechanism, advisory in nature, is close to agreement in principle. An international cooperation and assistance mechanism, with a fund and a matching database, is also within reach. Progress on assistance and response under Article VII is real. On compliance and verification, the most likely outcome is a mandate to continue work, possibly through an open ended working group, rather than an agreed instrument. That is a modest result. It is not nothing, because it keeps the question alive, and because the first two mechanisms, if properly funded, would give the treaty something it has never had: a permanent capacity to learn and to deliver. 13. Limitations of This Study This article is a synthesis, not an empirical investigation. It does not test hypotheses, and it produces no new data on compliance, participation, or capability. Its account of the negotiations relies on public documents and on published analysis, which means that it cannot capture the private bargaining that determines many outcomes. Its assessment of what the Working Group will deliver is a judgement based on the visible pattern of sessions, and negotiations can change direction quickly. Finally, the study deliberately avoids all technical discussion of agents, production, and weaponisation, which means that readers seeking a technical risk assessment will need to look elsewhere, in appropriately controlled settings. 14. Conclusion and Recommendations The Biological Weapons Convention is a short treaty carrying a very large prohibition. For fifty years it has done something valuable: it has held the line that biology must not be turned into a weapon, and it has held that line so successfully that no government today argues otherwise in public. That is a genuine achievement of international law, and students should not be persuaded by the fashionable view that the treaty is a paper tiger. At the same time, the Convention is under strain that its architecture was never built to bear. The science has changed. The politics have hardened. The treaty still cannot establish a fact, cannot conduct an investigation without the Security Council, cannot advise itself on technology, and cannot deliver on the cooperation it promised in 1972. The current Working Group is the most serious attempt in twenty years to fix some of this, and it is working under a consensus rule that gives every state a veto. Six recommendations follow from the analysis. First, treat #transparency as the immediate priority. Simplify the confidence building measure forms, fund national focal points, provide regional assistance, and make submission the default expectation. This costs little and would improve the treaty's only information flow within a single reporting cycle. Second, establish the science and technology review mechanism, and design it for regional balance from the start. An advisory body is a small thing on paper and a large thing in practice, because it gives the Convention a memory and a capacity to think. Third, build the international cooperation and assistance mechanism, and fund it. Article X is not a favour to developing states. It is the other half of the bargain and it is the foundation of the national capacity that everything else depends on. Fourth, pursue #compliance assurance in modules rather than as a single protocol. Voluntary transparency visits, peer review of biodefence programmes, standardised and comparable declarations, laboratory accreditation, and structured use of open source information can each be adopted separately, and each raises the cost of concealment. Fifth, strengthen #national_implementation. Model legislation, legal drafting assistance, and a designated national authority in every State Party would do more for the regime's day to day effectiveness than another decade of protocol debate. Sixth, build the institution. The Implementation Support Unit needs to become an organisation with the staff and budget to run the mechanisms that States Parties keep inviting it to run. The Convention will not be saved by a single dramatic negotiation. It will be maintained, or lost, through a long series of unglamorous decisions about forms, funds, focal points, and advisory bodies. Fifty years after entry into force, the task is not to reinvent the prohibition, which remains sound, but to build the machinery that the prohibition has always lacked. References Attal-Juncqua, A., Dods, G., Crain, N., et al. (2024). Shaping the future US bioeconomy through safety, security, sustainability, and social responsibility. Trends in Biotechnology, 42(6), 671-673. https://doi.org/10.1016/j.tibtech.2023.11.015 Bao, Y., Angeles, K. A., Hlashwayo, D. F., Thorat, K., Mandona, M., Masisi, K., Ahmad, F., and Kassegne, A. (2025). Enhancing the BWC through scientific communities and networks of new actors in the era of emerging science and technology. Health Security. https://doi.org/10.1177/23265094251387073 Boland, L., du Pont, J., and Lim, Y. B. (2025). The BWC at 50: exploring OSINT opportunities and challenges in the BWC ecosystem. Health Security. https://doi.org/10.1177/23265094251387084 Cropper, N. R., Rath, S., Teo, R. J. C., Warmbrod, K. L., and Lancaster, M. J. (2023). A modular-incremental approach to improving compliance verification with the Biological Weapons Convention. Health Security. https://doi.org/10.1089/hs.2023.0078 Dando, M. R. (2023). The chemical and biological nonproliferation regime after the COVID-19 pandemic: dealing with the scientific revolution in the life sciences. London: Palgrave Macmillan. Drobysz, S. (2020). Verification and implementation of the Biological and Toxin Weapons Convention. The Nonproliferation Review, 27(4-6), 487-497. https://doi.org/10.1080/10736700.2020.1823102 Duque Estrada Meyer, F. S., et al. (2025). Survival by design: the BWC's autopoietic response at fifty to a disaggregated biosecurity ecosystem. Health Security. https://doi.org/10.1177/23265094251381803 Feakes, D., et al. (2025). The long and unfinished road to institutionalization of the Biological Weapons Convention: a perspective from the Implementation Support Unit. Health Security. Ingesson, T. (2025). Large-scale nightmare or precision tool for covert operations? An assessment of the risks of genetic technology in the near future. International Journal of Intelligence and CounterIntelligence. International Science Council. (2025). Strengthening international cooperation and assistance under the Biological Weapons Convention. Paris: International Science Council. https://doi.org/10.24948/2025.06 Kim, S. J., Huot, V., and Neilley, V. (2025). Blockchain in biosecurity: a case study for strengthening the Biological Weapons Convention. Health Security. https://doi.org/10.1177/23265094251381802 Krasny, J. (2024). The Biological Weapons Convention: challenges and opportunities. GCSP Policy Brief No. 15. Geneva: Geneva Centre for Security Policy. Krasny, J. (2025). Strengthening the Biological Weapons Convention: transparency and denial as complementary approaches. Journal for Peace and Nuclear Disarmament. https://doi.org/10.1080/25751654.2025.2602260 Lentzos, F. (2022). AI and biological weapons. In T. Reinhold and N. Schornig (Eds.), Armament, arms control and artificial intelligence: the Janus-faced nature of machine learning in the military realm. Berlin: Springer. Millett, P., Alexanian, T., Appleton, E., Diggans, J., Montague, M., and Titus, A. (2022). Feasibility of on-site verification. Working paper. https://doi.org/10.2139/ssrn.4213018 Shang, L., Sheff, L., and Dando, M. (2023). A key role for scientists in strengthening the Biological Weapons Convention. Frontiers in Political Science, 5, 1265008. https://doi.org/10.3389/fpos.2023.1265008 Sharma, R. M., and Hobson, T. (2025). Gaps in transparency: understanding limited participation in confidence-building measures under the Biological Weapons Convention. The Nonproliferation Review, 32(4-6), 283-309. https://doi.org/10.1080/10736700.2025.2591572 Shearer, M. P., Potter, C., Vahey, R. A., Connell, N. D., and Gronvall, G. K. (2023). BWC assurance: increasing certainty in BWC compliance. The Nonproliferation Review. https://doi.org/10.1080/10736700.2023.2178099 Yassif, J. M., Korol, S., and Kane, A. (2023). Guarding against catastrophic biological risks: preventing state biological weapon development and use by shaping intentions. Health Security, 21(4), 258-265. https://doi.org/10.1089/hs.2022.0145 #BWC_1972 #biological_weapons_convention #biological_disarmament #toxin_weapons #arms_control_law #nonproliferation_regime #global_biosecurity #Geneva_disarmament #treaty_verification #dual_use_dilemma #life_sciences_governance #WMD_prohibition #international_security_studies #BWC_at_fifty #strengthening_the_BWC

The Convention on the Prohibition of the Development, Production, Stockpiling and Use of Chemical Weapons and on their Destruction, commonly called the Chemical Weapons Convention (CWC), was opened for signature in Paris on 13 January 1993 and entered into force on 29 April 1997. It is the first multilateral treaty to ban an entire category of weapons of mass destruction and to attach to that ban a permanent international organisation with the power to enter the territory of member states and inspect their military and civilian facilities. This article offers a structured, student-facing review of the Convention. It examines four things in turn. First, the legal architecture of the treaty, with particular attention to the General Purpose Criterion, the definition of a chemical weapon, and the three Schedules of chemicals. Second, the institutional design of the Organisation for the Prohibition of Chemical Weapons (OPCW) and the way its three organs share authority. Third, the verification system in practice, covering destruction verification, industry inspections, the never-used challenge inspection, and the newer investigative and attribution mechanisms. Fourth, the pressures now bearing on the regime: the Syrian file, the use of nerve agents in assassination attempts, the argument over central nervous system acting chemicals, the rapid convergence of chemistry with the life sciences and with artificial intelligence, and the wider erosion of consensus among states. The article argues that the CWC has succeeded on the measure it set for itself in 1993, since every declared stockpile in the world has now been verifiably destroyed, but that the treaty is being asked to do a different job in 2026 than the one it was designed to do. Its future depends less on destroying old weapons and more on preventing the re-emergence of new ones, on holding violators to account, and on keeping a fractured membership working together. The article closes with policy recommendations and a research agenda for students entering this field. Keywords: Chemical Weapons Convention; OPCW; disarmament; verification; General Purpose Criterion; attribution; dual-use chemistry; international law 1. Introduction Few areas of international law can point to a result as clear as the one achieved by the #Chemical_Weapons_Convention. When the treaty entered into force in 1997, eight states had declared possession of chemical weapons and the world held tens of thousands of tonnes of nerve agent, blister agent, and the munitions built to deliver them. On 7 July 2023, the last munition in the declared global stockpile, an M55 rocket filled with the nerve agent GB, was destroyed at a facility in Kentucky in the United States. With that single act, every chemical weapon that any state had formally declared to the #OPCW had been eliminated under international supervision. No other #disarmament treaty has ever completed the task it was written to perform. Yet the same period saw #chemical_weapons used in Syria on a scale not witnessed since the Iran-Iraq war of the 1980s, saw a nerve agent deployed on the streets of an English cathedral city, saw another used against a Russian opposition politician, and saw a nerve agent smeared on the face of a man in a Malaysian airport terminal. The paradox is uncomfortable and it is the reason this subject remains alive. The stockpiles are gone. The weapons are not. This article is written for students who are meeting the CWC for the first time in a serious way, whether from international law, international relations, security studies, chemistry, or public policy. The aim is to explain what the treaty actually says, how its enforcement machinery works and where it does not work, and why informed observers disagree about its condition. Three questions organise the discussion. The first question is legal. What exactly does the Convention prohibit, and how did its drafters write a ban broad enough to cover chemicals that had not yet been invented? The answer lies in a piece of drafting called the #General_Purpose_Criterion, which is arguably the single most important idea in the whole text and which students routinely overlook. The second question is institutional. The CWC did something unusual for a #disarmament_treaty. It created its own permanent organisation, gave that organisation a standing corps of inspectors, and gave those inspectors a right of entry into the territory of sovereign states. How is that authority structured, who controls it, and what happens when a member state refuses to cooperate? The third question is political. A treaty is only as strong as the willingness of its members to enforce it. Since roughly 2013 the OPCW has been operating in an environment in which its findings are publicly disputed by permanent members of the United Nations Security Council, in which decisions that were once taken by #consensus are now taken by contested vote, and in which the organisation has been asked to name perpetrators rather than merely count barrels. What has that done to the regime? The article proceeds as follows. Section 2 reviews the historical and scholarly background. Section 3 sets out the method. Sections 4 and 5 examine the legal architecture and the institutional design. Section 6 describes the #verification system as it operates in practice. Section 7 assesses what the regime has achieved. Section 8 examines the compliance crises that have defined the past decade. Section 9 turns to the challenges now emerging from science, technology, and geopolitics. Sections 10 and 11 discuss the findings and offer recommendations. Section 12 notes the limits of the analysis, and Section 13 concludes. 2. Background and Literature Review 2.1 From Ypres to The Hague The modern legal effort to control poison in war predates the Convention by almost a century. The Hague Declaration of 1899 forbade projectiles whose sole object was the diffusion of asphyxiating gases. That prohibition failed almost immediately under the pressure of industrial warfare. Large-scale gas warfare on the Western Front from 1915 onward produced the imagery that still shapes public revulsion at these weapons, and it produced the #Geneva_Protocol of 1925. The Geneva Protocol was an important normative achievement and a weak legal instrument. It banned the use in war of asphyxiating, poisonous, and other gases, but it said nothing about development, production, or stockpiling. Many states ratified it with reservations that reserved the right to retaliate in kind, which converted a prohibition on use into something closer to a no-first-use pledge. States continued to build arsenals throughout the interwar years, throughout the Second World War, and throughout the Cold War. Edwards and colleagues (2022) trace this trajectory and note that the twentieth century saw not only the expansion of stockpiles but the deliberate refinement of agents for incapacitation, area denial, and assassination. The negotiation of a comprehensive ban took roughly two decades in the Conference on Disarmament in Geneva. Progress was slow because the core problems were hard. How do you define a chemical weapon when almost any toxic chemical could serve as one? How do you verify that a state is not producing nerve agent in a plant that also makes pesticides? How intrusive an inspection will sovereign states tolerate? The use of chemical weapons in the Iran-Iraq war, including the attack on the Kurdish town of Halabja in 1988, sharpened the political urgency. The end of the Cold War removed the strategic reason for the two superpowers to keep their arsenals. The text was finalised in 1992 and opened for signature in January 1993. 2.2 The scholarly conversation The academic literature on the CWC divides, broadly, into four conversations. The first is doctrinal and legal. It asks what the treaty text means, how its ambiguities should be resolved, and how it interacts with #international_humanitarian_law and #international_human_rights_law. The standard reference work in this tradition is the article-by-article commentary edited by Krutzsch, Myjer, and Trapp, and the tradition continues in edited collections such as the research handbook assembled by Myjer and Marauhn (2022), which situates the CWC within the wider body of arms control law. The second conversation is about #norms. Scholars in this tradition ask why states have largely refrained from using chemical weapons even when they possessed them and even when using them might have been militarily convenient. The concept of a chemical weapons taboo, a normative prohibition that operates independently of enforcement, has been influential. Cross (2024), writing about the parallel case of biological weapons, offers a useful and sceptical examination of what a weapons taboo actually is and whether it does the explanatory work that is often claimed for it. Students should treat the taboo literature as a live argument rather than a settled conclusion. The third conversation is about #verification and institutional performance. Kelle (2023) provides the most useful single point of entry for students. He argues that the CWC regime has undergone a shift in its centre of gravity, moving from the verification of destruction, which was the dominant activity of the first twenty-five years, toward the #attribution of use, which he treats as a new and contested regime norm produced by the Syrian crisis. Ghionis and Kelle (2024) extend this analysis to the period after the Fifth Review Conference of 2023 and identify five areas where the regime needs work: upholding the norm against use, adapting verification, capacity building, organisational governance, and external engagement. The fourth conversation is about science, technology, and the future of the ban. This is the most rapidly moving body of work. Crowley and Dando (2022a) analyse the decision taken by states parties in December 2021 on the aerosolised use of central nervous system acting chemicals. Their book on toxins and bioregulators (2022b) argues that the convergence of chemistry and biology has created a class of agents that sit awkwardly between the CWC and the Biological and Toxin Weapons Convention. Their more recent monograph (2025) proposes what they call holistic arms control, which draws not only on disarmament treaties but on human rights law, humanitarian law, drug control conventions, medical ethics, and the codes of conduct of scientific bodies. Davison (2025) provides a detailed legal history of the same problem, tracing a thirty-year effort to close a loophole in the treaty text. Costanzi and Koblentz (2021) examine the technical and diplomatic problem of bringing a whole family of previously unlisted nerve agents under the Convention's declaration and verification rules. Urbina and colleagues (2022) demonstrated, in a widely discussed study, that machine learning models built for drug discovery could be inverted to generate toxic molecules, which raised the question of whether #dual_use control regimes built around physical materials can cope with threats that are informational. Noort and Hotchkiss (2025) describe the network of designated laboratories on which the OPCW depends for its analytical credibility. 2.3 The gap this article addresses Each of these conversations is valuable and each is, for a newcomer, partial. Legal scholarship tends to assume institutional knowledge. Institutional analysis tends to assume legal knowledge. The science and technology literature tends to assume both. The purpose of this article is integrative: to give a student a single, coherent account that connects the treaty text to the institution, the institution to its verification practice, and that practice to the political and scientific pressures now reshaping it. 3. Methodology This is a qualitative, doctrinal and analytical review. It does not generate new empirical data. It synthesises three categories of material. The first category is the primary legal text: the Convention itself, its three annexes, and the decisions taken by the #Conference_of_the_States_Parties and the #Executive_Council that have interpreted or supplemented the text. Decisions of the policy-making organs are treated as authoritative interpretations of the Convention where they were adopted under the procedures the treaty provides. The second category is the reporting of the OPCW Technical Secretariat, including the reports of its investigative bodies and the advice of its #Scientific_Advisory_Board. These are treated as evidence of the organisation's own findings and reasoning. Where those findings are politically contested, the article states that they are contested and identifies who contests them. The third category is peer-reviewed and expert scholarly literature, drawn principally from the period 2021 to 2026 and listed in the references. Three methodological limits should be stated at the outset. First, a great deal of the relevant information is classified by national governments and is unavailable to open-source researchers. Any assessment of covert programmes rests on inference. Second, the CWC is an intensely politicised subject, and the same facts are read very differently in The Hague, in Washington, and in Moscow. This article reports the findings of the OPCW's mandated bodies and reports the existence of disagreement, without treating disagreement as equivalent to disproof. Third, developments in this field move quickly and any account has a shelf life. 4. The Legal Architecture of the Convention 4.1 The core obligations Article I of the CWC contains the general obligations. Each state party undertakes never, under any circumstances, to develop, produce, otherwise acquire, stockpile, or retain #chemical_weapons, or to transfer them directly or indirectly to anyone. It undertakes never to use them. It undertakes never to engage in any military preparations to use them. It undertakes never to assist, encourage, or induce anyone else to engage in any activity prohibited to a state party. And it undertakes to destroy the chemical weapons it owns or possesses, to destroy any it has abandoned on the territory of another state, and to destroy its chemical weapons production facilities. Article I also contains a provision that students frequently miss. Paragraph 5 states that each state party undertakes not to use riot control agents as a method of warfare. This creates an asymmetry that runs through the whole treaty: tear gas may lawfully be used by police against a crowd in a domestic context, but may not lawfully be used by soldiers against an enemy in an armed conflict. The phrase "never under any circumstances" carries legal weight. It means the prohibition admits of no reprisal exception, no military necessity exception, and no derogation in emergency. This is unusually absolute for a treaty obligation and it is deliberate. 4.2 The General Purpose Criterion The intellectual heart of the Convention lies in Article II, and specifically in the way it defines a chemical weapon. The definition does not consist of a list of banned substances. It could not, because such a list would be obsolete the moment a chemist synthesised something new. Instead, the Convention defines a #toxic_chemical as any chemical which, through its chemical action on life processes, can cause death, temporary incapacitation, or permanent harm to humans or animals. This covers all such chemicals, regardless of their origin and regardless of how or where they were produced. It covers chemicals not yet discovered. The Convention then defines a chemical weapon as a toxic chemical or its precursor, together with the munitions and devices designed to deliver it, except where intended for purposes not prohibited under the Convention and as long as the types and quantities are consistent with such purposes. This inverted structure is called the #General_Purpose_Criterion. Its effect is to shift the legal question away from the identity of the molecule and toward the purpose and the quantity. Chlorine is a chemical weapon when it is loaded into a barrel and dropped from a helicopter onto a residential neighbourhood. The same chlorine is not a chemical weapon when it is used to purify municipal drinking water. The molecule is identical. The purpose is not. The purposes not prohibited are set out in Article II, paragraph 9. They are: industrial, agricultural, research, medical, pharmaceutical, or other peaceful purposes; protective purposes, meaning those directly related to protection against toxic chemicals and chemical weapons; military purposes not connected with the use of chemical weapons and not dependent on the toxic properties of chemicals as a method of warfare; and #law_enforcement, including domestic riot control purposes. Students should note the last of these four. The law enforcement exception has become the most contested clause in the entire Convention, and Section 9 returns to it at length. The General Purpose Criterion is what makes the CWC a durable instrument rather than a periodically obsolete one. Any new toxic chemical developed as a weapon is prohibited from the moment of its creation, whether or not the OPCW has ever heard of it. This is a considerable drafting achievement and it deserves more attention in teaching than it usually receives. 4.3 The Schedules If the General Purpose Criterion establishes what is prohibited, the Annex on Chemicals establishes what is monitored. This distinction confuses students more than any other feature of the treaty, so it is worth stating plainly. The Schedules are not a list of banned chemicals. They are a list of chemicals subject to declaration and inspection. A chemical can be entirely absent from every Schedule and still be a chemical weapon under Article II if it is used as one. The Annex divides listed chemicals into three Schedules according to risk. #Schedule_1 chemicals have few or no uses other than as chemical weapons, or as precursors to them. Sarin, VX, sulfur mustard, and the A-series agents commonly known by the name Novichok fall here. Production is capped at very small aggregate quantities for research, medical, pharmaceutical, or protective purposes. Facilities producing them are subject to intensive monitoring. Transfers to non-states-parties are forbidden. #Schedule_2 chemicals have some legitimate commercial use but pose a significant risk, often as precursors. Declaration and inspection requirements are substantial but less severe than for Schedule 1. #Schedule_3 chemicals are produced in large commercial volumes and have wide civilian application, but have been used as weapons in the past or could serve as precursors. Phosgene and hydrogen cyanide are examples. Reporting thresholds are high and inspection is comparatively light. Beyond the Schedules lies a fourth and much larger category: discrete organic chemicals, or DOCs. Plants producing these in quantities above a threshold must be declared, and a subset are inspectable. This category matters because the global chemical industry has grown enormously since 1993 and the majority of what it produces is unscheduled. The Annex on Chemicals may be amended without reopening the whole treaty, through a technical change procedure. This procedure was used for the first time in November 2019, when states parties agreed to add the A-series nerve agents and a related family of compounds to Schedule 1. The amendment took effect in June 2020. Costanzi and Koblentz (2021) analyse the diplomatic and technical logic of that amendment and argue for a family-based approach to scheduling, which covers structural classes of compounds rather than individually named molecules. The argument matters because naming individual molecules invites evasion by minor structural modification. 4.4 Declarations, destruction, and national implementation Article III requires each state party, within thirty days of the Convention entering into force for it, to declare whether it owns or possesses chemical weapons, whether it has any on its territory, whether it has abandoned any on the territory of another state, and whether it has or has had a chemical weapons production facility. It must also declare riot control agents in its possession. Articles IV and V govern the #destruction of stockpiles and of production facilities. Destruction is carried out under continuous international monitoring. Production facilities must be destroyed or, under strict conditions and with the approval of the Conference, converted to peaceful purposes. Article VI governs activities not prohibited, which in practice means the #chemical_industry. It requires each state party to adopt the measures necessary to ensure that toxic chemicals within its territory are developed, produced, and used only for purposes not prohibited, and it subjects declared industrial facilities to verification. Article VII requires each state party to enact #national_implementing_legislation. This obligation is easy to state and difficult to fulfil. It requires a state to criminalise the prohibited activities under its own law, to extend that criminal law to its nationals abroad, to designate a National Authority as the point of contact with the OPCW, and to establish export and import controls. Many states, particularly those with limited legislative and administrative capacity, have struggled with it. Weak national implementation is the quiet vulnerability of the regime, because a prohibition that exists in international law but not in domestic criminal law cannot easily be used to prosecute a non-state actor. Articles X and XI are the cooperative pillars of the treaty. Article X provides for #assistance_and_protection: a state party that is attacked or threatened with chemical weapons may request assistance, and the OPCW maintains capacity to deliver it. Article XI provides for #international_cooperation in the peaceful uses of chemistry and states that the Convention shall not be applied so as to hamper the economic or technological development of states parties. Developing countries have consistently argued that Article XI has been under-implemented relative to the resources poured into verification. This grievance is not merely rhetorical. It shapes voting behaviour and it explains why the geographic distribution of the OPCW's benefits is a recurring theme at every Review Conference. Article XII sets out the measures the Conference may take to redress a situation and ensure compliance, including restricting or suspending a state party's rights and privileges, recommending collective measures, and, in cases of particular gravity, bringing the issue to the attention of the United Nations General Assembly and Security Council. 5. The Organisation for the Prohibition of Chemical Weapons 5.1 Why a dedicated organisation matters The CWC did not entrust its implementation to an existing body. Article VIII established the #OPCW, based in The Hague, as an autonomous international organisation with its own staff, its own budget, and its own inspectorate. This was a significant institutional choice. The Biological and Toxin Weapons Convention of 1972, by contrast, has no verification organisation at all, and the difference in the two regimes' capacity to detect and respond to violations is stark. Students comparing the two conventions should treat institutional design as the primary explanatory variable. The OPCW was awarded the Nobel Peace Prize in 2013. 5.2 The three organs The Conference of the States Parties is the plenary body. All members sit in it, each with one vote. It meets annually and it is the ultimate decision-making authority: it oversees implementation, approves the budget, elects the Executive Council, appoints the Director-General, and may take the measures under Article XII described above. The Convention states that decisions on matters of substance should be taken by consensus as far as possible, but where consensus cannot be reached the Conference may decide by a two-thirds majority of members present and voting. This fallback provision was, for many years, largely theoretical. It is no longer. The Executive Council is the executive organ. It consists of forty-one states parties, elected by the Conference for two-year terms, with seats distributed among five regional groups. It supervises the work of the Technical Secretariat, considers concerns about compliance, and can recommend action to the Conference. Substantive decisions require a two-thirds majority of all its members. The Technical Secretariat, headed by the Director-General, does the actual work. It receives and analyses declarations, plans and conducts inspections, maintains the inspectorate, and provides technical support to states parties. Its inspectors are the only international officials with a treaty-based right to enter the chemical facilities of sovereign states. Two subsidiary structures deserve mention. The Scientific Advisory Board provides the Director-General with specialised advice on developments in science and technology relevant to the Convention. Its influence is considerable and often underestimated: the SAB's assessments have driven the debate on incapacitating chemicals, on the scheduling of new agents, and most recently on artificial intelligence. The Centre for Chemistry and Technology, opened in 2023, gives the organisation laboratory and training capacity it previously lacked. The Secretariat's analytical credibility rests on a network of designated laboratories in member states, which are accredited only after passing demanding proficiency tests. Noort and Hotchkiss (2025) describe this network and note the confidence it provides that any misuse of chemicals as weapons can be properly investigated. A student who wants to understand why OPCW findings carry weight should look at this laboratory network rather than at the political statements that surround the findings. 6. The Verification System in Practice The Convention's verification regime has four components. They developed at different times and they are in very different states of health. 6.1 Verification of destruction This was the dominant activity of the OPCW's first twenty-five years and it consumed the largest share of its resources. Inspectors maintained continuous or near-continuous presence at destruction facilities, verified the quantity and identity of agent destroyed, and confirmed the irreversibility of the process. The states that declared chemical weapons stockpiles to the OPCW were Albania, India, Iraq, Libya, the Republic of Korea, the Russian Federation, the Syrian Arab Republic, and the United States. All declared Category 1 stockpiles have now been destroyed. The final destruction event, at the Blue Grass Chemical Agent-Destruction Pilot Plant in Kentucky on 7 July 2023, eliminated the last munition in the original United States stockpile of more than thirty thousand tonnes of chemical agent. This is the achievement on which the regime's reputation rests, and it should be stated without hedging. #Destruction_verification worked. It was slow, it was expensive, deadlines were extended more than once, and the process generated its own diplomatic friction. But it was completed, and it was completed under international supervision that all parties accepted. 6.2 Industry verification The second component is routine inspection of declared #chemical_industry facilities under Article VI. Its purpose is not to catch weapons but to build confidence that plants capable of producing scheduled chemicals are not doing so for prohibited purposes, and to deter diversion. The system faces a structural problem. The chemical industry of 2026 is vastly larger, more globally distributed, and more technologically sophisticated than the industry of 1993. The number of declarable facilities has grown while the OPCW's inspection budget has not grown proportionately. This means the probability that any given facility is inspected in any given year is low. The regime therefore relies substantially on the deterrent effect of the possibility of inspection, on the accuracy of national declarations, and on the willingness of states to police their own industries. There is a second and deeper problem. The verification effort is concentrated on Scheduled chemicals, but a state seeking to reconstitute a weapons capability today would be unlikely to route it through Schedule 1 or Schedule 2 chemicals, precisely because those are the ones that are watched. Ghionis and Kelle (2024) and the wider expert community have argued for greater attention to unscheduled discrete organic chemicals, and for verification approaches that focus on capability and behaviour rather than on molecular identity. This is a difficult argument to win, because expanding industry verification imposes costs on the chemical sector and on the states that host it. 6.3 The challenge inspection Article IX contains the most intrusive mechanism in the treaty. Any state party may request a #challenge_inspection of any facility or location on the territory of any other state party, for the purpose of clarifying and resolving questions of possible non-compliance. The inspected state may not refuse. The Executive Council can block the inspection only by a three-quarters majority vote taken within twelve hours, a deliberately high bar designed to prevent the mechanism from being neutralised by political protection. The Convention builds in a compromise called managed access, which allows the inspected state to take steps to protect sensitive equipment, information, and areas unrelated to the Convention, while still satisfying the inspection team that the compliance concern has been addressed. In almost three decades, the challenge inspection has never been used. Not once. Why not? Several explanations circulate in the literature. States may fear that requesting an inspection would invite a retaliatory request. States may fear the diplomatic cost of an accusation that turns out to be unfounded. States may prefer to raise concerns through the softer consultation and clarification procedure of Article IX, paragraph 2, which is what the United Kingdom and forty-four other states did in the Navalny case, and which Russia rejected. And states may calculate that a challenge inspection which the Executive Council fails to support would demonstrate weakness rather than strength. Whatever the reason, the result is that the treaty's sharpest instrument has never been tested. Students should sit with the implications of that. A mechanism that has never been used has no track record, and a mechanism with no track record has uncertain deterrent value. Some scholars argue that its very existence deters; others argue that thirty years of non-use has quietly signalled that it will never be invoked. The argument is unresolved and it is a good subject for a dissertation. 6.4 Investigation of alleged use, and the turn to attribution The fourth component was originally modest. Article IX also provides for investigations of alleged use, and a separate mechanism exists under the authority of the United Nations Secretary-General. Neither was designed to name perpetrators. The Syrian crisis changed this. The OPCW established a Fact-Finding Mission in 2014 with a mandate to determine whether chemical weapons had been used, but explicitly not to say by whom. A joint OPCW-United Nations Joint Investigative Mechanism operated from 2015 to 2017 with an attribution mandate, and it did attribute several attacks, but its mandate was allowed to lapse when its renewal was vetoed in the Security Council. In June 2018, at a special session of the Conference of the States Parties, states parties adopted a decision entitled "Addressing the Threat from Chemical Weapons Use" which created the Investigation and Identification Team (#IIT). The IIT's mandate is to identify the perpetrators of chemical weapons use in Syria in cases where the Fact-Finding Mission has determined that use or likely use occurred. The IIT is not a court. It does not determine individual criminal responsibility. It reports its findings to states parties and to the Security Council. Kelle (2023) treats this development as a critical juncture in the evolution of the regime, arguing that it created what he calls a new #attribution norm within the CWC. That is a significant claim and it deserves scrutiny. The decision creating the IIT was adopted by vote rather than by consensus, over the objection of Russia, Syria, Iran, and others, who argued that the Conference had exceeded its powers under the Convention and had usurped the role of the Security Council. Supporters argued that the Conference was acting squarely within Article VIII, which empowers it to oversee implementation and to promote the object and purpose of the Convention, and that a Security Council paralysed by veto could not be allowed to paralyse the OPCW as well. Students should be able to argue both sides of this. The legal question is whether a plenary organ may, by majority vote, assign itself a function the treaty does not expressly provide for, when that function is arguably implicit in its supervisory mandate. Reasonable international lawyers disagree. 7. Achievements: What the Regime Has Delivered It is easy, in a field dominated by crises, to lose sight of what the CWC has accomplished. Four achievements stand out. Near-universal membership. One hundred and ninety-three states are party to the Convention. Only four remain outside it: Egypt, Israel, which signed but has not ratified, the Democratic People's Republic of Korea, and South Sudan. This means the ban covers roughly ninety-eight per cent of the world's population and roughly the same proportion of global chemical industry capacity. #Universality is not complete, and the four outliers include states in a region of acute tension, but the CWC is closer to universal than almost any other security treaty. Universality also matters for a reason students often miss. Unlike the Treaty on the Non-Proliferation of Nuclear Weapons, the CWC does not divide the world into legitimate possessors and prohibited non-possessors. Every state party has the same obligation. There is no chemical weapons equivalent of the recognised nuclear weapon state. This non-discriminatory structure is one reason the treaty attracted such wide adherence, and it is a point worth making in any comparative essay on #weapons_of_mass_destruction regimes. Verified destruction of all declared stockpiles. As noted, this was completed in July 2023. It is worth restating that this is unprecedented. An entire category of weapons of mass destruction, held by states in enormous quantity, was eliminated through an international legal process and verified by international inspectors on the territory of the possessor states. A functioning verification bureaucracy. The OPCW has conducted thousands of inspections in states across every region. It maintains a designated laboratory network capable of forensic analysis to a standard that has survived legal and scientific scrutiny. It has deployed teams into active conflict zones. These are hard operational tasks and the organisation performs them competently. Reinforcement of the norm. Whatever one thinks of the taboo literature, the practical fact is that chemical weapons use is now treated internationally as a scandal that demands a response, rather than as an unfortunate but ordinary feature of war. Sanctions have followed use. Investigations have followed use. Diplomatic isolation has followed use. This was not always so. 8. Compliance Crises 8.1 The Syrian file Syria acceded to the Convention in 2013 in the aftermath of the sarin attack on Ghouta, under intense international pressure and pursuant to a framework endorsed by United Nations Security Council Resolution 2118. The declared portion of its arsenal was removed and destroyed, largely at sea and in facilities abroad, in an operation that was itself a considerable logistical achievement. The problem is what came afterward. The OPCW Technical Secretariat has consistently reported that Syria's initial declaration could not be considered accurate and complete, because of gaps, inconsistencies, and discrepancies identified by the Declaration Assessment Team. And the Fact-Finding Mission repeatedly determined that chemical weapons, including chlorine and sarin, continued to be used in Syria after accession. The Investigation and Identification Team has now issued five reports. The first, in 2020, concerned attacks at Ltamenah in March 2017. The second, in 2021, concerned Saraqib in February 2018. The third, in 2023, concerned Douma in April 2018. The fourth, in February 2024, concluded that there were reasonable grounds to believe that the Islamic State in Iraq and the Levant had used sulfur mustard at Marea in 2015, which was the first IIT finding against a #non_state_actor. The fifth, released in January 2026, concluded that there were reasonable grounds to believe that the Syrian Arab Air Force was responsible for an attack at Kafr Zeita on 1 October 2016. In April 2021 the Conference of the States Parties suspended certain of Syria's rights and privileges under Article XII, including its right to vote. This was the first time such a measure had been imposed. The fall of the Assad government in December 2024 opened a new phase. The OPCW Director-General visited Damascus in February 2025 and the new authorities affirmed recognition of the organisation's mandates, including the identification of perpetrators. In June 2025 the Secretariat consolidated its Syrian work under a new Office of Special Missions, bringing the mandates of the Declaration Assessment Team, the Fact-Finding Mission, and the IIT under one umbrella. Since November 2025 the OPCW has maintained a continuous presence in the country. Deployments have identified previously undeclared chemical weapons and a substantial number of sites suspected of links to the former programme, well in excess of the number Syria had originally declared. Two lessons follow. The first is that a verification system depends fundamentally on the declaration it is given. If the initial declaration is false, inspectors are auditing a fiction. The second is more hopeful: political change can unlock a file that seemed permanently closed. The OPCW's institutional persistence, its refusal to close the Syrian dossier through a decade of obstruction, is what made it possible to act when the political situation shifted. 8.2 Nerve agents used for assassination A second category of violation involves the use of nerve agents against individuals rather than in warfare. In February 2017, VX was used to kill a man in a Malaysian airport terminal. In March 2018, an agent of the A-series family, widely referred to as Novichok, was used in Salisbury in the United Kingdom against Sergei and Yulia Skripal; a British civilian later died after handling the discarded container. In August 2020, a chemical of the same family was used against the Russian opposition politician Alexei Navalny. The OPCW confirmed the analytical findings in the Salisbury case through a technical assistance visit and confirmed the presence of a cholinesterase inhibitor with structural characteristics similar to Schedule 1 chemicals in the Navalny case. These incidents produced two responses. Diplomatically, they produced the schedule amendment of 2019, which brought the A-series agents formally under declaration and verification requirements for the first time. Costanzi and Koblentz (2021) argued that even this amendment leaves gaps, because listing named structural families still permits evasion by modification, and they proposed a broader family-based approach. Politically, they produced deadlock. The United Kingdom and forty-four other states parties invoked the consultation and clarification procedure of Article IX in the Navalny case, which is the softest of the treaty's compliance tools. Russia rejected the request and responded with counter-accusations. No challenge inspection was requested. The episode illustrates the gap between the enforcement powers the treaty grants on paper and the political conditions required to use them. 8.3 Terrorism and non-state actors The Convention was written for states. Its declaration, inspection, and destruction machinery presupposes a government, a territory, and a chemical industry. It is poorly configured for a cult in a Tokyo subway, an insurgent group in Iraq, or a lone individual with an internet connection. The Aum Shinrikyo sarin attack in 1995, before the Convention entered into force, remains the reference case. The IIT's 2024 finding against the Islamic State demonstrated that a non-state armed group in the contemporary period both produced and used a chemical warfare agent in the field. The regime's answer to this problem is indirect: Article VII national implementation, national export controls, and, outside the Convention, United Nations Security Council Resolution 1540. The answer is only as good as the weakest national legal system. This is why the unglamorous work of helping states draft implementing legislation is more important to counter-#terrorism than the more visible work of inspection. 8.4 Allegations under examination At the time of writing, allegations have been reported concerning the use of chemical agents in the conflict in Sudan. The OPCW has acknowledged receipt of information and indicated it would examine it, and the state concerned has denied the allegations. Students should be careful here. An allegation reported in the press, however credible the source, is not a finding of the OPCW's mandated investigative bodies. The discipline of distinguishing an allegation from a determination is central to serious work in this field, and it is a discipline that is frequently abandoned in public debate. 9. Emerging Challenges 9.1 The law enforcement exception and central nervous system acting chemicals This is the most legally interesting problem in contemporary CWC practice and students should understand it properly. Recall the structure of Article II. Toxic chemicals are lawful when intended for purposes not prohibited, and one of those purposes is law enforcement, including domestic riot control. Recall also that Article II, paragraph 7, defines a #riot_control_agent as a chemical, not listed in a Schedule, which produces rapidly in humans sensory irritation or disabling physical effects that disappear within a short time following the end of exposure. Here is the ambiguity. The law enforcement exception in paragraph 9 is not expressly limited to riot control agents. So the question arises: may a state use a toxic chemical that is not a riot control agent, something far more potent, for a law enforcement purpose? The stakes became concrete in October 2002, when Russian forces used an aerosolised fentanyl derivative to end a hostage siege at a Moscow theatre. The armed hostage-takers were incapacitated. So were the hostages. More than one hundred and twenty of them died. Riches and colleagues later identified carfentanil and remifentanil in samples from casualties. The problem with this class of chemicals, as the OPCW's Scientific Advisory Board has repeatedly explained, is dosage. A pharmaceutical that is safe under controlled medical conditions, with a known patient weight and a trained anaesthetist present, has a very low safety margin when it is sprayed into a room containing hundreds of people of different ages, body weights, and states of health, who cannot be individually monitored and who cannot leave. When such a chemical is disseminated in airborne form, the delivered dose cannot be controlled. The SAB has assessed that certain #CNS_acting_chemicals are extremely toxic, with some as lethal as nerve agents. They are also, as the SAB has noted, not riot control agents within the meaning of the treaty, because they act primarily on the central nervous system and their effects are not confined to temporary sensory irritation. After roughly two decades of argument, the Conference of the States Parties adopted a decision on 1 December 2021 stating its understanding that the aerosolised use of CNS-acting chemicals is inconsistent with law enforcement as a purpose not prohibited under the Convention. Crowley and Dando (2022a) described this as a substantial advance. It was adopted, however, by vote rather than consensus: eighty-five in favour, ten against, including China, Iran, and Russia, and thirty-three abstentions. Davison (2025), who traces the thirty-year history of this effort in detail, asks the right question. The loophole has shrunk. Has it closed? Probably not entirely, and students should be able to explain why. Four gaps remain. The decision does not define what counts as a CNS-acting chemical, nor provide an indicative list. It addresses only chemicals acting on the central nervous system, so a future weapon employing a toxin or bioregulator acting on some other physiological pathway would not be covered. It addresses only aerosolised use, which on a strict reading excludes gases and vapours as well as other delivery routes such as dart guns. And it restricts itself to munitions and devices specifically designed to cause harm, which may leave general-purpose delivery equipment outside its scope. Crowley and Dando (2025) argue that the answer to these residual gaps cannot come from the CWC alone. Their proposed framework of holistic arms control draws simultaneously on the Convention, on the Biological and Toxin Weapons Convention, on human rights law, on humanitarian law, on the United Nations drug control conventions, on medical ethics, and on the professional codes of scientific bodies. Whether or not one accepts their prescription, their diagnosis is hard to dispute: a governance gap that sits between two treaties will not be closed by either treaty acting alone. 9.2 Riot control agents and human rights A related problem concerns riot control agents themselves. Their use in law enforcement is lawful under the Convention, subject to types and quantities consistent with that purpose. But the same chemicals are frequently used in ways that raise serious #human_rights concerns: in confined spaces such as prisons and detention centres where those exposed cannot disperse, in excessive quantities against peaceful assemblies, and increasingly through wide-area delivery systems including drones and remote-operated launchers. The OPCW is not a human rights body and has generally been reluctant to enter this territory. But the Convention's own language, requiring types and quantities consistent with law enforcement purposes, arguably imports a proportionality standard that is not self-evidently satisfied when tear gas is dropped from an aircraft onto a crowd. Edwards and colleagues (2022) call for the OPCW to establish a process to determine which delivery mechanisms are prohibited under the Convention and to develop guidance on the appropriate use of the permitted ones. This is a plausible reform and it has not happened. 9.3 Convergence, artificial intelligence, and the informational threat The deepest long-term challenge to the CWC is that the science it must govern is no longer the science it was designed to govern. Chemistry and biology have converged. Toxins and bioregulators, which are produced by living organisms but which act as poisons, fall within the scope of both the CWC and the Biological and Toxin Weapons Convention, and are governed adequately by neither. Crowley and Dando (2022b) develop this argument at length. The practical risk is that a governance gap opens precisely where the science is moving fastest. More recently, #artificial_intelligence has entered the picture. Urbina and colleagues (2022) reported a striking result. A machine learning model designed to generate molecules with therapeutic properties and to penalise toxicity could be inverted, by reversing the sign of the toxicity term, so that it generated toxic molecules instead. Within hours it had produced tens of thousands of candidate structures, including known nerve agents and many novel compounds predicted to be more toxic still. The authors did not synthesise anything, and predicted toxicity is not demonstrated toxicity, and the very substantial practical barriers to actually producing and weaponising such a compound remain. But the paper made a conceptual point that cannot be unmade: the design of toxic molecules is becoming an informational activity, and a regime built on controlling physical materials and physical facilities is not obviously equipped to control information. The OPCW has responded. A Global Conference on the role of artificial intelligence in advancing implementation of the Convention was co-hosted with Morocco in Rabat in October 2024. The Scientific Advisory Board established a Temporary Working Group on Artificial Intelligence which began its mandate on 1 January 2025 and which delivered its report in March 2026. The Executive Council added a standing agenda item on emerging technologies. These are sensible institutional responses and they are, at this stage, exploratory rather than regulatory. There is a second side to this. AI is not only a threat vector. It offers the Secretariat tools for analysing declarations, detecting anomalies in industry data, supporting chemical forensics, and improving the efficiency of an inspectorate whose resources have not kept pace with the industry it monitors. Whether the technology is on balance a threat or an asset to the regime is an open question, and it is one of the most promising areas for new student research. 9.4 The politics of consensus The final challenge is not scientific. It is political. The OPCW was designed to operate by consensus. For most of its history it did. Since 2018, its most consequential decisions have been taken by contested vote: the creation of the IIT, the suspension of Syria's rights, the decision on CNS-acting chemicals. The Fifth Review Conference of May 2023 was unable to adopt a consensus final document, as the Fourth had also failed to do. This matters for reasons beyond diplomatic aesthetics. An organisation whose decisions are taken over the objection of a substantial minority of its members faces a legitimacy problem, and a legitimacy problem eventually becomes an implementation problem. States that voted against a decision may implement it grudgingly, or not at all. States in the middle, particularly in the developing world, may come to see the organisation as an instrument of one bloc rather than as a common institution. Ghionis and Kelle (2024) identify organisational governance and external engagement as priority areas for exactly this reason. The counter-argument is equally serious. If consensus is treated as a requirement rather than an aspiration, then any single state, including a state under investigation, can veto action against itself. That is not a functioning compliance regime; it is a compliance regime with an off switch operated by the accused. The states that pushed the IIT decision to a vote made precisely this argument. There is no clean resolution. The CWC is caught between two goods, inclusiveness and effectiveness, and it cannot fully have both. Recognising this trade-off honestly is more useful than pretending it away, and it is a better essay question than most of what appears on undergraduate examination papers in this field. 10. Discussion Three observations emerge from the foregoing. The regime has changed jobs. The CWC was written to eliminate existing arsenals and it did that. The task now is prevention of re-emergence, which is a fundamentally different problem requiring different tools. Destruction verification was a counting exercise with a defined endpoint. Prevention of re-emergence is an open-ended intelligence and monitoring exercise against an adversary who is trying not to be found, in an industry that is growing faster than the inspectorate. Kelle's (2023) framing of a shift from destruction to attribution captures part of this, but the shift is broader than attribution alone. It extends to industry monitoring, to science and technology review, and to the governance of research that no inspector will ever see. The treaty's greatest strength and its greatest weakness are the same clause. The General Purpose Criterion makes the prohibition future-proof: anything used as a chemical weapon is a chemical weapon. But the criterion turns on purpose, and purpose is invisible. You cannot inspect an intention. The consequence is that the Convention's substantive scope is enormous while its verification scope is narrow, and the gap between the two is where every contemporary controversy lives. The law enforcement exception, the CNS-acting chemicals debate, the unscheduled discrete organic chemicals problem, and the AI question are all, structurally, the same problem in different clothing. Institutional resilience is real but not unlimited. The OPCW has absorbed a decade of assault on its findings and its impartiality and has continued to function. It kept the Syrian file open for eleven years and was in a position to act within weeks when the political situation changed. That is a considerable institutional achievement and it should be credited. But an organisation cannot indefinitely take contested decisions by majority vote against a determined minority without cost. The regime's political capital is finite and it is being spent. 11. Policy Recommendations The following recommendations are drawn from the literature reviewed and are offered as arguable propositions rather than settled conclusions. Students are encouraged to contest them. Complete the Syrian file. The current period of cooperation with the new Syrian authorities is an opportunity that may not recur. Resources should be concentrated on establishing the full scope of the former programme, verifying the destruction of remaining stocks, and preserving evidence for future accountability processes. Define the terms of the 2021 decision. The Conference should task the Secretariat, with the assistance of the Scientific Advisory Board, to develop implementation guidance defining CNS-acting chemicals, providing an indicative list of chemicals of concern, and clearly demarcating the boundary between such chemicals and riot control agents. Extend the logic of the decision beyond aerosols and beyond the central nervous system. The gaps identified by Davison (2025) and by Crowley and Dando (2025) are foreseeable and should be closed before, rather than after, a weapon exploiting them is deployed. The precedent of the Protocol on Blinding Laser Weapons, which prohibited a weapon before its deployment, shows that pre-emptive prohibition is politically possible. Adopt a family-based approach to scheduling. Following Costanzi and Koblentz (2021), the Annex on Chemicals should cover structural families rather than enumerated compounds, in order to reduce the ease of evasion by minor modification. Rebalance verification toward unscheduled chemicals and toward behaviour. Inspection effort concentrated on the chemicals a violator would be least likely to use is inefficient. This requires additional resources and it requires the chemical industry to accept a heavier burden, which will be resisted. Institutionalise the science and technology review. The Temporary Working Group on Artificial Intelligence should not remain temporary. Reviewing convergent science is now a permanent function and should be resourced as one. Take Article XI seriously. Capacity building and international cooperation in the peaceful uses of chemistry are not a side payment to the developing world. They are the mechanism by which the regime retains the political support of the majority of its members, and that support is what allows difficult decisions to be taken at all. Pursue universality. Four states remain outside. Their accession would strengthen the regional security architecture more than it would weaken it, and the argument for accession should be made in those terms rather than in the language of pressure. Strengthen national implementation. Article VII compliance is the single most cost-effective investment available to the regime, because it is the only mechanism that reaches non-state actors, and it is chronically neglected. Invest in #disarmament_education. A treaty regime survives across generations only if each generation understands why it exists. The recent OPCW initiative to build a global academic network in support of the Convention is a step in this direction, and it deserves the support of universities. 12. Limitations Three limits should be acknowledged. First, this article relies on open sources; covert programmes are by definition not in the open literature, and any assessment of what states are secretly doing is inference rather than knowledge. Second, the article treats the findings of the OPCW's mandated bodies as authoritative while noting that some states dispute them; a reader who rejects that starting assumption will read the Syrian and Novichok sections very differently, and should. Third, the field is moving. The Syrian file in particular is in active flux, and any statement about it made in mid-2026 should be checked against subsequent OPCW reporting. 13. Conclusion The #Chemical_Weapons_Convention is the most successful disarmament treaty in history and it is under more strain than at any point since it entered into force. Both statements are true and neither cancels the other. The Convention did what it set out to do. Every declared stockpile in the world has been destroyed, verified by international inspectors, on the territory of the states that held them. One hundred and ninety-three states have accepted a legal obligation never, under any circumstances, to use these weapons, and have accepted a permanent international organisation with the right to come and check. That is not a small thing, and students who come to this subject through the headlines about Syria and Salisbury sometimes fail to see it. But the Convention was built for a world of large military stockpiles held by identifiable states, and that world is gone. The threat now is smaller, more distributed, and harder to see: a nerve agent in a perfume bottle, chlorine in an improvised barrel, a fentanyl derivative sprayed through a ventilation duct, a toxic molecule designed by software on an ordinary computer. Against these, a verification regime built around counting munitions and inspecting declared plants is a partial answer at best. What the regime has, and what it must protect, is the norm. The reason chemical weapons use produces scandal rather than shrugs is that ninety-eight per cent of humanity lives under a treaty that says these weapons are never, under any circumstances, acceptable. That norm is not self-executing. It is sustained by investigations that are actually carried out, by findings that are actually published, by perpetrators who are actually named, and by an institution that refuses to close a file simply because closing it would be more comfortable. The next twenty-five years of the Convention will be decided by whether its members are willing to pay the political costs of sustaining that norm. On the evidence of the past decade, the answer is uncertain. On the evidence of July 2023, it is not hopeless. Hashtags #Chemical_Weapons_Convention #CWC_1993 #OPCW #Chemical_Disarmament #Weapons_of_Mass_Destruction #Arms_Control #International_Law #Verification_Regime #General_Purpose_Criterion #Non_Proliferation #Chemical_Weapons_Ban #Global_Security #Peace_Studies #Disarmament_Education #International_Relations References Costanzi, S., and Koblentz, G. D. (2021). Strengthening controls on Novichoks: a family-based approach to covering A-series agents and precursors under the chemical-weapons nonproliferation regime. The Nonproliferation Review, 28(1-3), 95-113. https://doi.org/10.1080/10736700.2021.1975879 Cross, G. A. (2024). Is there a biological weapons taboo, and does it matter? The Nonproliferation Review, 31(1-3). https://doi.org/10.1080/10736700.2024.2415256 Crowley, M., and Dando, M. (2022a). Central nervous system weapons dealt a blow. Science, 375(6577), 153-154. https://doi.org/10.1126/science.abn5621 Crowley, M., and Dando, M. (2022b). Toxin and Bioregulator Weapons: Preventing the Misuse of the Chemical and Life Sciences. Palgrave Macmillan, Cham. Crowley, M., and Dando, M. (2025). Preventing Weaponization of CNS-acting Chemicals: A Holistic Arms Control Analysis. Royal Society of Chemistry, Croydon. Davison, N. (2025). Nerve agents by another name: the thirty-year effort to close a loophole on chemical weapons. In H. Krieger, P. Kalmanovitz, E. Lieblich, and S. Evdokimos Pantazopoulos (Eds.), Yearbook of International Humanitarian Law, Volume 26 (2023) (pp. 77-108). T.M.C. Asser Press, The Hague. https://doi.org/10.1007/978-94-6265-663-5_3 Edwards, B., Novossiolova, T., Crowley, M., Whitby, S., Dando, M., and Shang, L. (2022). Meeting the challenges of chemical and biological weapons: strengthening the chemical and biological disarmament and non-proliferation regimes. Frontiers in Political Science, 4, 805426. https://doi.org/10.3389/fpos.2022.805426 Ghionis, A., and Kelle, A. (2024). The Chemical Weapons Convention After its Fifth Review Conference: Key Issues for the European Union. EU Non-Proliferation and Disarmament Papers No. 87. Stockholm International Peace Research Institute, Stockholm. Ghionis, A., Kelle, A., and Garzon Maceda, M. (2023). Preparing for Success at the Fifth Review Conference of the Chemical Weapons Convention: A Guide to the Issues. United Nations Institute for Disarmament Research, Geneva. https://doi.org/10.37559/WMD/23/CWC/01 Kelle, A. (2023). The CWC at 25: from verification of chemical-weapons destruction to attribution of their use. The Nonproliferation Review, 28(4-6). https://doi.org/10.1080/10736700.2023.2180234 Lentzos, F., and Jakob, U. (2023). Chemical, biological and health security threats. In SIPRI Yearbook 2023: Armaments, Disarmament and International Security. Oxford University Press, Oxford. Meyer, P. (2021). Does the Conference on Disarmament have a future? Journal for Peace and Nuclear Disarmament, 4(2), 287-294. Myjer, E. P. J., and Marauhn, T. (Eds.) (2022). Research Handbook on International Arms Control Law. Edward Elgar Publishing, Cheltenham. https://doi.org/10.4337/9781788111904 Noort, D., and Hotchkiss, P. J. (2025). Scientific collaboration to rid the world of chemical weapons. Nature Reviews Chemistry, 9, 728-729. https://doi.org/10.1038/s41570-025-00762-y Revill, J., Borrie, J., and Lennane, R. (2022). Back to the Future for Verification in the Biological Disarmament Regime? WMD Compliance and Enforcement Series No. 14. United Nations Institute for Disarmament Research, Geneva. Urbina, F., Lentzos, F., Invernizzi, C., and Ekins, S. (2022). Dual use of artificial-intelligence-powered drug discovery. Nature Machine Intelligence, 4(3), 189-191. https://doi.org/10.1038/s42256-022-00465-9

The Treaty on the Non-Proliferation of Nuclear Weapons, opened for signature in 1968 and in force since 1970, remains the central legal instrument that seeks to stop the spread of nuclear weapons and the technology that makes them possible. This article offers students and early-career researchers a structured, evidence-based review of the treaty: how it came about, what it actually says, how it is verified, how it has performed, and why it is under so much pressure today. The study uses a qualitative, doctrinal and historical-institutional approach. It reads the treaty text alongside the record of eleven review conferences held between 1975 and 2026, the safeguards practice of the International Atomic Energy Agency, and the recent scholarly literature on the global nuclear order. Three findings stand out. First, the treaty's core exchange, often called the grand bargain, is legally uneven: the obligations placed on states without nuclear weapons are precise and verifiable, while the disarmament obligation placed on states with nuclear weapons is open-ended and hard to enforce. Second, the treaty's practical success in limiting membership of the nuclear club has been real but partial, and it depends heavily on institutions built around the treaty rather than on the treaty text itself. Third, the period from 2022 to 2026 has been unusually damaging: a major war involving a nuclear-armed state, the expiry of the last United States and Russia arms control treaty in February 2026, military strikes on safeguarded nuclear facilities in Iran, a near-total loss of inspector access in that country, and a third consecutive review conference that closed without an agreed final document in May 2026. The article argues that the treaty is neither dead nor healthy. It is a legal framework whose rules still hold, but whose political consensus has thinned to the point where its ability to manage the next proliferation crisis is genuinely in doubt. The conclusion sets out reform options and a research agenda for students. Keywords: nuclear non-proliferation; NPT; International Atomic Energy Agency; safeguards; nuclear disarmament; arms control; global nuclear order; nuclear energy governance 1. Introduction Few international agreements have been asked to carry as much weight as the #NPT. Signed in 1968 and entering into force on 5 March 1970, it is the legal backbone of what scholars call the #global_nuclear_order. Its purpose is stated plainly in its own preamble and articles: to prevent the further spread of #nuclear_weapons, to promote #nuclear_disarmament, and to protect the right of states to use nuclear energy for peaceful purposes. With 191 states parties, it is one of the most widely joined security treaties in history. Only India, Israel, Pakistan and South Sudan have never joined, and North Korea announced its withdrawal in 2003. For students approaching this subject for the first time, the treaty is deceptively simple. It is short. It contains eleven articles. Most of its operative language can be read in twenty minutes. Yet almost every serious dispute in nuclear politics over the last fifty years can be traced back to a disagreement about what those eleven articles mean, who is required to do what, and what happens when a party breaks the rules. This article has four aims. The first is descriptive: to explain, in clear language, what the treaty says and how the machinery around it works. The second is historical: to trace the negotiation of the treaty and the eleven review conferences that have followed, because the treaty is best understood as a living political process rather than a fixed text. The third is analytical: to assess the treaty's record against the fears of the 1960s and the promises made in 1995, and to place that assessment inside the main theoretical debates in the field. The fourth is forward-looking: to describe the state of the regime in 2026, after a series of shocks that have left it visibly weakened, and to set out the reform proposals now on the table. The timing matters. In February 2026, the New Strategic Arms Reduction Treaty expired without a successor, leaving the two largest nuclear arsenals without any legally binding numerical limit for the first time since the early 1970s. In May 2026, the eleventh #review_conference of the treaty ended in New York without a consensus outcome document, the third consecutive failure. Meanwhile, the International Atomic Energy Agency has reported that it lost continuity of knowledge over Iran's declared stocks of #enriched_uranium after Iranian cooperation was suspended in July 2025 in the aftermath of military strikes on its nuclear facilities. Any honest account of the treaty written in 2026 has to hold two things together: the treaty's genuine, measurable achievements, and the seriousness of its present crisis. 1.1 Research questions The article is organised around four questions. What legal obligations does the treaty actually create, and how are they distributed between the two classes of states it recognises? How effective has the verification and enforcement machinery been in detecting and reversing violations? What explains the treaty's durability despite recurring political deadlock in its review process? What are the most credible reform paths available to states parties after the failure of the 2026 review conference? 1.2 Method and scope The method is qualitative and interpretive. It combines doctrinal reading of the treaty and its associated safeguards instruments with a historical-institutional analysis of the review process, drawing on the published record of review conference outcome documents, agency reporting, and recent peer-reviewed scholarship. The article does not attempt statistical testing of proliferation determinants, a large and separate literature. Nor does it take a normative position on whether nuclear deterrence is defensible. It aims instead to give students an accurate map of the terrain, including the places where scholars disagree sharply. A word on limits. Nuclear politics is a field where classified information matters and where states have strong incentives to misrepresent their capabilities and intentions. Any assessment rests on open sources, and open sources are incomplete. Readers should treat confident claims about hidden programmes, in either direction, with caution. 2. Origins: How the Treaty Was Built 2.1 The problem the treaty was designed to solve The atomic bombings of Hiroshima and Nagasaki in August 1945 created an immediate governance problem. The technology that produced the bombs was rooted in physics that could not be kept secret indefinitely, and in industrial processes that were expensive but not impossible for a determined mid-sized state to reproduce. Within two decades, five states had tested nuclear devices: the United States in 1945, the Soviet Union in 1949, the United Kingdom in 1952, France in 1960, and China in 1964. The early attempts at control failed. The Baruch Plan of 1946, which proposed international ownership of dangerous nuclear activities, collapsed in Cold War distrust. In its place came a very different logic, set out by President Dwight Eisenhower in his 1953 Atoms for Peace speech: rather than deny nuclear technology to the world, the nuclear powers would supply it for civilian purposes, in exchange for controls that would keep it out of weapons programmes. This bargain, technology in return for oversight, is the ancestor of the treaty's third pillar and of the International Atomic Energy Agency, established in 1957. By the early 1960s the fear of rapid spread was widespread. President John F. Kennedy famously worried in 1963 that fifteen, twenty or twenty-five states might hold nuclear weapons by the 1970s. That forecast, whatever its analytical basis, captured the mood. The Cuban Missile Crisis of 1962 had made the dangers of nuclear confrontation vivid. The Chinese test of 1964 raised the prospect of a chain reaction in Asia. West Germany's status, and the question of whether NATO would create a multilateral nuclear force, alarmed Moscow. The United Nations General Assembly responded in 1961 with the Irish Resolution, which called for an agreement under which nuclear states would not transfer weapons and non-nuclear states would not acquire them. That resolution is the direct ancestor of the treaty. 2.2 Negotiating the bargain, 1965 to 1968 Negotiations took place mainly in the Eighteen Nation Committee on Disarmament in Geneva. The crucial breakthrough came when Washington and Moscow, despite deep hostility elsewhere, discovered a shared interest in freezing the membership of the nuclear club. Both preferred a world in which their rivals' allies did not acquire independent arsenals. The superpower convergence of 1966 and 1967 made a treaty possible. But the non-nuclear majority, organised loosely through the #Non_Aligned_Movement and led diplomatically by states such as India, Sweden, Brazil, Mexico, Nigeria and Egypt, refused to sign a purely restrictive document. They made three demands. They wanted a commitment to #nuclear_disarmament by the nuclear powers, so that the treaty would not simply freeze an unequal world in place. They wanted a guarantee of access to civilian nuclear technology. And they wanted security assurances against nuclear attack or nuclear blackmail. The result was the #grand_bargain, and it is essential to understand it as a package rather than as a list of separate rules. States without nuclear weapons accepted a binding, verifiable ban on acquiring them. In exchange, they received a right to peaceful nuclear technology under safeguards, and a promise from the nuclear-armed states to negotiate in good faith toward disarmament. The treaty was opened for signature on 1 July 1968 and entered into force on 5 March 1970, with an initial duration of twenty-five years. India refused to sign, calling the treaty discriminatory because it divided the world into nuclear haves and have-nots on the basis of a date rather than on any principle of law or security. Israel and Pakistan also stayed outside. That criticism, that the treaty legalises inequality, has never gone away, and it remains the single most important political fault line inside the regime. 2.3 The definition that created two classes of states Article IX(3) defines a nuclear-weapon state as one that manufactured and exploded a nuclear weapon or other nuclear explosive device before 1 January 1967. This captures exactly five states: the United States, the Soviet Union (whose obligations and rights passed to the Russian Federation), the United Kingdom, France and China. These are the #nuclear_weapon_states, often called the P5 because they are also the permanent members of the Security Council. Everyone else who joins is a non-nuclear-weapon state. Two points deserve emphasis for students. First, the definition is historical, not normative. The treaty does not say that these five states are entitled to nuclear weapons forever; it says that they possessed them on a certain date and are therefore subject to a different set of rules. Second, the definition cannot be updated. India, Pakistan, Israel and North Korea possess nuclear weapons but cannot become nuclear-weapon states under the treaty without amending it, and amendment requires, in practice, the agreement of all five recognised nuclear powers and the Board of Governors of the agency. This is why the treaty and the actual distribution of nuclear weapons in the world have drifted apart. 3. The Legal Architecture: Reading the Treaty 3.1 Articles I and II: the core prohibitions Article I binds each nuclear-weapon state not to transfer nuclear weapons or explosive devices to any recipient, and not to assist, encourage or induce any #non_nuclear_weapon_states to manufacture or acquire them. Article II is the mirror image: each non-nuclear-weapon state undertakes not to receive, manufacture or otherwise acquire nuclear weapons, and not to seek or receive assistance in manufacturing them. These two articles are the heart of the #nonproliferation pillar. They are unusually clear by the standards of international law. But two ambiguities have generated decades of argument. The first concerns #nuclear_sharing arrangements within NATO, under which American weapons are stationed in the territory of non-nuclear allies and, in wartime, could in principle be delivered by allied aircraft. The United States and its allies argue that no transfer of control occurs in peacetime and that the treaty's negotiating record accommodates the practice. Russia and many non-aligned states reject that reading. The dispute resurfaced sharply after Russia announced the deployment of nuclear weapons in Belarus in 2023, a step Moscow defended by pointing to the NATO precedent. The second ambiguity concerns the phrase "manufacture." At what point does research, or the accumulation of technology and materials, become manufacture of a weapon? The treaty does not say. This is the legal space in which the concept of #nuclear_latency lives: a state can build a full civilian fuel cycle, including enrichment and reprocessing, and come very close to weapons capability without formally violating Article II. Japan is the standard example of a latent capability held under full transparency; Iran before 2015 is the standard example of latency that alarmed its neighbours. 3.2 Article III: safeguards Article III requires every non-nuclear-weapon state party to accept safeguards, negotiated with the agency, on all source and special fissionable material in all peaceful nuclear activities on its territory. The purpose is stated in the article itself: to verify that material is not diverted from peaceful uses to weapons. Article III(2) adds an export control rule: parties must not supply source or special fissionable material, or equipment especially designed for processing such material, to any non-nuclear-weapon state unless the material is under safeguards. This is the legal foundation for the #comprehensive_safeguards_agreement, whose standard model was published by the agency as document INFCIRC/153 in 1972. It is also the foundation for the international #export_controls system, including the Zangger Committee and the #Nuclear_Suppliers_Group, which was created in 1975 in direct response to India's 1974 nuclear test. Note the asymmetry. Safeguards under Article III are required only of non-nuclear-weapon states. The five nuclear-weapon states have concluded voluntary offer agreements, under which the agency may apply safeguards to selected civilian facilities, but their military stocks are entirely outside the system. This is legally coherent, since the treaty was never designed to verify nuclear-weapon states' arsenals, but politically it feeds the recurring charge of double standards. 3.3 Article IV: peaceful uses Article IV states that nothing in the treaty affects the inalienable right of all parties to develop research, production and use of nuclear energy for peaceful purposes, without discrimination and in conformity with Articles I and II. It also obliges parties to facilitate the fullest possible exchange of equipment, materials and scientific information for #peaceful_uses, with particular regard to the needs of developing countries. Article IV is the most contested provision in the treaty, and students should understand why. Read broadly, it guarantees a right to the complete #fuel_cycle, including uranium enrichment and spent fuel reprocessing, which are precisely the technologies that produce weapons-usable material. Read narrowly, it grants a conditional right, exercisable only in conformity with the non-proliferation obligations, and says nothing that specifically guarantees enrichment. Iran has consistently taken the broad reading. The United States and several European states have leaned toward the narrow one. Neither reading has been authoritatively settled, and the treaty contains no court to settle it. 3.4 Article V: peaceful nuclear explosions Article V, now effectively dead, promised non-nuclear-weapon states access to the potential benefits of peaceful nuclear explosions. In the 1960s, engineers seriously proposed using nuclear devices to excavate canals and harbours. The idea collapsed as the environmental and proliferation problems became clear, and Article V is today a historical curiosity that illustrates how much the technical assumptions of 1968 have changed. 3.5 Article VI: the disarmament obligation Article VI is one sentence, and it is the most argued-over sentence in arms control. It obliges each party to pursue negotiations in good faith on effective measures relating to cessation of the nuclear arms race at an early date and to nuclear disarmament, and on a treaty on general and complete disarmament under strict and effective international control. Three features of #Article_VI matter. First, the obligation falls on all parties, not only on nuclear-weapon states, although in practice it is the nuclear-armed states whose conduct is judged against it. Second, the obligation is to pursue negotiations in #good_faith, not to disarm by a fixed date. There is no deadline and no benchmark in the text. Third, it links nuclear disarmament to general and complete disarmament, which some nuclear-weapon states have historically read as a condition and most non-nuclear states reject as an excuse. The International Court of Justice addressed this in its 1996 advisory opinion on the legality of the threat or use of nuclear weapons. The Court concluded unanimously that there exists an obligation to pursue in good faith and bring to a conclusion negotiations leading to nuclear disarmament in all its aspects. The phrase "bring to a conclusion" is significant: it suggests that endless process without result would not satisfy the obligation. But the #ICJ has no enforcement power here, and the opinion is advisory. 3.6 Article VII: nuclear-weapon-free zones Article VII affirms the right of groups of states to conclude regional treaties assuring the total absence of nuclear weapons in their territories. This has produced a genuine and underrated success. Five #nuclear_weapon_free_zones now cover most of the southern hemisphere and a large part of Asia: Latin America and the Caribbean under the Treaty of Tlatelolco (1967), the South Pacific under Rarotonga (1985), Southeast Asia under Bangkok (1995), Africa under Pelindaba (1996), and Central Asia under Semipalatinsk (2006). Mongolia has a recognised single-state nuclear-weapon-free status. The zones matter for two reasons. They add legally binding regional prohibitions on top of the treaty's global one, and their protocols provide a route through which nuclear-weapon states give negative security assurances, that is, promises not to use or threaten nuclear weapons against zone members. The unratified protocols, and the reservations attached to some ratifications, are a persistent source of complaint. The zone that does not exist is the one that has caused the most damage inside the treaty: the proposed #Middle_East_WMD_free_zone, promised in a resolution adopted at the 1995 conference as part of the price for indefinite extension. Its non-realisation has poisoned at least three review conferences. 3.7 Article VIII and the review process Article VIII provides for amendment, which is practically impossible, and for review conferences every five years. The review process is the treaty's political engine. It has no power to change the treaty, and its outcome documents are political rather than legally binding, but it is the only place where all parties meet to assess implementation and to negotiate shared interpretations. When it works, as in 1995, 2000 and 2010, it clarifies and strengthens the rules. When it fails, as in 2005, 2015, 2022 and 2026, the failure itself becomes a political fact that damages confidence in the regime. 3.8 Articles IX and X: membership and withdrawal Article IX contains the definition discussed above and the entry-into-force provisions. #Article_X contains the withdrawal clause, which allows any party to withdraw if it decides that extraordinary events related to the subject matter of the treaty have jeopardised its supreme interests. Withdrawal requires three months notice to all other parties and to the Security Council, with a statement of the extraordinary events involved. The withdrawal clause is a live problem. North Korea announced withdrawal in 2003, after having built a nuclear programme in violation of its safeguards agreement while inside the treaty, and then tested a device in 2006. Its legal status remains contested; a number of states still list it as a party in some contexts. The core question raised by that episode is whether a state can violate the treaty from within, gain the material benefits of membership, and then walk away without consequence. Several proposals have been made to address this, including a declaration that a withdrawing state remains responsible for violations committed while a party, and that nuclear material and equipment supplied under the treaty must be returned or remain under safeguards. None has been adopted. Article X also contained the provision that produced the most consequential decision in the treaty's history: twenty-five years after entry into force, a conference would decide whether the treaty should continue indefinitely or be extended for fixed periods. That conference met in 1995. 4. The Machinery: How Verification Actually Works The treaty text is only the top layer. Underneath it sits a system of institutions, agreements and technical practices without which the treaty would be a statement of intent. Students often skip this material because it is technical. That is a mistake: almost every real dispute in the regime, from Iraq in 1991 to Iran in 2026, has been a dispute about verification. 4.1 The International Atomic Energy Agency The agency was created in 1957, eleven years before the treaty, with a dual mandate: to promote peaceful nuclear technology and to apply #safeguards that ensure it is not diverted to weapons. The treaty made agency safeguards compulsory for non-nuclear-weapon state parties, and in doing so turned a voluntary technical service into the verification arm of a global security regime. The agency's governing bodies are the General Conference of all member states and a Board of Governors of thirty-five states, which is the body that makes findings of non-compliance and refers cases to the Security Council. The secretariat, headed by a Director General, carries out inspections. It is important for students to see the political structure clearly: technical findings are produced by professional inspectors, but the decisions about what those findings mean, and what to do about them, are made by governments voting in the Board. 4.2 Comprehensive safeguards agreements Under the standard comprehensive safeguards agreement, a state declares all of its nuclear material and facilities. The agency then applies material accountancy, containment and surveillance to verify that declared material is not diverted. Inspectors count, measure and seal. Cameras watch. Records are audited. The technical goal is the timely detection of the diversion of a significant quantity of material, defined for practical purposes as roughly the amount needed for one device. The limitation of this system is obvious once stated: it verifies the correctness of a state's declaration, but it was not originally designed to verify its completeness. If a state simply does not declare a facility, classical safeguards may never find it. That is exactly what happened in Iraq, whose clandestine enrichment programme was uncovered only after the 1991 Gulf War, despite Iraq being a party in apparently good standing with an agreement in force. 4.3 The Additional Protocol The Iraqi shock produced the most important reform in safeguards history. In 1997 the Board approved the Model Additional Protocol, which grants the agency significantly broader rights: expanded declarations covering the whole nuclear fuel cycle including research and manufacturing, access to any location on a site at short notice, environmental sampling, and the use of satellite imagery and other open-source information. The #Additional_Protocol changes what the agency can conclude. With a comprehensive agreement alone, the agency can normally state only that declared material remains in peaceful use. With an Additional Protocol in force and implemented, the agency can draw the far stronger broader conclusion that there is no indication of undeclared nuclear material or activities in the state as a whole. Roughly 140 states have Additional Protocols in force. They are, however, voluntary. There is no treaty obligation to accept one, and several states with significant nuclear activities, including Iran since 2021, do not implement one. The empirical evaluation of the protocol is mixed rather than triumphant. Its expanded access has demonstrably improved the agency's ability to detect undeclared activity, but it did not prevent the Iranian crisis, and its adoption has stalled well short of universality (Gibbons and Robinson 2021). 4.4 Export controls Article III(2) requires safeguards as a condition of supply, but it does not list what counts as sensitive equipment. That gap was filled outside the treaty. The Zangger Committee produced a trigger list in the 1970s. The Nuclear Suppliers Group, formed after India's 1974 test, went further, covering #dual_use items and eventually adopting guidelines requiring comprehensive safeguards as a condition of supply. Export control regimes are effective but politically awkward. They are informal, they are dominated by supplier states, and they are viewed by many developing countries as cartels that restrict the very technology access that Article IV promises. The 2008 exemption granted by the group to India, a non-party with nuclear weapons, damaged the regime's credibility badly. It signalled that a state could stay outside the treaty, test weapons, and still obtain civilian nuclear cooperation if its strategic relationships were right. 4.5 Enforcement and the Security Council The treaty has no enforcement mechanism of its own. Article XII.C of the agency's statute allows the Board, on finding non-compliance, to report to the Security Council and the General Assembly. What happens next is a political decision by the #UN_Security_Council, where the five veto powers are also the five nuclear-weapon states. This structure has three consequences. Enforcement is selective, because it depends on great-power agreement. Enforcement is slow, because coalition-building takes years. And enforcement is inconsistent, because the same five states that police the regime are the states whose disarmament record is under criticism. Every serious analysis of #compliance in this field returns to that structural point. 5. The Review Process: Eleven Conferences, 1975 to 2026 The treaty is reviewed every five years. The record is worth learning in outline, because the pattern of success and failure is itself analytically revealing. 5.1 The first cycles, 1975 to 1990 The 1975 conference produced a consensus final declaration. The 1980 and 1990 conferences did not, largely because non-nuclear states objected to the lack of progress on a comprehensive test ban and on Article VI generally. The 1985 conference produced a final document. The pattern established early was that the review process succeeds when the nuclear-armed states can point to visible movement on disarmament, and fails when they cannot. 5.2 The 1995 Review and Extension Conference This was the decisive moment. The treaty was due to expire, and the parties had to decide whether to extend it indefinitely, for one fixed period, or for a series of fixed periods. The nuclear-weapon states, led by the United States, ran a major diplomatic campaign for #indefinite_extension. Many non-aligned states wanted rolling extensions precisely because periodic renewal would preserve leverage over the disarmament agenda. The outcome was a package adopted without a vote. The treaty was extended indefinitely. In exchange, the conference adopted a set of Principles and Objectives for Nuclear Non-Proliferation and Disarmament, a decision to strengthen the review process, and a Resolution on the Middle East calling for a zone free of weapons of mass destruction in the region. Two consequences followed, and both still shape the regime. The first is that the non-nuclear majority gave away its principal source of leverage in return for political commitments that were not legally binding (Onderco 2021; Tannenwald 2024). The second is that the Middle East resolution created an expectation that has never been fulfilled, and that has since been used both as a bargaining chip and as a blocking device in successive conferences. 5.3 2000 and the thirteen practical steps The 2000 conference is generally rated the high point of the review process. It produced a consensus document containing thirteen practical steps toward disarmament, including an unequivocal undertaking by the nuclear-weapon states to accomplish the total elimination of their nuclear arsenals. That phrase removed the argument that Article VI required only endless negotiation without result. 5.4 2005: collapse The 2005 conference achieved essentially nothing and spent most of its time on procedure. The atmosphere had been damaged by the American withdrawal from the Anti-Ballistic Missile Treaty, the invasion of Iraq on the basis of claims about weapons of mass destruction that proved false, and the revelation of the A. Q. Khan network, which had sold centrifuge technology to Libya, Iran and North Korea. 5.5 2010: the action plan The 2010 conference recovered, producing a consensus outcome with a 64-point action plan across all three pillars and an agreement to convene a conference on a Middle East zone in 2012. The mood was helped by the signing of the New Strategic Arms Reduction Treaty between the United States and Russia earlier that year, which showed that the two largest arsenals could still be negotiated down. 5.6 2015: failure over the Middle East The promised 2012 Middle East conference never happened. In 2015, the conference collapsed on the final day when the United States, supported by the United Kingdom and Canada, refused to accept text on convening a Middle East conference to a deadline without Israeli consent. Arab states regarded this as a breach of the 1995 bargain. The failure was not about disarmament in the abstract; it was about a specific, twenty-year-old promise. 5.7 2022: failure over Ukraine The tenth conference, postponed repeatedly by the pandemic, met in August 2022, six months after Russia's full-scale invasion of Ukraine. Negotiators came close to a consensus text. Russia blocked it, objecting to language on the seizure of the Zaporizhzhia nuclear power plant and on Ukrainian control over nuclear facilities on its own territory. The episode showed something important: a treaty process that requires consensus can be held hostage by the single state whose behaviour is the subject of the text (Mukhatzhanova 2022). 5.8 2026: the third consecutive failure The eleventh review conference met at United Nations headquarters in New York from late April to 22 May 2026. Expectations were low from the start, since the 2023, 2024 and 2025 preparatory committee meetings had all failed to agree recommendations. The conference met against the background of the war in Ukraine, the June 2025 Israeli and American strikes on Iranian nuclear facilities, the reimposition of United Nations sanctions on Iran in September 2025, and the expiry of the New Strategic Arms Reduction Treaty in February 2026. The conference president, Ambassador Do Hung Viet of Vietnam, took the unusual step of not tabling the draft final document at all, judging that consensus did not exist and preferring not to force a public blocking. He later described the outcome as disappointing but not surprising. The most contested issue in the final week was country-specific language, above all on Iran, alongside familiar disagreements over the pace of disarmament. Reporting on the negotiations indicates that the five nuclear-weapon states worked to remove the more specific disarmament commitments from the draft, while many non-nuclear-weapon states argued that the nuclear-armed states had shown less flexibility than the rest of the membership. Three points deserve emphasis. First, failure to agree a document is not the same as failure of the treaty. A consensus final document has been achieved in only four of eleven conferences, in 1975, 1985, 2000 and 2010, and the treaty survived all the other failures. Second, three consecutive failures nonetheless represent a pattern rather than an accident, and they coincide with the erosion of nearly every other pillar of #arms_control. Third, the substantive agenda that did attract broad support, including greater #transparency in national reporting by nuclear-armed states and practical #nuclear_risk_reduction measures, still exists and can be pursued outside a consensus document. The next review conference is scheduled for 2031. 6. Compliance and Crisis: The Case Record The treaty's performance is best judged not by the elegance of its text but by what happened when states tried to break it. Eight cases are essential. 6.1 South Africa: the only state to disarm South Africa built six nuclear devices in secret during the 1970s and 1980s, then dismantled them, acceded to the treaty in 1991, and accepted agency verification of the dismantlement. It remains the only state to have built an arsenal and then given it up. The case matters because it shows that #verification of disarmament is technically possible, and because the decision to disarm was driven by a change in the country's security and political situation rather than by treaty pressure alone. 6.2 Iraq: the failure that reformed safeguards Iraq was a party with a safeguards agreement in force. After the 1991 war, inspectors found an extensive undeclared enrichment programme. The agency had been inspecting declared facilities and finding nothing wrong, because it had neither the mandate nor the tools to look elsewhere. The result was the Additional Protocol and a shift in safeguards thinking from narrow accountancy toward assessment of a state's nuclear programme as a whole. 6.3 North Korea: the treaty's clearest defeat #North_Korea joined the treaty in 1985, delayed its safeguards agreement, was found in non-compliance in 1993, threatened withdrawal, negotiated the Agreed Framework with the United States in 1994, and then, after that arrangement collapsed amid mutual accusations, announced withdrawal in January 2003. It tested a device in 2006 and has since built an arsenal with an active production infrastructure and a declared doctrine of use. The case exposes each weakness of the regime in turn: delay in bringing safeguards into force, reliance on a bilateral deal outside the treaty, the ambiguity of Article X, the inability of the Security Council to impose costs sufficient to change behaviour, and the absence of any mechanism to reverse #breakout once it has occurred. 6.4 Libya: the case that worked Libya pursued a covert weapons programme, buying centrifuge designs and components through the A. Q. Khan network. In 2003, after negotiations with the United States and the United Kingdom, and in a regional context shaped by the invasion of Iraq, Libya agreed to give up the programme. Equipment was removed and the agency verified. The case is regularly cited by advocates of coercive diplomacy. It is also cited, by others, as a warning, because Muammar Gaddafi was overthrown with Western military assistance eight years later, a sequence that other governments have studied closely when weighing whether disarmament buys security. 6.5 Syria In 2007 Israel destroyed a structure at Dair Alzour that the agency later assessed was very likely an undeclared reactor. Syria did not cooperate with the subsequent investigation and was found in non-compliance by the Board in 2011. The case remains formally unresolved, which is one reason it appears in the working papers of every review conference. 6.6 Iran: the central case of the present era The Iranian file is the longest-running and most consequential compliance dispute in the treaty's history, and students should know its main phases. Undeclared enrichment facilities at Natanz and a heavy-water project at Arak were revealed in 2002. Between 2003 and 2005, negotiations with three European states produced a temporary suspension. In September 2005 the Board found Iran in non-compliance with its safeguards agreement, and in 2006 the file was referred to the Security Council, which adopted a series of sanctions resolutions between 2006 and 2010. In July 2015, Iran and six powers concluded the Joint Comprehensive Plan of Action. The #JCPOA capped enrichment at 3.67 percent, limited the enriched uranium stockpile to 300 kilograms, restricted centrifuge numbers and types, redesigned the Arak reactor, and imposed an unusually intrusive verification regime that included provisional application of the Additional Protocol. Security Council Resolution 2231 endorsed the deal and set out a #snapback mechanism under which the earlier sanctions could be reimposed if Iran failed to perform. The United States withdrew from the agreement in 2018 and reimposed sanctions. From 2019 Iran began exceeding the limits. In February 2021 it stopped implementing the Additional Protocol. It accumulated uranium enriched to 60 percent, a level with no plausible civilian justification and technically close to weapons grade, becoming the only non-nuclear-weapon state party ever to do so. In June 2025, the Board adopted a resolution finding Iran in non-compliance with its safeguards agreement. Days later, Israel and then the United States conducted military strikes on Iranian nuclear facilities. The agency withdrew its inspectors for safety reasons, and Iran formally suspended cooperation in July 2025. In August 2025 the three European parties invoked the snapback mechanism, and the previously terminated Security Council resolutions returned to force at the end of September 2025. Iran rejected the legality of that step. Through 2026, agency reporting has described a near-total loss of access, an unverified stockpile of highly enriched uranium, and a loss of continuity of knowledge over declared material, while also reporting no evidence of an active weaponisation programme. The Iranian case matters for the treaty as a whole for three reasons. It tests whether verified restraint can be rewarded, and the collapse of the 2015 agreement suggests a discouraging answer. It raises the unresolved legal question of whether safeguarded facilities may lawfully be attacked, and what happens to inspection when they are. And it is now the single issue most capable of blocking consensus in the review process, as the 2026 conference showed. 6.7 The states outside: India, Israel, Pakistan India tested in 1974 and again in 1998. Pakistan tested in 1998. Israel has never confirmed or denied an arsenal but is universally assessed to hold one. All three have nuclear weapons, none is a party, and none accepts comprehensive safeguards. Their existence poses a problem the treaty cannot solve on its own terms. The treaty binds only its parties. It does not make possession by non-parties unlawful. Yet a regime whose central claim is that nuclear weapons should not spread is weakened when three states acquired them outside the regime and paid, over time, relatively modest and reversible costs. The 2008 civil nuclear arrangement between the United States and India, and the accompanying supplier group exemption, sharpened the problem by showing that the price of staying outside could be negotiated down. 6.8 Ukraine and the Budapest Memorandum Ukraine inherited a large part of the Soviet arsenal in 1991 and gave it up, acceding to the treaty as a non-nuclear-weapon state in 1994. In exchange it received assurances from the United States, the United Kingdom and Russia, recorded in the #Budapest_Memorandum, to respect its independence and existing borders and to refrain from the threat or use of force against it. Russia's annexation of Crimea in 2014 and its full-scale invasion in 2022 broke those assurances. The consequences for the regime are hard to overstate. The lesson drawn in several capitals was that refraining from nuclear weapons may not buy security, and that assurances offered in exchange may not be honoured. The occupation of the #Zaporizhzhia plant added a second lesson: safeguarded civilian facilities can become instruments and targets in war. 7. The Regime Under Stress, 2020 to 2026 The years since 2020 have brought a cluster of pressures that arrived together. Taken separately, each would be manageable. Taken together, they amount to the most serious strain the treaty has faced. 7.1 The collapse of bilateral arms control The treaty has always depended on progress outside itself. When the two largest arsenals were shrinking under negotiated agreements, the disarmament pillar looked credible. That external support has now gone. The Intermediate-Range Nuclear Forces Treaty ended in 2019. The Open Skies Treaty unravelled in 2020 and 2021. Russia suspended its participation in the New Strategic Arms Reduction Treaty in February 2023, ending on-site inspections and data exchanges, though it said it would continue to observe the numerical limits. The treaty could be extended only once, and it expired on 5 February 2026. The expiry of #New_START leaves the United States and Russia without any legally binding cap on their deployed strategic forces for the first time since the early 1970s. In September 2025 President Putin proposed that both sides continue to observe the central limits informally for one year. The United States did not formally accept, and after expiry it stated that it would seek a new and modernised agreement, arguing that any future framework must also cover China's expanding arsenal. Russian officials have said they will observe the former limits so long as the United States does the same. This is a political understanding at best, without verification, and it can be abandoned at any time. The effect on the treaty is direct. Article VI is judged by results, and there are now no results to point to. #Strategic_stability between the largest arsenals now rests on unilateral restraint and mutual observation rather than on law. 7.2 Modernisation and expansion Every nuclear-armed state is modernising. Several are expanding. Recent assessments by the Stockholm International Peace Research Institute describe arsenals that are being upgraded and, in some cases, growing, with China's expansion the most rapid and India's posture shifting toward canisterised, ready-to-launch systems and submarine patrols. #Modernization is not in itself a treaty violation. Article VI requires good-faith negotiation, not a freeze on maintenance. But the political effect on the review process is severe, because non-nuclear-weapon states see hundreds of billions of dollars committed to arsenals that are supposed to be on a path to elimination. 7.3 The prohibition treaty and the split in the disarmament movement Frustration with the pace of Article VI implementation produced the #humanitarian_initiative of 2013 to 2015, a series of conferences that reframed nuclear weapons as a humanitarian and public health problem rather than only a strategic one. That movement, driven by non-nuclear states, the Red Cross movement and civil society, led to the Treaty on the Prohibition of Nuclear Weapons, adopted in 2017 and in force since January 2021. The #TPNW bans the development, testing, production, possession, stationing and use of nuclear weapons for its parties. No nuclear-armed state has joined. No state under an extended deterrence arrangement has joined. Its supporters argue that it fills the legal gap left by Article VI, stigmatises possession, and gives non-nuclear states a way to act rather than wait. Its critics, including most NATO members, argue that it is unverifiable in practice, that it risks undermining the older treaty by creating a rival forum, and that it cannot deliver disarmament because it does not include any state that has weapons to give up. Scholars are divided. Some read the prohibition treaty as counter-institutionalisation that will eventually pull the older regime toward its own goals (Tannenwald 2024). Others argue that contestation of this kind has historically strengthened rather than weakened the older treaty, because it forces the nuclear-armed states to justify themselves and to make concessions in order to preserve #legitimacy (Smetana and O'Mahoney 2022). What is not in dispute is that the two treaties now coexist uneasily, and that the disagreement runs through the review process. 7.4 The test ban and the fissile material treaty Two long-promised measures remain undelivered. The Comprehensive Nuclear-Test-Ban Treaty was adopted in 1996 but has never entered into force, because it requires ratification by a specific list of forty-four states, several of which, including the United States, China, Egypt, Iran and Israel, have not ratified, while India, Pakistan and North Korea have not signed. Russia withdrew its ratification in 2023, aligning its status with that of the United States. The #CTBT nonetheless operates a global monitoring system that detects tests with high reliability, and a testing moratorium has held among the five recognised nuclear powers since the 1990s. A treaty banning the production of fissile material for weapons, the #FMCT, has been on the agenda of the Conference on Disarmament since 1993 and has never begun formal negotiation, blocked by disagreements over whether existing stocks would be covered. Both are cited at every review conference as evidence that the disarmament pillar is stalled. Both are also examples of a general lesson: the treaty depends on instruments it cannot itself create. 7.5 AUKUS and naval nuclear propulsion In 2021 Australia, the United Kingdom and the United States announced that Australia would acquire conventionally armed, nuclear-powered submarines. Because naval reactors in the American and British designs use highly enriched uranium, the arrangement would involve transferring weapons-usable material to a non-nuclear-weapon state. The legal route runs through Article 14 of Australia's comprehensive safeguards agreement, which allows for arrangements under which safeguards may be adjusted for nuclear material used in non-proscribed military activity such as #naval_nuclear_propulsion. The three partners say no material will be transferred before an Article 14 arrangement is concluded with the agency, that the arrangement will be submitted to the Board, and that the outcome will set the highest non-proliferation standard. The Director General has confirmed that the agency's role here falls within its existing legal framework. China and Russia object, arguing that #AUKUS sets a precedent under which non-nuclear states could remove significant quantities of material from routine inspection under a military exemption. Australia and its partners reply that the exemption is explicit in the standard safeguards model, that it was foreseen by the drafters, and that Brazil has been pursuing a comparable arrangement for its own naval programme. Students should notice the structure of the argument, because it recurs. Both sides are making legal claims that are defensible on the text. The dispute is really about precedent, trust, and whether a loophole that has never been used should be opened now. 7.6 The revival of nuclear energy The peaceful uses pillar is enjoying a revival, driven by climate policy, energy security and rising electricity demand. Many states that had no plans for #nuclear_energy a decade ago are now considering it, and interest in #small_modular_reactors has grown quickly. The agency's own projections for installed capacity to 2050 have been revised upward in recent years. This is a treaty success in one sense: Article IV is doing what it was designed to do. But growth in civilian nuclear power increases the number of states with nuclear expertise, materials and infrastructure. If new programmes are built with imported fuel and without national enrichment or reprocessing, the proliferation risk stays low. If they are built with domestic fuel cycles, #nuclear_latency spreads. The governance question of the next decade is therefore not whether nuclear energy expands, but on what terms. 7.7 Attacks on nuclear facilities Between 2022 and 2026 the world saw sustained military activity around the Zaporizhzhia plant in Ukraine and direct strikes on safeguarded nuclear facilities in Iran. The legal position is contested. Additional Protocol I to the Geneva Conventions restricts attacks on nuclear electrical generating stations in certain circumstances, and a general assembly resolution and agency statements have condemned attacks on safeguarded facilities. Yet the practice of the last four years suggests an emerging reality in which such facilities are treated as legitimate targets. For the treaty this is corrosive in a specific way. Article IV promises safe access to peaceful nuclear technology. If accepting safeguards makes a state's facilities into a map for an attacker, the incentive structure of the treaty is inverted. This argument is now made openly in Tehran and has been noticed elsewhere. 7.8 Emerging technologies Three technological developments are now regular agenda items. #Artificial_intelligence raises questions about decision support in nuclear command and control, about the speed of escalation, and about whether machine learning tools lower the barrier to weapons-relevant knowledge. Advanced conventional weapons, including precision strike and missile defence, blur the line between nuclear and non-nuclear deterrence and complicate arms control arithmetic (Futter and Zala 2021). Cyber operations threaten the integrity of early warning and command systems, and of safeguards data itself, which is part of the wider field of #nuclear_security. None of these is addressed in the 1968 text. The review process has begun to discuss them, and the 2026 conference did generate broad interest in the topic, but the treaty has no mechanism for adapting its rules to new technology short of political agreement in a body that cannot reach political agreement. 7.9 Alliance anxiety and the hedging debate Perhaps the most dangerous development is the least visible. In South Korea, opinion polling has repeatedly shown majority support for an indigenous arsenal. In parts of Europe, the reliability of American #extended_deterrence is being debated openly, along with the possibility of European nuclear arrangements. In the Middle East, Saudi officials have said that if Iran acquires a weapon, they would seek one. These are not programmes. They are conversations. But the treaty's stability has always rested on the fact that most capable states concluded that they were safer without nuclear weapons than with them, in large part because of alliance guarantees. If that calculation changes in even two or three capable states, the risk of a #proliferation_cascade becomes real. The literature calls the intermediate position #hedging: keeping options open, building capability, and deferring the decision. 8. Theoretical Perspectives The treaty is a standing test case for the main theories of international relations, and a good student answer should be able to use more than one lens. 8.1 Realism Realists explain the treaty as an artefact of power. It exists because the two superpowers wanted it, it survives because the strong find it useful, and it constrains the weak more than the strong. On this account, states acquire nuclear weapons when their security environment demands it and refrain when it does not; the treaty is a coordination device, not a cause of restraint. The realist prediction for the current period is straightforward: as great-power competition intensifies and alliance guarantees look less reliable, more states will hedge, and the treaty will not stop them. The realist account explains the non-parties well. It explains the North Korean case well. It struggles to explain why so many capable states, from Japan to Germany to Brazil to Sweden, have chosen to stay non-nuclear for decades even when they had both the capability and plausible security motives. 8.2 Liberal institutionalism Institutionalists focus on what the regime does that states cannot do alone: it provides information through safeguards, it reduces uncertainty about others' intentions, it lowers the transaction costs of cooperation, and it creates a focal point around which expectations converge. Verification is the key mechanism. If I can be reasonably confident that my neighbour is not building a weapon, my own incentive to build one falls. This perspective explains the durability of the regime and the centrality of the #IAEA. Its weakness is enforcement: the theory assumes that violations will be detected and punished, and the record of punishment is patchy. 8.3 Constructivism and the normative account Constructivists argue that the treaty helped create and sustain a norm, that nuclear weapons are not ordinary weapons and that acquiring them is illegitimate. The related idea of the #nuclear_taboo holds that the non-use of nuclear weapons since 1945 reflects not only deterrence but a normative prohibition that has grown stronger over time. On this account, the regime's real product is not compliance enforced by inspection but a widely shared understanding that possession is shameful and that restraint is normal. The prohibition treaty is best understood in these terms: it is an attempt to strengthen stigma when law and negotiation have stalled. Critics respond that norms are cheap when interests align and collapse when they do not, and they point to nuclear signalling during the war in Ukraine as evidence that the taboo is thinner than claimed. The more careful position in the recent literature is that the taboo is real but conditional, that it operates strongly on decision-makers in ordinary times and weakly in extreme ones, and that public opinion is more permissive than the elite discourse suggests (Smetana and Wunderlich 2021). 8.4 Critical and postcolonial readings A fourth body of work asks who the regime serves. It notes that the treaty divides the world into two legal classes on the basis of a 1967 date, that the enforcement mechanism sits with five states that are also the permanent members of the Security Council, and that pressure has fallen far more heavily on states outside the Western alliance system than on those inside it. The charge of hypocrisy is not merely rhetorical; it has measurable political effects, because it makes it harder to build coalitions for enforcement when they are needed. Even scholars sympathetic to the regime accept that the perception of double standards is one of the principal threats to its long-term survival (Gibbons 2022). 8.5 The current debate: fragile or antifragile The most useful contemporary framing for students is a direct disagreement in the literature. One position holds that the regime is a failed promise whose internal contradictions are steadily undermining it, that the asymmetry between hard non-proliferation rules and soft disarmament commitments has produced polarisation that the review process cannot resolve, and that the trajectory is downward (Tannenwald 2024). The other holds that the regime is antifragile, that contestation and periodic crisis are how it adapts, and that predictions of collapse have been made after every failed conference since 1980 and have all been wrong so far (Smetana and O'Mahoney 2022). The evidence supports both, which is why the debate is interesting. Membership is near universal and has never fallen. Only one state has ever left. The number of nuclear-armed states has grown far more slowly than 1960s forecasts predicted. At the same time, the disarmament pillar is stalled, the review process has failed three times in a row, verification in the most important compliance case has collapsed, and the last bilateral limits have expired. A student who can hold both facts together, and say what evidence would settle the question, is doing the analysis properly. 9. Assessing the Record 9.1 The case for success Four claims can be defended with evidence. First, the spread of nuclear weapons has been far slower than expected. In 1963 a spread to twenty or twenty-five states within a decade was considered plausible by serious people. Today nine states hold nuclear weapons. Whatever else is true, the forecast of rapid diffusion did not happen. Second, the regime has produced reversals. South Africa dismantled an arsenal. Ukraine, Belarus and Kazakhstan gave up inherited weapons. Argentina and Brazil abandoned competitive programmes and built a bilateral inspection system. Libya gave up a covert programme. Sweden, Switzerland, South Korea, Taiwan and others halted programmes at various stages. Some of these outcomes had little to do with the treaty; several had a great deal to do with it. Third, the safeguards system works, within its limits. It has caught violations. It has provided reliable assurance about the vast majority of the world's civilian nuclear material for decades, at a cost that is trivial compared with military budgets. Fourth, the treaty created a default. Non-possession is the normal condition, and a state that wants weapons must justify departing from a rule that nearly every other state accepts. That is not nothing. It changes the domestic politics of the decision in every capital. 9.2 The case for failure Four counter-claims are equally defensible. First, the treaty did not stop the states that were determined. Israel, India, Pakistan and North Korea all have weapons. The treaty's success is largely a success with states that were not going to build weapons anyway, which raises a hard question about causation that the literature has never fully resolved. Second, the disarmament pillar has not delivered. Global arsenals fell sharply after the Cold War, but that decline has stalled. Every possessor is modernising. Some are expanding. The unequivocal undertaking of 2000 has not been honoured in any meaningful sense, and with the expiry of the last bilateral limits in 2026 there is no negotiated ceiling at all. Third, enforcement is unreliable. North Korea violated the treaty, withdrew, built weapons, and remains a functioning state with an arsenal. The costs imposed were real but insufficient. The lesson available to any future violator is not encouraging. Fourth, the political consensus that sustains the regime is visibly thinning. Three failed review conferences, an open split over the prohibition treaty, an unresolved Middle East file, and the erosion of confidence in security assurances after Ukraine all point the same way. 9.3 The counterfactual problem Students writing about this subject should be careful about one methodological trap. It is not possible to observe the world in which the treaty does not exist. Claims that the treaty prevented proliferation and claims that it did not are both, strictly speaking, claims about a counterfactual that cannot be measured directly. The best available evidence is indirect and comes from three sources: the documented cases where treaty membership and safeguards changed a state's calculation or capability; the systematic historical work on why states with capability chose restraint; and the study of alliance guarantees, which suggests that #extended_deterrence has done a great deal of the work often credited to the treaty. A careful answer therefore says that the treaty was one necessary component of a wider system of restraint, and that it is unlikely to have been sufficient on its own. 10. Reform: What Could Actually Be Done Reform proposals fall into five groups. None requires amending the treaty, which is politically impossible. 10.1 Strengthening the review process The most modest proposals are procedural. States could agree to structured, in-person discussion of the national reports submitted by nuclear-weapon states, which would create a form of #accountability where none currently exists; this was in the draft outcome document that was withdrawn in 2026. They could agree standard reporting templates so that reports become comparable. They could move away from the requirement of consensus for procedural outcomes, so that a small number of states cannot block the whole process. They could allow the president of a conference to issue a summary on his or her own authority when consensus fails, so that four weeks of work produce a usable record. 10.2 Closing the withdrawal gap Proposals here include an agreed understanding that a withdrawing state remains legally responsible for violations committed while a party; that nuclear material, equipment and facilities acquired under the treaty must remain under safeguards or be returned; and that the Security Council will automatically take up any notice of withdrawal. None of these requires amending Article X, since they can be adopted as an agreed interpretation. All have been blocked by states that see withdrawal as a sovereign right that must not be qualified. 10.3 Strengthening safeguards Making the Additional Protocol the universal standard, formally or in practice, is the single most valuable technical reform available. Related proposals include making the protocol a condition of nuclear supply through the suppliers group, and giving the agency the resources it needs as the number of facilities under safeguards grows with the expansion of civilian nuclear power. A separate technical agenda concerns #disarmament_verification, the science of confirming that warheads have actually been dismantled, which is a precondition for any serious future disarmament agreement and is currently underdeveloped. 10.4 Managing the fuel cycle If the risk lies in enrichment and reprocessing, the logical answer is to make national fuel cycles unnecessary. Proposals include fuel banks, of which the agency operates one in Kazakhstan; assured supply arrangements; multinational ownership of enrichment facilities; and take-back arrangements for spent fuel that keep #plutonium out of national hands. The obstacle is Article IV: any scheme that looks like a permanent denial of rights to some states will be rejected. Any scheme that is genuinely non-discriminatory would have to include the suppliers themselves, and they have shown no interest. 10.5 Risk reduction as the realistic near-term agenda With disarmament stalled and arms control expired, many practitioners now argue that the achievable agenda is risk reduction: crisis communication channels, pre-launch notification arrangements, agreed restraint on attacks against nuclear facilities and command systems, transparency about doctrine, and reaffirmation of the principle that a nuclear war cannot be won and must never be fought. These measures do not reduce arsenals. They reduce the probability of use by accident, misjudgement or escalation. In 2026 this is the part of the agenda with the broadest support and the fewest ideological obstacles. 11. A Study Guide This section is aimed at students preparing essays or examinations. 11.1 Key terms Grand bargain. The exchange at the heart of the treaty: non-acquisition and safeguards by non-nuclear states in return for peaceful nuclear cooperation and a disarmament commitment by nuclear states. Nuclear-weapon state. A state that manufactured and exploded a nuclear device before 1 January 1967. Five states qualify. The category cannot be enlarged. Comprehensive safeguards agreement. The agreement every non-nuclear-weapon state party must conclude with the agency, covering all nuclear material in all peaceful activities. Additional Protocol. A voluntary supplementary agreement giving the agency broader access and information rights, and the basis for concluding that there are no undeclared activities. Breakout. The time required for a state to convert declared or covert capability into enough weapons-usable material for a device. Latency. The condition of having the technical capacity to build weapons quickly without having done so. Snapback. The mechanism in Security Council Resolution 2231 allowing the automatic reimposition of previously lifted sanctions on Iran, invoked in 2025. Negative security assurance. A promise by a nuclear-weapon state not to use or threaten nuclear weapons against a non-nuclear-weapon state, usually given through the protocols to regional zone treaties. 11.2 Common misconceptions The treaty bans nuclear weapons. It does not. It bans acquisition by its non-nuclear parties and requires good-faith negotiation toward disarmament by all. The prohibition treaty of 2017 bans possession, but only for its own parties, none of which has nuclear weapons. The treaty makes India, Pakistan and Israel illegal possessors. It does not. They are not parties, and a treaty binds only its parties. The agency polices nuclear weapons. It does not. It verifies that declared civilian material is not diverted, and, where an Additional Protocol is in force, looks for undeclared activity. It has no jurisdiction over the arsenals of nuclear-weapon states. A failed review conference means the treaty has failed. It does not. The obligations remain in force regardless of what a conference agrees. But repeated failure is evidence of political decay, and evidence should not be dismissed. Article IV guarantees a right to enrich uranium. This is contested. The text guarantees a right to peaceful nuclear energy in conformity with Articles I and II, and says nothing explicit about enrichment. Both the broad and the narrow reading have serious defenders. 11.3 Essay questions Was the indefinite extension of 1995 a strategic error by the non-nuclear-weapon states? Argue with reference to the leverage they surrendered and the commitments they obtained. Does the Treaty on the Prohibition of Nuclear Weapons strengthen or undermine the older regime? Use the contestation literature. Can any verification system detect a determined, well-resourced covert programme? Use Iraq, Iran and North Korea. Assess the claim that the collapse of the Budapest Memorandum has done more damage to non-proliferation than any failed review conference. If you were advising the president of the 2031 review conference, what single procedural reform would you prioritise, and why? 11.4 How to use sources in this field Three cautions. First, distinguish between what the agency reports and what governments say about it; the two are frequently conflated in press coverage. Second, note that arsenal figures are estimates produced by a small number of open-source researchers, and treat them as estimates. Third, be alert to the political position of the institution publishing a claim; almost every organisation active in this field, including states, has a settled view of what the answer should be. 12. Limitations and Future Research This article rests on open sources and on a qualitative reading of the record. It cannot resolve the causal question of how much restraint the treaty itself produced, as opposed to alliance guarantees, technical difficulty, cost, or domestic politics. That question deserves more systematic comparative work, particularly on the states that stopped programmes. Three further research agendas are worth naming. The first concerns the legal status of attacks on safeguarded facilities, an area where practice is now running ahead of doctrine. The second concerns the governance of nuclear energy expansion, and specifically whether the arrival of many new entrants with small reactors can be reconciled with the existing safeguards model and its resource base. The third concerns #artificial_intelligence and nuclear decision-making, where the policy discussion is currently far ahead of the empirical evidence. 13. Conclusion The treaty was designed to do three things: stop the spread of nuclear weapons, move the world toward disarmament, and share the benefits of nuclear technology. On the first, it has performed better than its architects feared, though not as well as its supporters claim, and not without four significant failures. On the second, it has not delivered, and in 2026 the disarmament pillar is in a worse condition than at any point since the treaty entered into force. On the third, it has succeeded to the point where its success now creates new problems. The regime's central weakness is structural rather than accidental. The treaty asks non-nuclear states to accept a permanent, verified, enforceable restriction in exchange for a promise that is neither verified nor enforceable, made by the states that also control the enforcement machinery. That arrangement held for fifty years because the alternative looked worse to almost everyone, because alliance guarantees substituted for national arsenals, and because the nuclear-armed states did, for a period, deliver visible reductions. Two of those three supports are now weakening at the same time. The situation in 2026 is therefore serious but not hopeless. The legal obligations remain in force, and no state other than North Korea has ever left. Membership is close to universal, and the goal of full #NPT_universality, though distant, has not been abandoned. The safeguards system continues to function everywhere except in one country. The failure of a review conference to produce a document does not repeal a treaty. What has been lost is not law but confidence, and confidence is rebuilt by action rather than by text. If the states parties want the regime to survive the next crisis, the most plausible path runs through modest, concrete steps rather than grand declarations: transparency and reporting that can be examined rather than merely filed, risk reduction measures that lower the chance of nuclear use by accident, a serious effort to restore verified limits between the largest arsenals, universal adoption of the Additional Protocol, and a genuine answer to the question that the treaty has never faced squarely, which is what happens to a state that cheats and leaves. None of that is easy. All of it is possible without changing a word of the treaty. The alternative is a slow drift toward a world where the treaty still exists on paper while more states quietly decide that the promise made to them in 1968 was not kept, and that the safest course is to keep their options open. Hashtags #Treaty_on_the_Non_Proliferation_of_Nuclear_Weapons #NPT_1968 #Nuclear_Non_Proliferation #Nuclear_Arms_Control #Global_Nuclear_Governance #NPT_Review_Conference_2026 #IAEA_Safeguards_Regime #Disarmament_Studies #International_Security_Studies #Nuclear_Politics #Weapons_of_Mass_Destruction #Nuclear_Treaty_Law #Peace_and_Conflict_Studies #Student_Research_Resources References Arms Control Association. (2026). 2026 NPT Review Conference stymied by disputes. Arms Control Today, 56(5), 18-21. Bencini, L., van der Kwast, H. C., & Zlauvinen, G. (2023). Reflections on review conferences: The Non-Proliferation Treaty, the Biological Weapons Convention and the Chemical Weapons Convention. United Nations Institute for Disarmament Research. https://doi.org/10.37559/WMD/23/RevCon/01 Bidgood, S., & Potter, W. C. (2023). Nuclear arms control and the erosion of the disarmament dialogue. In A. Vicente, P. Sinovets, & J. Theron (Eds.), Russia's war on Ukraine: The implications for the global nuclear order (pp. 15-32). Springer. Futter, A., & Zala, B. (2021). Strategic non-nuclear weapons and the onset of a Third Nuclear Age. European Journal of International Security, 6(3), 257-277. https://doi.org/10.1017/eis.2021.2 Gibbons, R. D. (2022). The hegemon's tool kit: US leadership and the politics of the nuclear nonproliferation regime. Cornell University Press. Gibbons, R. D., & Robinson, T. (2021). Twenty-five years safer? Assessing the IAEA's Model Additional Protocol and its role in international politics. The Nonproliferation Review, 28(4-6), 297-318. https://doi.org/10.1080/10736700.2022.2027162 Herrera, M. (2026). What future for the NPT? Scenarios for the treaty going forward. Journal for Peace and Nuclear Disarmament, 9(sup1), 63-76. https://doi.org/10.1080/25751654.2026.2639044 Hunt, J. R. (2022). The nuclear club: How America and the world policed the atom from Hiroshima to Vietnam. Stanford University Press. https://doi.org/10.1515/9781503631724 International Atomic Energy Agency. (2025). Energy, electricity and nuclear power estimates for the period up to 2050. IAEA Reference Data Series No. 1. International Atomic Energy Agency. (2026). Implementation of the NPT safeguards agreement and relevant provisions of United Nations Security Council resolutions in the Islamic Republic of Iran (Report GOV/2026/33). IAEA Board of Governors. Kerr, P. K. (2026). Iran's nuclear program: Tehran's compliance with international obligations (CRS Report R40094). Congressional Research Service. Kristensen, H. M., & Korda, M. (2024). Nuclear notebook: United States nuclear weapons, 2024. Bulletin of the Atomic Scientists, 80(3), 182-208. Mukhatzhanova, G. (2022). The 10th NPT Review Conference: Why it was doomed and how it almost succeeded. Arms Control Today, 52(8), 6-13. Onderco, M. (2021). Networked nonproliferation: Making the NPT permanent. Stanford University Press. https://doi.org/10.11126/stanford/9781503628922.001.0001 Onderco, M., & Portela, C. (2023). External drivers of EU differentiated cooperation: How change in the nuclear nonproliferation regime affects member states' alignment. Contemporary Security Policy, 44(1), 150-175. https://doi.org/10.1080/13523260.2022.2146336 Onderco, M., Smetana, M., van der Meer, S., & Etienne, T. W. (2021). When do the Dutch want to join the nuclear ban treaty? Findings of a public opinion survey in the Netherlands. The Nonproliferation Review, 28(1-3), 149-163. https://doi.org/10.1080/10736700.2021.1978156 Smetana, M., & O'Mahoney, J. (2022). NPT as an antifragile system: How contestation improves the nonproliferation regime. Contemporary Security Policy, 43(1), 24-49. https://doi.org/10.1080/13523260.2021.1978761 Smetana, M., & Wunderlich, C. (2021). Nonuse of nuclear weapons in world politics: Toward the third generation of nuclear taboo research. International Studies Review, 23(3), 1072-1099. https://doi.org/10.1093/isr/viab002 Sokov, N. (2026). The end of New START: Short- and medium-term options. Vienna Center for Disarmament and Non-Proliferation. Stockholm International Peace Research Institute. (2025). SIPRI yearbook 2025: Armaments, disarmament and international security. Oxford University Press. Stockholm International Peace Research Institute. (2026). SIPRI yearbook 2026: Armaments, disarmament and international security. Oxford University Press. Tannenwald, N. (2024). The nuclear nonproliferation regime as a failed promise: Contestation and self-undermining dynamics in a liberal order. Global Studies Quarterly, 4(2), ksae025. https://doi.org/10.1093/isagsq/ksae025 Vicente, A., Sinovets, P., & Theron, J. (Eds.). (2023). Russia's war on Ukraine: The implications for the global nuclear order. Springer.

The Charter of the United Nations, signed at San Francisco on 26 June 1945 and in force from 24 October 1945, is the single most consequential multilateral treaty of the modern era. It created a permanent organisation with universal aspirations, outlawed the unilateral use of armed force between states, established a centralised system of #collective_security, and set out a framework for the peaceful settlement of disputes. This article examines the Charter as both a legal instrument and a living political institution. It traces the historical conditions that produced the text, analyses its structure and the legal hierarchy it creates, evaluates the operation of its principal organs, and assesses eight decades of practice in the fields of enforcement, peacekeeping, human rights, and decolonisation. The article argues that the Charter is best understood as a constitutional framework whose formal text has changed only marginally while its operational meaning has been transformed repeatedly through institutional practice. It also argues that the widening gap between the Charter's promise of impartial #collective_security and the selective behaviour of the Security Council constitutes the central legitimacy problem of contemporary #global_governance. Drawing on primary legal materials, the jurisprudence of the International Court of Justice, and recent United Nations policy instruments including Our Common Agenda (2021), A New Agenda for Peace (2023), and the Pact for the Future (2024), the study concludes that the Charter's normative core remains indispensable even where its enforcement machinery is blocked, and that realistic reform should focus on the working methods, accountability, and representativeness of the Security Council rather than on textual amendment, which the Charter itself has made extraordinarily difficult to achieve. Keywords: United Nations Charter; collective security; use of force; Security Council; international law; peaceful settlement of disputes; institutional reform. 1. Introduction Very few legal documents have shaped daily life across the whole planet as quietly and as thoroughly as the #UN_Charter. Its provisions determine when a state may lawfully send soldiers across a border, whether a trade embargo is legal, how a ceasefire is monitored, and which government is entitled to speak for a nation in the halls of world diplomacy. It is the reference point invoked by governments that go to war and by governments that condemn war. It has been cited by liberators and by occupiers, by great powers and by the smallest island states. Eight decades after it was drafted, it remains the closest thing humanity possesses to a written constitution for international society. The Charter was written by people who had watched the world burn twice within a single generation. Between 1914 and 1945, industrialised warfare killed tens of millions of human beings, destroyed the economies of entire continents, and culminated in the systematic murder of civilian populations and the first use of nuclear weapons against cities. The delegates who assembled in San Francisco in the spring of 1945 were not idealists writing in an abstract seminar room. They were survivors of catastrophe attempting to design machinery that would make catastrophe less likely. The opening words of the Preamble, in which the peoples of the United Nations declare their determination to save succeeding generations from the scourge of war, are not decorative rhetoric. They are a statement of purpose written by people who had counted the dead. The central claim advanced here is that the Charter should be understood as a constitutional framework rather than as an ordinary contract between states. Ordinary treaties bind their parties and no one else. The Charter, by contrast, claims priority over conflicting agreements, purports to bind non-members in matters necessary for the maintenance of international peace, creates organs capable of issuing decisions binding on all members, and has generated a body of secondary law and institutional practice that no single member state can unilaterally undo. That constitutional quality is the source both of its authority and of its deepest problems, because a constitution that cannot be amended in practice and whose enforcement organ is structurally capable of paralysis will inevitably attract charges of #legitimacy failure. The article develops that argument across the sections that follow. 2. Methodology and Scope This study adopts a doctrinal and historical institutional approach. Doctrinal legal analysis is used to interpret the text of the Charter itself, applying the interpretive rules reflected in the Vienna Convention on the Law of Treaties, which requires that a treaty be interpreted in good faith according to the ordinary meaning of its terms in their context and in the light of its object and purpose, with subsequent practice taken into account. Because the Charter is the constituent instrument of an international organisation, its interpretation is also informed by the doctrine of implied powers and by the evolutionary practice of the organs themselves, a point confirmed by the International Court of Justice in the advisory opinions on Reparation for Injuries in 1949 and Certain Expenses in 1962. 3. Historical Origins: From the Concert of Europe to San Francisco 3.1 The long search for order The idea that states might submit their quarrels to common rules is far older than the United Nations. The Peace of Westphalia in 1648 is conventionally treated as the origin of the modern system of sovereign states. The Concert of Europe that followed the Napoleonic Wars established the habit of consultation among great powers, and the Hague Peace Conferences of 1899 and 1907 produced the first serious attempts to codify the laws of war and to build machinery for arbitration. None of these arrangements outlawed war. They regulated it. War remained a legitimate instrument of national policy, and the right to wage it was regarded as an attribute of #sovereign_equality itself. 3.2 The League of Nations and its failure The First World War destroyed that assumption. The Covenant of the League of Nations, adopted in 1919 as part of the Treaty of Versailles, was the first serious attempt to build a permanent organisation dedicated to preventing war. It required members to submit disputes to arbitration, judicial settlement, or Council inquiry, and imposed a waiting period before resort to hostilities. It created an Assembly, a Council, a Secretariat, and, alongside it, the Permanent Court of International Justice. The League failed, and understanding why is essential to understanding the choices made in 1945. Four weaknesses stand out. The Covenant did not prohibit war outright; it merely delayed it and made certain kinds of war unlawful while leaving others permissible, and the Kellogg-Briand Pact of 1928, which renounced war as an instrument of national policy, contained no enforcement machinery whatsoever. Decision rules required unanimity in most cases, so any member could paralyse action. The great powers were never all inside the organisation at the same time; the United States never joined, Germany and Japan withdrew in the 1930s, and the Soviet Union was expelled in 1939. Above all, the League had no reliable means of enforcement. When Japan invaded Manchuria in 1931 and when Italy invaded Ethiopia in 1935, the League produced reports and half hearted sanctions, and the aggressors simply walked away. The lesson drawn by the architects of the post war order was blunt: an organisation for peace must include the strongest states, must be able to act without requiring unanimity of the whole membership, and must possess coercive teeth. That lesson explains almost everything distinctive about the #UN_Charter. 3.3 The wartime diplomacy that produced the text The drafting process unfolded across a series of wartime meetings. The Atlantic Charter of August 1941, agreed between Franklin Roosevelt and Winston Churchill aboard warships off Newfoundland, set out common principles including the renunciation of territorial aggrandisement, respect for the right of peoples to choose their form of government, and the establishment of a wider and permanent system of general security. The Declaration by United Nations of 1 January 1942, signed by twenty six states fighting the Axis powers, gave the alliance its name; the term United Nations was originally the name of a wartime coalition, not of a peace organisation. The Moscow Declaration of October 1943, issued by the foreign ministers of the United States, the United Kingdom, the Soviet Union, and China, committed the four powers to establishing a general international organisation based on the principle of the sovereign equality of all peace loving states. The Dumbarton Oaks Conversations, held in Washington between August and October 1944, produced the first detailed blueprint. It is here that the essential architecture appears: a General Assembly of all members with broad but recommendatory powers, a Security Council with primary responsibility for peace and security and with binding powers, an Economic and Social Council, an International Court of Justice, and a Secretariat. Dumbarton Oaks left one question unresolved, and it was the most explosive one: the voting procedure in the Security Council. That question was settled at the Yalta Conference in February 1945, where the three leaders agreed on what became known as the Yalta formula, under which the permanent members would each hold a #veto over substantive decisions but would abstain from voting in disputes to which they were parties in matters of pacific settlement. 3.4 San Francisco, 1945 The United Nations Conference on International Organization opened in San Francisco on 25 April 1945 and closed on 26 June, when delegates of fifty states signed the Charter. Poland, unable to send a delegation in time because of disputes about the composition of its government, signed later and is counted among the fifty one original members. The Charter entered into force on 24 October 1945, once China, France, the Soviet Union, the United Kingdom, the United States, and a majority of the other signatories had deposited their instruments of ratification. That date is observed annually as United Nations Day. San Francisco was not a rubber stamp. The smaller states fought hard against the great power #veto, and the debates over it were bitter. Delegations from Latin America, Australia, New Zealand, and the smaller European states argued that the veto entrenched inequality in an organisation founded on sovereign equality. The Australian delegation in particular pressed for limiting the veto to enforcement action. The great powers refused to yield, and their position was, in the end, decisive for a simple reason: they made clear that there would be no organisation at all if the veto were removed. The smaller states accepted the bargain because an imperfect organisation containing the great powers seemed preferable to a perfect one that they would boycott. That trade off, between effectiveness and equality, is the original sin of the United Nations and the source of most of the criticism it has attracted ever since. Other San Francisco contributions were more constructive. The delegations of the smaller powers strengthened the human rights language, expanded the powers of the General Assembly, added the provisions on non self governing territories that became Chapter XI, and secured a more prominent role for regional arrangements. Non governmental organisations, present in significant numbers for the first time at a diplomatic conference of this kind, pressed successfully for the inclusion of #human_rights language in the Preamble and Article 1, and for the creation of the Commission on Human Rights under Article 68. 4. The Legal Nature and Architecture of the Charter 4.1 Structure of the instrument The Charter consists of a Preamble and one hundred and eleven articles distributed across nineteen chapters, with the Statute of the International Court of Justice annexed to it and forming an integral part of it. The chapters proceed in a logical sequence: purposes and principles, membership, organs, then a chapter devoted to each principal organ, then the substantive chapters on pacific settlement, enforcement, regional arrangements, economic and social cooperation, non self governing territories, the trusteeship system, the Secretariat, and finally miscellaneous provisions, transitional arrangements, amendment, and ratification. The drafting is deliberately terse. Article 2(4), which contains the single most important rule of modern international law, is a single sentence. Article 51, which preserves the right of #self_defence, is a single paragraph. This economy of language was intentional. The drafters knew they were writing a document that would have to accommodate circumstances they could not foresee, and they preferred general standards to detailed rules. The consequence is that the meaning of the Charter has been worked out over eight decades through practice, interpretation, and argument rather than being fixed at the moment of signature. 4.2 The Charter as a constituent instrument A constituent instrument is a treaty that creates an organisation and confers powers upon it. Such treaties are interpreted differently from ordinary bilateral agreements. The International Court of Justice held in Reparation for Injuries in 1949 that the United Nations possesses international legal personality distinct from that of its members and may bring international claims, because such capacities are necessarily implied as essential to the performance of its duties. In Certain Expenses in 1962 the Court held that when the organisation takes action appropriate for the fulfilment of one of its stated purposes, the presumption is that such action is not ultra vires. These holdings gave the Charter an expansive, purposive character. The organisation may do what is reasonably necessary to achieve its objects, even where no express power is stated. 4.3 Article 103 and the hierarchy of obligations Article 103 provides that in the event of a conflict between the obligations of members under the Charter and their obligations under any other international agreement, their Charter obligations prevail. This is a rule of extraordinary reach. It means, in effect, that the Charter sits above the ordinary body of #treaty_law. Combined with Article 25, under which members agree to accept and carry out the decisions of the Security Council, it means that a binding Council decision can override conflicting treaty commitments. The consequences of that supremacy have been tested in practice, most visibly in litigation concerning the implementation of Council sanctions within domestic and regional legal orders, where courts have insisted that respect for fundamental rights cannot simply be displaced by an assertion of Charter supremacy. The unresolved tension between Charter supremacy and the protection of individual rights remains one of the most difficult problems in contemporary #international_law. Article 103 does not, however, place the Charter above peremptory norms of general international law, known as #jus_cogens, such as the prohibitions of genocide, slavery, torture, and aggression. The dominant scholarly view, supported by the reasoning of several international tribunals, is that no organ of the United Nations may lawfully require states to violate a peremptory norm. 4.4 Membership and universality Article 4 provides that membership is open to peace loving states which accept the obligations of the Charter and which, in the judgment of the organisation, are able and willing to carry them out. Admission is effected by the General Assembly upon the recommendation of the Security Council, which means that a permanent member can block admission. During the Cold War this power was used repeatedly, producing long queues of applicants and eventually package deals in which blocs of states were admitted together. Membership grew from fifty one in 1945 to one hundred and ninety three today, the great expansion coming with #decolonization in the 1950s and 1960s and again with the dissolution of the Soviet Union and Yugoslavia in the early 1990s. 5. Purposes and Principles: Articles 1 and 2 5.1 The purposes Article 1 sets out four purposes. The first is to maintain international peace and security, and to that end to take effective collective measures for the prevention and removal of threats to the peace and for the suppression of acts of aggression, and to bring about by peaceful means, and in conformity with the principles of justice and international law, adjustment or settlement of international disputes. The second is to develop friendly relations among nations based on respect for the principle of equal rights and #self_determination of peoples. The third is to achieve international cooperation in solving problems of an economic, social, cultural, or humanitarian character, and in promoting and encouraging respect for human rights and fundamental freedoms for all without distinction as to race, sex, language, or religion. The fourth is to be a centre for harmonising the actions of nations in the attainment of these common ends. Three observations follow. First, peace is listed first and is given the most operational language, which reflects the priorities of 1945. Second, the reference to justice and international law appears in the clause on peaceful settlement but is conspicuously absent from the clause on collective measures, a drafting choice that has been read as preserving the political discretion of the Security Council. Third, the inclusion of human rights and self determination in a security treaty was not obvious in 1945 and has proved enormously consequential, providing the legal foothold from which the entire human rights machinery of the organisation subsequently grew. 5.2 The principles Article 2 lists seven principles that govern the organisation and its members. Article 2(1) establishes the #sovereign_equality of all members. This is the foundational premise of the system and it is, on its face, contradicted by the composition of the Security Council. The tension is legal as well as rhetorical, and there is no wholly satisfying doctrinal resolution of it. The honest answer is that the Charter contains a compromise between the formal equality of states and the material inequality of power. Article 2(2) requires members to fulfil in good faith the obligations they have assumed under the Charter. Good faith is not an empty formula; it is a general principle of law that shapes the interpretation and performance of every other obligation. Article 2(3) requires all members to settle their international disputes by peaceful means in such a manner that international peace and security, and justice, are not endangered. This is the positive counterpart to the prohibition of force. Article 2(4) is the heart of the Charter. Members shall refrain in their international relations from the threat or use of force against the territorial integrity or political independence of any state, or in any other manner inconsistent with the purposes of the United Nations. This provision does something the Covenant of the League never did and the Kellogg-Briand Pact never enforced: it prohibits not merely war in a technical sense but the threat or #use_of_force as such. The International Court of Justice held in the Nicaragua case in 1986 that the prohibition of the use of force also exists as a rule of customary international law binding independently of the Charter, and it is widely regarded as a peremptory norm from which no derogation is permitted. Article 2(5) requires members to give the organisation every assistance in any action it takes in accordance with the Charter and to refrain from assisting any state against which preventive or enforcement action is being taken. Article 2(6) is remarkable. It provides that the organisation shall ensure that states which are not members act in accordance with these principles so far as may be necessary for the maintenance of international peace and security. A treaty ordinarily cannot bind third parties. Article 2(6) asserts that the Charter's core rules apply to everyone, and it is one of the strongest textual indications that the drafters regarded the document as constitutional in character. Article 2(7) provides that nothing in the Charter authorises the organisation to intervene in matters which are essentially within the domestic jurisdiction of any state, but it adds a crucial exception: this principle shall not prejudice the application of enforcement measures under Chapter VII. The domestic jurisdiction clause has been the great battleground of the organisation's history. Governments accused of atrocities routinely invoke it. The practice of the organisation has steadily narrowed it, on the reasoning that matters governed by international obligations, and in particular grave and systematic violations of #human_rights that threaten international peace, are by definition no longer essentially domestic. 6. The Principal Organs Article 7 establishes six principal organs: the General Assembly, the Security Council, the Economic and Social Council, the Trusteeship Council, the International Court of Justice, and the Secretariat. 6.1 The General Assembly The #General_Assembly is the plenary body in which all one hundred and ninety three members sit and each has one vote, in accordance with Article 18. It is the closest thing the world has to a parliament of states, and its egalitarian voting rule is the purest expression of sovereign equality in the Charter. Its powers, however, are principally recommendatory. Article 10 authorises it to discuss any question within the scope of the Charter and to make recommendations. Article 11 permits it to consider general principles of cooperation in the maintenance of international peace. Article 12 originally restrained it from making recommendations on a dispute while the Security Council is exercising its functions in respect of that dispute, though the practice of both organs has relaxed this restriction considerably, and the Court noted in its 2004 advisory opinion on the Wall in the Occupied Palestinian Territory that the two organs increasingly deal with the same matters in parallel. The Assembly does hold two genuinely binding powers. Under Article 17 it approves the budget and apportions expenses among the members, and the Court confirmed in Certain Expenses that this power is real and that members are legally obliged to pay. Under Article 4 it admits new members. It also elects the non permanent members of the Security Council, elects the members of the Economic and Social Council, appoints the Secretary General on the recommendation of the Council, and, together with the Council, elects the judges of the International Court of Justice. The Assembly's most creative act was the #Uniting_for_Peace resolution of 3 November 1950. Adopted during the deadlock produced by the Korean conflict, it declared that if the Security Council, because of the lack of unanimity of its permanent members, fails to exercise its primary responsibility for the maintenance of international peace and security in a case where there appears to be a threat to the peace, breach of the peace, or act of aggression, the General Assembly shall consider the matter immediately with a view to making appropriate recommendations to members for collective measures. This mechanism, under which an emergency special session can be convened, has been invoked repeatedly, most recently and most consequentially in relation to the invasion of Ukraine in 2022. Its legal status remains contested. It does not give the Assembly the power to authorise the use of force in a binding manner, but it does give the Assembly the power to speak with the authority of the overwhelming majority of states when the Council is paralysed, and that political function has proved durable. 6.2 The Security Council The #Security_Council is the organ on which the Charter's system of enforcement depends. Article 24 confers on it primary responsibility for the maintenance of international peace and security and states that in carrying out its duties under that responsibility it acts on behalf of the members. Article 25 provides that members agree to accept and carry out its decisions in accordance with the Charter. It is the only organ that can issue decisions binding on all member states. Its composition is set out in Article 23. It consists of fifteen members, of which five are permanent: China, France, the Russian Federation, the United Kingdom, and the United States. The remaining ten are elected by the General Assembly for two year terms with due regard to the contribution of members to the maintenance of peace and to equitable geographical distribution. Retiring members are not eligible for immediate re election. The original Council had eleven members; an amendment adopted in 1963 and in force from 1965 enlarged it to fifteen. Voting is governed by Article 27. Each member has one vote. Procedural matters are decided by an affirmative vote of nine members. All other matters require an affirmative vote of nine members including the concurring votes of the permanent members, with the proviso that in decisions under Chapter VI and under Article 52(3) a party to a dispute shall abstain from voting. The literal text appears to require the positive concurrence of all five permanent members, which would mean that an abstention defeats a resolution. In practice, from the earliest years, the Council treated abstention and absence as not blocking adoption, and the International Court of Justice confirmed in its 1971 advisory opinion on Namibia that this consistent practice had been generally accepted by members and constituted a valid interpretation of the Article. This is perhaps the clearest example in the history of the organisation of the text of the Charter being modified by practice without formal amendment. The #veto is the defining feature and the defining problem of the Council. It was designed to ensure that the organisation would never be dragged into enforcement action against a great power, on the theory that such action would mean world war rather than police action. Whatever one thinks of that logic, its consequence is structural: whenever a permanent member or its close ally is implicated in a conflict, the Council's enforcement machinery cannot operate. The Cold War produced hundreds of vetoes and near total paralysis. The 1990s produced a burst of unprecedented activity. The period since roughly 2011 has seen a return to blockage, most acutely over Syria, Ukraine, and Gaza. 6.3 The Economic and Social Council The Economic and Social Council, established under Chapter X, coordinates the economic, social, cultural, educational, and health work of the organisation and of the specialised agencies. It has fifty four members elected by the General Assembly, following two enlargements by amendment. Under Article 62 it may make or initiate studies and make recommendations; under Article 63 it concludes relationship agreements with the specialised agencies; and under Article 71 it may make arrangements for consultation with non governmental organisations, a provision that has become the principal gateway through which civil society engages the United Nations system. The Council has long been criticised as unwieldy, and much of the substantive work it was intended to coordinate has migrated to other bodies, including the committees of the General Assembly, the specialised agencies, the international financial institutions, and the forum that reviews progress on the Sustainable Development Goals. 6.4 The Trusteeship Council The Trusteeship Council administered the international trusteeship system established under Chapters XII and XIII, which applied to territories detached from enemy states, former League mandates, and territories voluntarily placed under the system. Its purpose, stated in Article 76, included the promotion of the progressive development of the trust territories towards self government or independence. It completed its work when Palau, the last remaining trust territory, achieved independence, and the Council suspended operation on 1 November 1994. Its formal abolition would require a Charter amendment, which is why it survives on paper. Proposals to repurpose it, for example as a guardian of the global commons or of the interests of future generations, have been floated repeatedly, including in the recent reform discussions, but none has yet been adopted. 6.5 The International Court of Justice The #ICJ is the principal judicial organ of the United Nations under Article 92, and its Statute is annexed to the Charter. All members are ipso facto parties to the Statute. In contentious cases the Court settles legal disputes between states, but only where the states concerned have consented to its jurisdiction, whether by special agreement, by a compromissory clause in a treaty, or by a declaration under the optional clause. In advisory proceedings it gives opinions on legal questions at the request of the General Assembly, the Security Council, or other authorised bodies. The Court's weakness is enforcement. Article 94(1) obliges each member to comply with the Court's decision in any case to which it is a party, and Article 94(2) provides that if a party fails to perform, the other party may have recourse to the Security Council. Since the Council is subject to the veto, a permanent member that ignores a judgment cannot be coerced through this route. It is the same structural problem that afflicts the enforcement chapter, appearing in judicial dress. 6.6 The Secretariat and the Secretary General The Secretariat is the permanent staff of the organisation, headed by the #Secretary_General, who is described in Article 97 as the chief administrative officer. Article 100 requires the Secretary General and the staff to seek and receive instructions from no government and to refrain from any action which might reflect on their position as international officials responsible only to the organisation; members undertake to respect the exclusively international character of these responsibilities. Article 99 gives the Secretary General a political power of great potential significance: the authority to bring to the attention of the Security Council any matter which in his opinion may threaten the maintenance of international peace and security. This is the only express provision of the Charter that grants an individual the right to place a matter before the Council. It has been used sparingly, precisely because its power derives from its rarity. Its most recent formal invocation came in December 2023, when the Secretary General wrote to the Council concerning the humanitarian catastrophe in Gaza. The office has always oscillated between the roles suggested by its two words, secretary and general. Some incumbents have interpreted the position as primarily administrative; others have used the moral authority of the office and the discretion inherent in good offices, mediation, and public advocacy to shape outcomes. The method of selection, in which the Council recommends a candidate and the Assembly appoints, means that no one can hold the office without the acquiescence of all five permanent members, and this structural fact places a permanent constraint on how far any Secretary General can go in confronting the powerful. 7. Chapter VI: The Pacific Settlement of Disputes #Chapter_VI, comprising Articles 33 to 38, deals with the peaceful settlement of disputes whose continuance is likely to endanger the maintenance of international peace and security. Article 33(1) requires the parties to such a dispute first of all to seek a solution by negotiation, enquiry, mediation, conciliation, arbitration, judicial settlement, resort to regional agencies or arrangements, or other peaceful means of their own choice. This list is not exhaustive and the parties are free to choose among the methods; the obligation is one of means, not of result. Article 33(2) empowers the Security Council, when it deems necessary, to call upon the parties to settle their dispute by such means. Article 34 authorises the Council to investigate any dispute or any situation which might lead to international friction. Article 35 allows any member, and in certain circumstances a non member, to bring a dispute to the attention of the Council or the Assembly. Article 36 permits the Council to recommend appropriate procedures or methods of adjustment, and instructs it to take into consideration that legal disputes should as a general rule be referred by the parties to the International Court of Justice. Article 37 provides that if the parties fail to settle a dispute by the means indicated in Article 33 they shall refer it to the Council, which may recommend such terms of settlement as it may consider appropriate. Article 38 allows the Council to make recommendations on any dispute at the request of all the parties. The vocabulary of Chapter VI is the vocabulary of recommendation. The Council calls upon, investigates, and recommends. It does not, under this chapter, command. This textual restraint reflects a deliberate design principle: coercion is reserved for Chapter VII and for situations that have already crossed the threshold of a threat to the peace. In practice, Chapter VI has proved less important than its drafters expected, partly because states rarely bring their own disputes to the Council voluntarily, and partly because the Council's most effective tools of #conflict_resolution have developed outside the formal chapter structure altogether. The good offices of the Secretary General, special envoys, mediation support, and quiet diplomacy have done far more preventive work than the formal machinery of Articles 33 to 38. The principle of #peaceful_settlement nonetheless remains the legal bedrock of the system, since it is the affirmative obligation that gives meaning to the negative prohibition in Article 2(4). 8. Chapter VII: Collective Security and the Use of Force 8.1 The design #Chapter_VII, comprising Articles 39 to 51, is the enforcement heart of the Charter and the element that most sharply distinguishes it from the League Covenant. Article 39 is the gateway. The Security Council shall determine the existence of any threat to the peace, breach of the peace, or act of aggression, and shall make recommendations, or decide what measures shall be taken in accordance with Articles 41 and 42, to maintain or restore international peace and security. Everything under Chapter VII depends on an Article 39 determination. Crucially, the determination is a matter for the Council's political judgment. There is no definition of a threat to the peace in the Charter, and the Council has never accepted that its determinations are subject to judicial review. Article 40 permits the Council to call upon the parties to comply with provisional measures, such as a ceasefire or a withdrawal, without prejudice to the rights of the parties. Article 41 empowers the Council to decide what measures not involving the use of armed force are to be employed to give effect to its decisions, including complete or partial interruption of economic relations and of communications and the severance of diplomatic relations. This is the legal basis for #sanctions. Article 42 provides that should the Council consider that measures under Article 41 would be inadequate or have proved inadequate, it may take such action by air, sea, or land forces as may be necessary to maintain or restore international peace and security. Articles 43 to 47 describe how the forces required for Article 42 action were to be provided. Article 43 envisaged special agreements under which members would make armed forces, assistance, and facilities available to the Council. Article 47 established a Military Staff Committee, composed of the chiefs of staff of the permanent members, to advise and assist the Council on military requirements and to be responsible for the strategic direction of any armed forces placed at its disposal. 8.2 The gap between design and practice The Article 43 agreements were never concluded. Cold War distrust made it inconceivable that the Soviet Union and the United States would place national forces under a joint command structure. The Military Staff Committee still exists on paper and meets as a formality, but it has never performed its intended function. The result is that the Charter's most ambitious design feature, a standing enforcement capability under international command, has never existed. The organisation responded with improvisation, and the improvisation has been remarkably durable. Two devices in particular emerged. The first is the authorisation technique. Rather than commanding a United Nations army it does not possess, the Council authorises willing member states, or coalitions, or regional organisations, to use force. The classic formula, first used in Resolution 678 of 29 November 1990 concerning the Iraqi occupation of Kuwait, is that the Council authorises member states cooperating with the government concerned to use all necessary means to achieve a stated objective. This delegation model preserves the Council's legal monopoly on the authorisation of force while outsourcing the actual fighting to those with the capability and the will. It has been used in Somalia in 1992, in Haiti, in Bosnia, in East Timor, in Sierra Leone, in Cote d'Ivoire, and in Libya in 2011. The delegation model raises serious problems of accountability. Once force is authorised, the Council has limited practical ability to control how the mandate is interpreted or when it ends. The Libyan intervention of 2011, authorised by Resolution 1973 to protect civilians and to enforce a no fly zone, is the paradigm case: several members subsequently argued that the operation exceeded the mandate by pursuing regime change, and the resulting bitterness has poisoned Council politics ever since, contributing directly to the deadlock over Syria. The second device is the expansion of the meaning of a threat to the peace. In 1945 the phrase would have been understood to mean an inter state military threat. The Council has since determined that internal armed conflicts, humanitarian catastrophes, mass displacement, the overthrow of democratically elected governments, terrorism, the proliferation of weapons of mass destruction, and, in some instances, epidemic disease can constitute threats to international peace and security. This expansion has allowed the Council to act in situations the drafters never contemplated, and it has simultaneously eroded the boundary set by Article 2(7). It is a striking example of a constitutional text acquiring new meaning without a single word being changed. 8.3 Sanctions Sanctions under Article 41 became the Council's workhorse after the Cold War. The comprehensive economic sanctions imposed on Iraq in the 1990s produced severe humanitarian consequences for the civilian population and provoked an intense debate about the ethics of collective punishment. The response was a shift towards targeted sanctions aimed at individuals and entities rather than whole economies: asset freezes, travel bans, arms embargoes, and commodity restrictions on specific actors. Targeted sanctions solved one problem and created another. Listing an individual and freezing that person's assets indefinitely, on the basis of confidential intelligence and without any right of appeal, raises acute questions of due process. The Council responded by creating a focal point for delisting requests and, for the counter terrorism regime, an Office of the Ombudsperson with authority to review delisting petitions. Whether these mechanisms satisfy the requirements of fair procedure remains contested, and the litigation on this question in domestic and regional courts illustrates the practical limits of Article 103 supremacy when it collides with entrenched constitutional rights. 8.4 Article 51 and self defence Article 51 provides that nothing in the Charter shall impair the inherent right of individual or collective self defence if an armed attack occurs against a member, until the Security Council has taken measures necessary to maintain international peace and security, and that measures taken in the exercise of this right shall be immediately reported to the Council. Four elements have generated persistent controversy. The first is the trigger. The text requires that an armed attack occur, and the International Court of Justice held in Nicaragua that not every use of force amounts to an armed attack; the concept covers the most grave forms of the use of force, and lesser uses of force, while unlawful, do not open the door to forcible self defence. This creates a gap between the prohibition in Article 2(4) and the exception in Article 51, and states have persistently sought to narrow it. The second is timing. The traditional formulation, drawn from nineteenth century practice, permits anticipatory action where the necessity of self defence is instant, overwhelming, and leaves no choice of means and no moment for deliberation. A broader doctrine of preventive self defence against distant and speculative threats, asserted prominently in the early 2000s, has not been accepted by the general body of states and finds no support in the text. The third is the nature of the attacker. The text does not say that the armed attack must come from a state. After 2001, Council resolutions recognising the right of self defence in response to terrorist attacks, and the subsequent practice of states conducting operations against non state armed groups on the territory of states unable or unwilling to suppress them, generated an intense doctrinal debate. The unable or unwilling formula is asserted by some states and rejected by others; it cannot yet be described as settled law. The fourth is the limits. Self defence, even when lawfully triggered, is governed by the customary requirements of necessity and proportionality and by international humanitarian law once hostilities begin. The reporting requirement, and the words until the Security Council has taken measures, confirm that the right is meant to be a temporary bridge until #collective_security takes over, not a standing licence. 8.5 Humanitarian intervention and the Responsibility to Protect The Charter permits the use of force in only two circumstances: self defence under Article 51 and authorisation by the Security Council under Chapter VII. There is no third exception for humanitarian purposes in the text. This produced the great dilemma of the 1990s. The failure to prevent the genocide in Rwanda in 1994, and the failure to protect the safe area of Srebrenica in 1995, demonstrated the moral cost of inaction. The NATO air campaign over Kosovo in 1999, undertaken without Council authorisation because of the certainty of a veto, demonstrated the legal cost of action outside the Charter framework. An independent commission famously described that campaign as illegal but legitimate, a formula that captures the dilemma precisely and resolves it not at all. The attempt to escape the dilemma produced the doctrine of the #Responsibility_to_Protect, developed by an international commission in 2001 and endorsed in modified form by the General Assembly in the World Summit Outcome Document of 2005. It has three pillars: each state has the responsibility to protect its own population from genocide, war crimes, ethnic cleansing, and crimes against humanity; the international community has a responsibility to assist states in meeting that obligation; and if a state manifestly fails, the international community is prepared to take collective action through the Security Council in accordance with the Charter. The critical point, often lost in political debate, is that the 2005 formulation created no new legal exception to Article 2(4). It routed the responsibility through the Security Council. It therefore does not solve the problem of the veto; it restates it in more urgent language. After Libya in 2011 the doctrine's political capital was substantially depleted, though its first two pillars, concerning prevention and assistance, continue to shape the practice of the organisation. 9. Peacekeeping: The Chapter That Does Not Exist 9.1 Origins of an improvisation The word #peacekeeping appears nowhere in the Charter. The activity that has become the most visible face of the United Nations in the world was invented because the enforcement machinery of Chapter VII could not function. Dag Hammarskjold, the second Secretary General, described it with characteristic wit as belonging to Chapter Six and a Half: more than the pacific settlement of disputes envisaged in Chapter VI, less than the coercive enforcement envisaged in Chapter VII. The first armed #peacekeeping operation in the modern sense was the United Nations Emergency Force, established by the General Assembly in November 1956 to supervise the withdrawal of forces from Egyptian territory during the Suez crisis. The circumstances are instructive. Two permanent members of the Security Council, the United Kingdom and France, were parties to the conflict and vetoed Council action. The Assembly, acting through the Uniting for Peace procedure, filled the vacuum. The force that emerged was built on three principles that became the classical doctrine of peacekeeping: consent of the host state, impartiality between the parties, and the non use of force except in self defence. The legal foundation of these operations was contested from the beginning. Several states, including the Soviet Union and France, refused to pay for the Emergency Force and for the operation in the Congo, arguing that the General Assembly had no authority to establish such missions and that the expenses were therefore not expenses of the organisation within the meaning of Article 17(2). The International Court of Justice, in the Certain Expenses advisory opinion, rejected that argument and held that expenditures authorised by the Assembly for the maintenance of international peace and security, where the action taken was appropriate to the fulfilment of a stated purpose of the organisation, were expenses of the organisation and had to be borne by the members. The financial crisis that followed nevertheless nearly destroyed the organisation and left a lasting lesson: legal authority without political consent is fragile. 9.2 From classical to multidimensional Classical peacekeeping was essentially a military task: interposing lightly armed soldiers between the armies of consenting states along an agreed line, monitoring a ceasefire, and reporting violations. The end of the Cold War transformed both the number and the nature of these missions. Between 1988 and 1994 the Council authorised more operations than in the preceding four decades combined, and the tasks assigned to them expanded far beyond ceasefire observation to include the disarmament of combatants, the supervision of elections, the reform of police and judicial institutions, the protection of human rights, the facilitation of humanitarian relief, and, in some cases, the temporary administration of territory. This expansion brought a crisis. Missions deployed on classical principles, with the consent of parties who had no intention of keeping their promises, found themselves present at atrocities they were neither mandated nor equipped to prevent. The independent inquiries into the failures in Rwanda and in Srebrenica reached convergent conclusions: mandates were unclear, resources were inadequate, the doctrine of impartiality had been misapplied to mean neutrality between victim and perpetrator, and member states had been unwilling to accept the risks that effective action would have required. The response was doctrinal reform. The Brahimi Report of 2000 argued that mandates must be clear, credible, and achievable, that impartiality means adherence to the principles of the Charter rather than equal treatment of all parties in all cases, and that peacekeepers who witness atrocities must be presumed to be authorised to stop them. A further review in 2015 emphasised the primacy of politics, arguing that peace operations are not a substitute for a political strategy and cannot succeed in the absence of a genuine peace to keep. 9.3 Protection of civilians and the robust turn Contemporary mandates routinely authorise missions, acting under Chapter VII, to use all necessary means to protect civilians under imminent threat of physical violence within their areas of deployment and within their capabilities. This robust turn has blurred the line between peacekeeping and enforcement, and in theatres where missions have been authorised to conduct offensive operations against named armed groups it raises difficult questions about whether their personnel retain protected status under international humanitarian law. The practical difficulties are severe. Protection mandates create expectations that missions frequently cannot meet, and the gap between the promise of protection and its delivery corrodes the #legitimacy of the organisation among the very people it is meant to serve. Missions are chronically under resourced, troop contributing countries impose national caveats on what their contingents may be asked to do, and host governments have increasingly restricted missions or demanded their withdrawal. The closure of the mission in Mali at the request of the host authorities and the drawdown of the mission in the Democratic Republic of the Congo illustrate a broader phenomenon: consent, the first principle of peacekeeping, is being withdrawn in an environment where host states have alternative security partners. 9.4 Peacebuilding The recognition that ending violence is not the same as building peace led to the creation in 2005 of a peacebuilding architecture: a Peacebuilding Commission as an intergovernmental advisory body, a Peacebuilding Fund, and a support office in the Secretariat. The concept of #peacebuilding rests on the well documented finding that a large proportion of civil wars recur within a decade of a settlement, and that the risk of recurrence is shaped by the quality of institutions, the inclusiveness of political settlements, and the presence of sustained international engagement after the cameras leave. The concept of sustaining peace, endorsed by parallel resolutions of the Assembly and the Council in 2016, reframes the task as continuous rather than post conflict, encompassing prevention as much as reconstruction. 10. Chapter VIII: Regional Arrangements #Chapter_VIII, comprising Articles 52 to 54, addresses the role of regional arrangements and agencies. Article 52 recognises the existence of such arrangements for dealing with matters relating to the maintenance of international peace and security as are appropriate for regional action, provided their activities are consistent with the purposes and principles of the organisation, and it encourages the pacific settlement of local disputes through them. Article 53 contains the critical restriction: no enforcement action shall be taken under regional arrangements or by regional agencies without the authorisation of the Security Council. Article 54 requires that the Council be kept fully informed of activities undertaken or in contemplation. The design logic is subsidiarity combined with central control. Regional bodies may act first, because they are closer to the conflict and often better informed, but they may not coerce without a mandate from the Council. This preserves the Council's monopoly over the lawful authorisation of force. Practice has been uneven. The African Union has developed the most elaborate regional peace and security architecture, including a Peace and Security Council and provisions in its Constitutive Act permitting intervention in grave circumstances, and it has deployed substantial missions, sometimes with subsequent Council authorisation and sometimes with contested legal status. The financing of African led peace support operations through assessed contributions has long been contentious, and its partial resolution through recent Council decisions is among the more significant institutional developments of recent years. Sub regional organisations in West Africa have intervened in member states, sometimes acting first and seeking authorisation afterwards. Collective defence alliances such as NATO rest legally on Article 51 rather than Chapter VIII, since collective self defence requires no Council authorisation, but NATO has also acted as an implementing agent under Council mandates. 11. Self Determination, Decolonisation, and the Transformation of Membership When the Charter was signed, a large part of the world's population lived under colonial rule, and several of the founding members were the metropolitan powers of extensive empires. The Charter did not abolish colonialism. It contained two chapters that dealt with dependent territories in a manner that now appears cautious to the point of complicity, but which nonetheless planted the seeds of the empires' dissolution. Chapter XI, the Declaration Regarding Non Self Governing Territories, imposed on administering powers the obligation to recognise the principle that the interests of the inhabitants of such territories are paramount, to accept as a sacred trust the obligation to promote their well being, and to develop self government, taking due account of the political aspirations of the peoples. Article 73(e) required administering powers to transmit information about conditions in the territories to the Secretary General. That reporting obligation, apparently technical, proved to be the thin end of a very substantial wedge, because it created an international forum in which colonial administration became a matter of standing international scrutiny. Chapters XII and XIII established the trusteeship system for a limited category of territories, with the explicit objective in Article 76 of promoting progressive development towards self government or independence. The transformation came from the #General_Assembly. In 1960 the Assembly adopted the Declaration on the Granting of Independence to Colonial Countries and Peoples, which declared that the subjection of peoples to alien subjugation, domination, and exploitation constitutes a denial of fundamental human rights and is contrary to the Charter, and that all peoples have the right to #self_determination. Two years later it established the special committee that continues to monitor the remaining non self governing territories. Within two decades the membership of the organisation was transformed, and with it the balance of voting power in the Assembly, which became a forum in which the newly independent majority pressed for a new international economic order, for the condemnation of apartheid, and for the recognition of permanent sovereignty over natural resources. 12. From Peace to Human Security: The Widening of the Charter's Agenda 12.1 Human rights The Charter mentions #human_rights seven times but defines them nowhere. It contains no bill of rights and no enforcement mechanism. What it did was to establish, for the first time in a general treaty of universal aspiration, that the treatment by a government of the people under its control is a matter of legitimate international concern. Article 55 commits the organisation to promote universal respect for, and observance of, human rights and fundamental freedoms for all without distinction, and Article 56 commits members to take joint and separate action in cooperation with the organisation for the achievement of those purposes. From that foundation grew an enormous edifice: the #Universal_Declaration_of_Human_Rights of 1948, the two Covenants of 1966, the specialised conventions on racial discrimination, on discrimination against women, against torture, on the rights of the child, and on the rights of persons with disabilities, the treaty bodies that monitor them, the system of special procedures, the Office of the High Commissioner for Human Rights, and the Human Rights Council with its universal periodic review. The relationship between this machinery and the peace and security machinery has been uneasy. Human rights bodies have documented atrocities that the Security Council has then declined to address. The Council, for its part, has increasingly treated grave violations as evidence of a threat to the peace, has embedded human rights functions in peace operations, established commissions of inquiry, and referred situations to the International Criminal Court. Its referral power under the Rome Statute is a striking illustration of the Council's reach, and its selective use is a striking illustration of the Council's politics. 12.2 Development and the sustainable development agenda Chapter IX commits the organisation to promote higher standards of living, full employment, and conditions of economic and social progress and development. The framers understood, having watched the economic collapse of the 1930s feed the politics of the 1930s, that peace has material foundations. The development work of the organisation has evolved from technical assistance in the 1950s through the development decades and the Millennium Development Goals to the 2030 Agenda for #sustainable_development, whose sixteenth goal is explicitly devoted to peaceful, just, and inclusive societies. That goal is the clearest institutional expression of the proposition that development and #conflict_resolution are not separate agendas but a single one. 13. Amendment, Reform, and the Politics of the Impossible 13.1 The formal amendment procedure Article 108 provides that amendments to the Charter come into force for all members when they have been adopted by a vote of two thirds of the members of the General Assembly and ratified in accordance with their respective constitutional processes by two thirds of the members of the United Nations, including all the permanent members of the Security Council. Article 109 provides for a general conference to review the Charter, to be convened by a two thirds vote of the Assembly and a vote of any nine members of the Council, and stipulates that any alteration recommended by such a conference takes effect only when ratified by two thirds of the members including all the permanent members. The words including all the permanent members are the entire story. No amendment can enter into force over the objection of any one of the five. The permanent members therefore hold a veto not only over decisions but over the constitutional structure that grants them the veto. This is the closest thing in international law to a self entrenching clause, and it explains why the Charter has been amended only in the most marginal respects: the enlargement of the Security Council from eleven to fifteen members and the corresponding adjustment of the voting threshold, adopted in 1963 and effective in 1965; the enlargement of the Economic and Social Council, twice; and a technical amendment concerning the review conference. In eight decades, that is all. 13.2 The Security Council reform debate Every serious observer accepts that the composition of the Council reflects the world of 1945 rather than the world of today. Africa, with more than a quarter of the membership of the organisation and the majority of the items on the Council's agenda, has no permanent seat. Latin America has none. India, with the largest population on earth, has none. Two of the five permanent seats belong to Western European states. The distribution of permanent membership tracks the outcome of a war that ended before most of the world's present population was born. The reform debate has been formally under way since 1993, and negotiations have continued in an intergovernmental process since 2009. The principal camps are well defined and have hardly moved. One group, comprising Brazil, Germany, India, and Japan, seeks new permanent seats for themselves and for African states. A rival coalition, including Italy, Pakistan, Argentina, Mexico, and South Korea, opposes new permanent seats and proposes instead an expanded category of longer term, re electable, non permanent seats. The African common position demands two permanent seats with all the prerogatives of current permanent members, including the veto, and additional non permanent seats. A further group focuses primarily on the veto and on working methods rather than on composition. The impasse is structural. The aspirants to permanent membership are blocked by their regional rivals, who prefer no reform to a reform that elevates a neighbour. The African position insists on the veto, which the existing permanent members will not extend and which many reformers regard as the very problem to be solved. And any package must ultimately pass through Article 108, meaning that it requires the affirmative ratification of all five current permanent members. Three decades of negotiation have not produced a single agreed text. #UN_reform of the Council's composition is, on any realistic assessment, not achievable in the near term. 13.3 Reform that has actually happened Because formal amendment is blocked, reform has proceeded by other routes, and here the record is more encouraging than the rhetoric of stalemate suggests. Working methods have changed. The Council's procedures, once almost entirely closed, now include open debates, briefings by civil society representatives, and interaction with troop contributing countries. The selection of the Secretary General, historically conducted in secrecy among the permanent members, was opened in 2016 to public hearings with candidates before the General Assembly, a change achieved not by amending Article 97 but by the Assembly asserting a prerogative it had always possessed on paper. Restraint initiatives have gained adherents. A code of conduct promoted by a cross regional group commits its signatories to support timely action to prevent or end mass atrocities and not to vote against credible draft resolutions to that end. A separate initiative promoted by France and Mexico calls for voluntary restraint in the use of the veto in situations of mass atrocity. Neither is legally binding. Both create reputational costs for obstruction, which is not nothing. The most significant procedural innovation of recent years is the veto initiative adopted by the General Assembly in April 2022. Under this standing mandate, the Assembly is automatically convened within ten working days of the casting of a #veto in the Security Council to debate the situation concerned, and the vetoing member is invited to speak first. It does not abolish the veto or reverse its effect. It does impose a duty of public explanation, transforming what was once a silent act of obstruction into an occasion on which the vetoing state must justify itself before the whole membership. 13.4 Our Common Agenda, A New Agenda for Peace, and the Pact for the Future The most recent reform cycle began with the report Our Common Agenda in 2021, which set out a wide ranging vision of a renewed and more networked #multilateralism, encompassing future generations, the digital sphere, outer space, and the governance of global public goods. It was followed by a series of policy briefs, of which the most directly relevant here is A New Agenda for Peace, issued in 2023. That brief argued that the international security architecture is fragmenting, that prevention must be placed at the centre of the system, that peace operations must be reconceived for an era in which consent is fragile and geopolitical competition is intense, and that the strategic risks posed by nuclear weapons and by emerging technologies require urgent collective attention. An independent high level advisory board on effective multilateralism, reporting in 2023, made complementary recommendations on the governance of the global commons and on the representativeness of decision making bodies. The cycle culminated in the #Summit_of_the_Future in September 2024, at which the General Assembly adopted the #Pact_for_the_Future, together with a Global Digital Compact and a Declaration on Future Generations. The Pact contains commitments across the fields of sustainable development, international peace and security, science and technology, youth and future generations, and the transformation of global governance. On Security Council reform it contains stronger language than any previous consensus document, acknowledging the need to redress the historical injustice against Africa and to improve the representation of underrepresented regions, and committing members to intensify negotiations. An honest assessment must be sceptical. The Pact is a political declaration, not a treaty. It creates no new legal obligations and it does not amend the Charter. Its language on the Council reiterates aspirations that have been reiterated for three decades. Its value, if it has one, is agenda setting: it establishes benchmarks against which member states can be held to account, and it registers, at the highest political level, a collective admission that the institution is not fit for the purposes it now serves. 14. Critical Perspectives 14.1 The realist critique The realist tradition regards the Charter as an epiphenomenon. On this view, states comply with international rules when it suits their interests and disregard them when it does not; the Security Council is a device by which the powerful legitimate what they were going to do anyway; and the veto is not a defect but an accurate reflection of the underlying distribution of power. The evidence adduced is not trivial. Powerful states have repeatedly used force without authorisation and suffered no consequences, and the Council has never taken enforcement action against a permanent member. The reply is not that law always constrains. It is that the existence of the rule changes the terms on which power must operate. States that use force without authorisation go to considerable lengths to construct legal justifications, and the shape of those justifications is itself shaped by the Charter. No state today claims a general right to conquer. Every state that invades another asserts, however implausibly, that it is acting in self defence or at the invitation of a legitimate government. That universal hypocrisy is, in a modest but real sense, a tribute the powerful pay to the norm. 14.2 The Third World and postcolonial critique A more searching critique comes from scholars in the tradition of Third World approaches to international law. Their argument is not that the Charter is ineffective but that it is complicit. The organisation was founded by empires at the moment of their greatest reach, and its structures encoded the hierarchy of that moment. The permanent membership is a colonial artefact. Sovereignty, formally equal, operates in practice to protect the position of the strong. Interventions justified in the language of human rights reproduce, in the critics' view, the rhetorical structures of the civilising mission, and the economic institutions of the post war order entrenched patterns of dependency that formal political independence did not disturb. This critique deserves to be taken seriously rather than deflected. It is also complicated by an inconvenient fact that its more sophisticated proponents acknowledge: the newly independent states of Asia and Africa did not reject the organisation. They joined it, and they used the Charter's own language of sovereign equality and #self_determination as the principal instrument of their liberation. The Charter is thus both an instrument of hierarchy and the vocabulary in which hierarchy has been most effectively contested. 14.3 The legitimacy critique The most widely felt criticism is the simplest. The Council acts decisively in some crises and does nothing in others, and the difference is not explained by the gravity of the suffering. When the interests of a permanent member are engaged, the machinery stops. Vetoes blocked action on Syria for more than a decade, a paralysis that defined the #Security_Council in that period. The Council could not condemn the invasion of Ukraine because the invading state sits in the chair with a veto in its hand, an outcome so absurd that it functions as a reduction to absurdity of the whole arrangement. The Council's handling of the war in Gaza has produced repeated deadlock and a widespread perception, especially across the global South, of double standards. Selectivity of this kind is corrosive because #legitimacy in international institutions rests almost entirely on the perception of impartial rule application. An organisation that applies its rules to the weak and not to the strong will eventually be regarded, by the weak, as an instrument of the strong. That perception, more than any doctrinal weakness, is the gravest threat to the Charter system in the present decade. 14.4 The functionalist defence Against these critiques stands a defence that is unglamorous and largely accurate. Whatever the failures at the level of high politics, the Charter system does an enormous amount of quiet work that no one notices because it succeeds. Aircraft cross borders under rules agreed through a specialised agency. Disease outbreaks are reported and tracked, refugees are registered and fed, children are vaccinated, weather data is shared, radio spectrum is allocated, and many people are alive today because a peacekeeping mission stood between them and men with guns. Institutions are judged unfairly when their successes are invisible and their failures are televised. 15. The Charter Under Stress: Contemporary Case Material 15.1 Ukraine The invasion of Ukraine in February 2022 by a permanent member of the Security Council presented the Charter system with the scenario it was least equipped to handle. A draft resolution deploring the aggression was vetoed by the state committing it. The Council, using a procedural device under which the veto does not apply, referred the matter to an emergency special session of the General Assembly. On 2 March 2022 the Assembly adopted a resolution deploring in the strongest terms the aggression against Ukraine and demanding the immediate, complete, and unconditional withdrawal of forces, by an overwhelming majority. Further resolutions followed on the humanitarian consequences and on the purported annexation of Ukrainian territory. The Human Rights Council established a commission of inquiry, and the International Court of Justice, on an application invoking the Genocide Convention, indicated provisional measures ordering the suspension of military operations. The lesson is double edged. The enforcement machinery failed completely, exactly as its design guarantees when a permanent member is the aggressor. But the normative machinery functioned: the overwhelming majority of states, in a public and recorded vote, applied the Charter's rule against #aggression to a great power. Law without enforcement is weaker than law with it. It is not the same as no law at all. 15.2 Syria and Gaza The Syrian conflict produced more than a decade of Council paralysis, with repeated vetoes blocking condemnation, accountability referrals, and at times humanitarian access. The General Assembly responded by establishing an impartial mechanism to collect and preserve evidence of serious crimes for future prosecutions, an innovation that illustrates how blocked functions migrate to other organs. The war in Gaza since October 2023 has produced a comparable pattern of deadlock, with repeated vetoes on ceasefire language, an invocation of Article 99 by the Secretary General, provisional measures proceedings before the International Court of Justice, and an advisory opinion in 2024 on the legal consequences of the policies and practices of the occupying power. The gap between the volume of legal determination and the absence of enforcement is, in this case, especially stark. 15.3 The pattern Across these cases the same structure recurs. The prohibition of the #use_of_force is repeatedly affirmed and repeatedly violated. The Council is blocked whenever the interests of a permanent member are engaged. Other organs, the General Assembly, the Court, and the human rights bodies, expand to fill the vacuum, generating authoritative legal findings but no coercive remedy. The result is a system rich in law and poor in enforcement, in which the normative and the operational have come apart. 16. Discussion: What the Charter Is and Is Not Four conclusions emerge from this survey. First, the Charter is a constitution in function though not in form. It establishes organs, allocates competences, asserts supremacy over conflicting agreements, purports to reach non parties, and has generated a body of interpretive practice that no member can unilaterally revise. Understanding it as an ordinary contract between sovereign states misses most of what makes it powerful. Second, its formal rigidity has been compensated by informal flexibility. The text has scarcely changed; the institution has changed beyond recognition. Peacekeeping, the authorisation model of enforcement, the expansion of the threat to the peace, the abstention practice in Council voting, the emergency special session, the human rights machinery, the international tribunals, and the architecture of #peacebuilding were all built without amending a word. Constitutional change in this system happens through practice, acquiescence, and the accumulated weight of precedent. That is a strength, because it permits adaptation without the impossible hurdle of Article 108. It is also a weakness, because change achieved through practice can be reversed through practice. Third, the central defect is not in the norms but in the enforcement structure. Article 2(4) is a good rule, Article 51 a workable exception, Chapter VII a coherent design. The problem is that the organ entrusted with enforcement is constitutionally disabled from acting against the states most capable of breaking the rules, and is constitutionally protected from being reformed. Fourth, and notwithstanding all of the above, the Charter remains indispensable. It supplies the vocabulary in which international conduct is contested, the forum in which the weak can compel the strong to answer, the legal baseline against which aggression is measured, and the machinery through which a great deal of concrete good is done. The realistic ambition for the coming decade is not a rewritten Charter. It is a Council whose members exercise restraint in the use of the #veto, whose working methods are transparent, whose composition is somewhat more representative, and which is held to public account by a General Assembly that has learned to use the powers it already possesses. Modest as that sounds, it would represent the most significant strengthening of the #world_order in a generation. 17. Conclusion The Charter of the United Nations was drafted by people who had seen what happens when there are no rules. They produced a document that is imperfect, unequal, and in important respects unenforceable, and they produced it because the alternative was nothing. Eighty years later, its central prohibition remains the most important rule in #international_law, its organs remain the principal fora of global #diplomacy, and its promise remains unfulfilled. The task facing the present generation is not to venerate the Charter and not to discard it, but to hold the institution to the standard that the document itself sets. That standard is stated in its first sentence, and it has not changed: to save succeeding generations from the scourge of war. Whether the organisation built at San Francisco can be made to serve that purpose in an age of renewed great power rivalry is the central question of contemporary #global_governance, and it will not be answered in a summit declaration. It will be answered by what states do. Hashtags: #United_Nations #UN_Charter_1945 #Charter_of_the_United_Nations #international_law #collective_security #Security_Council_reform #peace_and_security #conflict_resolution #global_governance #multilateralism #use_of_force #veto_power #peacekeeping_operations #human_rights #self_determination #international_organisations #UN_at_80 #world_peace #diplomacy_studies #law_of_nations References Primary legal instruments Charter of the United Nations, signed at San Francisco, 26 June 1945, entered into force 24 October 1945. 1 UNTS XVI. Statute of the International Court of Justice, annexed to the Charter of the United Nations, 26 June 1945. Covenant of the League of Nations, 1919. General Treaty for the Renunciation of War as an Instrument of National Policy (Kellogg-Briand Pact), Paris, 27 August 1928. Vienna Convention on the Law of Treaties, Vienna, 23 May 1969, 1155 UNTS 331. Rome Statute of the International Criminal Court, 17 July 1998, 2187 UNTS 3. United Nations documents and resolutions United Nations General Assembly, Resolution 377 (V), Uniting for Peace, 3 November 1950. United Nations General Assembly, Resolution 1514 (XV), Declaration on the Granting of Independence to Colonial Countries and Peoples, 14 December 1960. United Nations Security Council, Resolution 678 (1990), 29 November 1990. United Nations, An Agenda for Peace: Preventive Diplomacy, Peacemaking and Peace-keeping, Report of the Secretary-General, A/47/277-S/24111, 17 June 1992. United Nations, Report of the Panel on United Nations Peace Operations (Brahimi Report), A/55/305-S/2000/809, 21 August 2000. International Commission on Intervention and State Sovereignty, The Responsibility to Protect, International Development Research Centre, Ottawa, 2001. United Nations General Assembly, Resolution 60/1, 2005 World Summit Outcome, 24 October 2005, paragraphs 138 to 140. United Nations, Uniting Our Strengths for Peace: Politics, Partnership and People, Report of the High-Level Independent Panel on Peace Operations, A/70/95-S/2015/446, 2015. United Nations Security Council, Resolution 1973 (2011), 17 March 2011. United Nations Secretary-General, Our Common Agenda, Report of the Secretary-General, A/75/982, United Nations, New York, 2021. United Nations General Assembly, Resolution ES-11/1, Aggression against Ukraine, 2 March 2022. United Nations General Assembly, Resolution 76/262, Standing mandate for a General Assembly debate when a veto is cast in the Security Council, 26 April 2022. United Nations, A Breakthrough for People and Planet: Effective and Inclusive Global Governance for Today and the Future, Report of the High-Level Advisory Board on Effective Multilateralism, United Nations University, 2023. United Nations Secretary-General, A New Agenda for Peace, Our Common Agenda Policy Brief 9, United Nations, New York, July 2023. United Nations General Assembly, Resolution 79/1, Pact for the Future, Global Digital Compact and Declaration on Future Generations, 22 September 2024. Decisions and advisory opinions of the International Court of Justice Reparation for Injuries Suffered in the Service of the United Nations, Advisory Opinion, ICJ Reports 1949, p. 174. Certain Expenses of the United Nations (Article 17, paragraph 2, of the Charter), Advisory Opinion, ICJ Reports 1962, p. 151. Legal Consequences for States of the Continued Presence of South Africa in Namibia (South West Africa) notwithstanding Security Council Resolution 276 (1970), Advisory Opinion, ICJ Reports 1971, p. 16. Military and Paramilitary Activities in and against Nicaragua (Nicaragua v. United States of America), Merits, Judgment, ICJ Reports 1986, p. 14. Legality of the Threat or Use of Nuclear Weapons, Advisory Opinion, ICJ Reports 1996, p. 226. Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory, Advisory Opinion, ICJ Reports 2004, p. 136. Accordance with International Law of the Unilateral Declaration of Independence in Respect of Kosovo, Advisory Opinion, ICJ Reports 2010, p. 403. Legal Consequences of the Separation of the Chagos Archipelago from Mauritius in 1965, Advisory Opinion, ICJ Reports 2019, p. 95. Allegations of Genocide under the Convention on the Prevention and Punishment of the Crime of Genocide (Ukraine v. Russian Federation), Provisional Measures, Order of 16 March 2022. Application of the Convention on the Prevention and Punishment of the Crime of Genocide in the Gaza Strip (South Africa v. Israel), Provisional Measures, Order of 26 January 2024. Legal Consequences arising from the Policies and Practices of Israel in the Occupied Palestinian Territory, including East Jerusalem, Advisory Opinion, 19 July 2024. Books and scholarly articles Klabbers, Jan. An Introduction to International Organizations Law. 4th edition. Cambridge: Cambridge University Press, 2022. Hurd, Ian. International Organizations: Politics, Law, Practice. 4th edition. Cambridge: Cambridge University Press, 2021. Corten, Olivier. The Law Against War: The Prohibition on the Use of Force in Contemporary International Law. 2nd edition. Oxford: Hart Publishing, 2021. Kress, Claus. The Ukraine War and the Prohibition of the Use of Force in International Law. Occasional Paper Series No. 13. Brussels: Torkel Opsahl Academic EPublisher, 2022. Simma, Bruno, Daniel-Erasmus Khan, Georg Nolte and Andreas Paulus (editors). The Charter of the United Nations: A Commentary. Oxford: Oxford University Press, 2012. Gray, Christine. International Law and the Use of Force. 4th edition. Oxford: Oxford University Press, 2018. Crawford, James. Brownlie's Principles of Public International Law. 9th edition. Oxford: Oxford University Press, 2019. Weiss, Thomas G. and Sam Daws (editors). The Oxford Handbook on the United Nations. 2nd edition. Oxford: Oxford University Press, 2018. Chesterman, Simon, David M. Malone and Santiago Villalpando (editors). The Oxford Handbook of United Nations Treaties. Oxford: Oxford University Press, 2019. Ruys, Tom, Olivier Corten and Alexandra Hofer (editors). The Use of Force in International Law: A Case-Based Approach. Oxford: Oxford University Press, 2018. Thakur, Ramesh. Reviewing the Responsibility to Protect: Origins, Implementation and Controversies. London: Routledge, 2019. Anghie, Antony. Imperialism, Sovereignty and the Making of International Law. Cambridge: Cambridge University Press, 2005. Chimni, B. S. International Law and World Order: A Critique of Contemporary Approaches. 2nd edition. Cambridge: Cambridge University Press, 2017. Mazower, Mark. Governing the World: The History of an Idea. London: Penguin, 2012. Schlesinger, Stephen C. Act of Creation: The Founding of the United Nations. Boulder: Westview Press, 2003.

This article examines the body of international agreements that states have built since the twentieth century to limit war, restrain weapons and protect civilians. It treats these agreements as one connected system rather than as a list of separate texts. The study covers the weapons of mass destruction regimes, including the Treaty on the Non-Proliferation of Nuclear Weapons, the Comprehensive Nuclear-Test-Ban Treaty, the Treaty on the Prohibition of Nuclear Weapons, the Chemical Weapons Convention and the Biological Weapons Convention. It then covers conventional and humanitarian instruments, including the Convention on Certain Conventional Weapons, the Anti-Personnel Mine Ban Convention, the Convention on Cluster Munitions and the Arms Trade Treaty. The article also looks at the newer areas of autonomous weapons, outer space and cyber operations, where law is still forming. The article argues three points. First, disarmament conventions have delivered measurable results, such as the destruction of declared chemical stockpiles, the clearance of contaminated land and the sharp fall in the number of nuclear warheads since the 1980s. Second, the system is now under visible strain. The last bilateral treaty limiting United States and Russian strategic nuclear forces expired on 5 February 2026, five European states have left the mine ban treaty since December 2025, and the Eleventh Review Conference of the Non-Proliferation Treaty closed in May 2026 without a consensus final document. Third, the crisis is political rather than legal. The rules did not fail on their own terms. States changed their view of what the rules cost them. The article ends with recommendations for repairing the system, and with study questions for students of international law, political science and peace studies. Keywords: disarmament law, arms control, non-proliferation, humanitarian disarmament, treaty compliance, international security 1. Introduction For most of recorded history, the decision to arm was treated as a private matter for each ruler. A state built what it could afford and used what it thought it needed. The idea that the international community could sit down together, write a rule, and then bind sovereign states to that rule in matters of national defence is young. It is barely more than a century old in any organised form, and it only became a dense legal system after 1945. That system is what this article calls the architecture of #peace, #security and #disarmament conventions. It is not a single building. It is a set of structures built at different times, by different coalitions, for different reasons, and joined together by a shared assumption: that unlimited weapons produce unlimited danger, and that mutual restraint written into #international_law is safer than restraint left to goodwill. The architecture matters because it is being tested harder now than at any point since the early 1980s. On 5 February 2026 the New Strategic Arms Reduction Treaty, known as #New_START, expired without a successor. It was the last legally binding cap on the strategic nuclear forces of the two largest arsenals in the world. Between December 2025 and February 2026, Estonia, Latvia, Lithuania, Finland and Poland completed their withdrawal from the 1997 Anti-Personnel Mine Ban Convention, the first withdrawals in that treaty's history. In May 2026, the Eleventh Review Conference of the Treaty on the Non-Proliferation of Nuclear Weapons ended in New York without an agreed final document, extending a record of failure that now stretches across sixteen years of review cycles. At the same time, the system is not dead. New law is still being made. The Treaty on the Prohibition of Nuclear Weapons entered into force in January 2021 and will hold its first review conference in New York from 30 November to 4 December 2026, with South Africa in the chair. States continue to meet in Geneva on autonomous weapons. Mine clearance operators continue to hand cleared land back to farmers in Cambodia, Angola and Iraq. The Organisation for the Prohibition of Chemical Weapons continues to verify the destruction of chemical stocks. So the picture is mixed, and a serious study has to hold both halves of it at once. This article does that. It is written for students, and it tries to explain the law in plain language without simplifying it into slogans. 1.1 Research questions The article is organised around four questions. What is the structure of the international system of disarmament and arms control conventions, and how did it come to look the way it does? What has this system actually achieved, measured against evidence rather than against ambition? Why is the system under such heavy strain in the middle of the 2020s, and is the strain legal, technical or political? What can be done, realistically, to preserve what works and rebuild what has broken? 1.2 Why the topic matters for students Students often meet this subject as a list of dates and acronyms to memorise. That is the least useful way to learn it. The subject is really about a hard question in political theory: how do you get sovereign actors, who do not trust each other, to give up capabilities they believe keep them alive? Every treaty discussed below is an answer to that question, and every crisis discussed below is the answer breaking down. Once you see the pattern, the acronyms become easy, because each one is a variation on the same problem. That problem is not going away. It will be waiting for whoever reads this article when they enter public service, journalism, humanitarian work, research or law. 2. Conceptual framework: four words that are often confused Before the treaties, the vocabulary. Four terms are used loosely in public debate and precisely in scholarship, and the difference matters. 2.1 Peace #Peace in the legal sense is not simply the absence of shooting. The scholarly tradition distinguishes negative peace, meaning the absence of direct violence, from positive peace, meaning the presence of conditions that make violence unlikely: functioning institutions, fair distribution, dispute settlement, and the ability of groups to pursue their interests without force. The #UN_Charter of 1945 reflects both ideas. It bans the threat or use of force in Article 2(4), which is negative peace, and it commits members to economic and social cooperation, which points toward positive peace. Disarmament law sits mostly on the negative side, but not entirely. Treaties like the mine ban convention contain obligations on clearance, risk education and assistance to survivors. Those obligations are about rebuilding life after violence, which is a positive-peace project carried inside an arms control instrument. 2.2 Security #Security is a contested word. The traditional meaning is national security: the protection of the state, its territory and its institutions from external attack. From the 1990s onward, scholars and United Nations bodies developed the idea of #human_security, which shifts the question from what threatens the state to what threatens people. A landmine does not threaten the survival of a state. It threatens the farmer who steps on it. Once the object of security shifts from the state to the person, entire categories of weapons become unacceptable that would otherwise be seen as ordinary military tools. That shift explains the entire humanitarian branch of disarmament law. 2.3 Arms control #Arms_control means managing weapons rather than removing them. It accepts that states will keep armed forces and tries to make the resulting competition safer, cheaper and more predictable. Ceilings on the number of deployed warheads, notification of missile tests, hotlines and inspection rights are all arms control measures. Their goal is #strategic_stability, meaning a situation where neither side has an incentive to strike first in a crisis. 2.4 Disarmament #Disarmament means elimination. It is the removal of a capability from the world, not just its regulation. The Chemical Weapons Convention is a disarmament treaty because it requires the destruction of stocks. The Anti-Personnel Mine Ban Convention is a disarmament treaty. New START was an arms control treaty, because it capped numbers without banning the weapon. The difference is not academic hair splitting. It determines what success looks like. An arms control regime succeeds when the competition remains stable. A disarmament regime succeeds only when the weapon is gone. When commentators say a treaty has failed, always ask which of the two standards they are applying. 2.5 A working typology For analysis, the conventions can be sorted into five families. Prohibition regimes, which ban a weapon outright. Examples: the Biological Weapons Convention, the Chemical Weapons Convention, the mine ban convention, the cluster munitions convention, the nuclear prohibition treaty. Non-proliferation regimes, which allow some states to keep a weapon while forbidding others to acquire it, usually in exchange for benefits. The Non-Proliferation Treaty is the main example, and the bargain it contains is the source of most of its political trouble. Quantitative arms control, which caps numbers and builds verification. New START and its predecessors are the model. Regulatory regimes, which do not ban a weapon but restrict how it may be used or transferred. The Convention on Certain Conventional Weapons and the Arms Trade Treaty belong here. Confidence-building and transparency regimes, which do not restrict capability at all but reduce the risk of misreading the other side. Registers, notifications, observation of exercises and incident hotlines fall in this group. Most real treaties mix these types. The clean categories are a teaching tool, not a description of the world. 3. Method and sources This is a doctrinal and analytical review, not an experiment. The method has three parts. First, primary legal analysis. The article works from the text of the conventions themselves, from their protocols, and from the official records of review conferences and meetings of states parties. Treaty text is the ground truth. Political commentary about a treaty is not the same thing as the treaty. Second, evidence on implementation. Claims about what a treaty has achieved are checked against systematic monitoring: the annual reporting of the Organisation for the Prohibition of Chemical Weapons, the Landmine Monitor and Cluster Munition Monitor produced by civil society, the annual data published by the Stockholm International Peace Research Institute, and the warhead estimates published in the Nuclear Notebook series. Third, interpretation. The article draws on the international relations literature on regimes, norms and compliance to explain why states join, obey, ignore or leave these agreements. Two limitations should be stated at the start. Verification data are incomplete by nature, because states with the most to hide report the least. And the situation is moving quickly. Events after mid-2026 are outside the scope of this article, and readers should check current sources before relying on any figure quoted here. 4. Historical evolution: how the architecture was built 4.1 The first attempts, 1868 to 1918 The earliest modern effort to restrain weapons by agreement is the 1868 St Petersburg Declaration, which banned exploding bullets under a certain weight. Its preamble contains the sentence that still underpins the entire field: the only legitimate aim of war is to weaken the military forces of the enemy, and weapons that uselessly aggravate suffering go beyond that aim. The Hague Conferences of 1899 and 1907 extended the idea. They produced rules on the conduct of hostilities and attempted, without much success, to limit new technologies such as aerial bombardment. They failed to prevent 1914, and their failure produced the first lesson of the field: rules on how to fight do not by themselves prevent the decision to fight. 4.2 The interwar experiment, 1919 to 1939 After the First World War, the 1925 Geneva Protocol banned the use in war of asphyxiating gases and of bacteriological methods. It was a real achievement, and chemical weapons were largely absent from the European battlefields of the Second World War. But the protocol had a hole in it. It banned use, not possession, and many states joined with reservations allowing retaliation in kind. Deterrence, not law alone, produced the restraint. The wider interwar disarmament project failed. The World Disarmament Conference of 1932 to 1934 collapsed. The naval treaties of the 1920s were abandoned in the 1930s. The second lesson of the field was learned here: disarmament agreements do not survive a collapse of the political relationships that produced them. This lesson is directly relevant to 2026. 4.3 The Cold War and the birth of modern arms control, 1945 to 1991 The atomic bombings of Hiroshima and Nagasaki in 1945 changed the subject permanently. From then on, the central question of international security was how to live with a weapon that could end organised life. The early answer was the #UN_Charter and #collective_security. The Charter banned aggressive force and gave the #UN_Security_Council the power to authorise collective responses. It also, in Article 26, gave the Council responsibility for plans to regulate armaments. That article has never been seriously used, which is itself a lesson about the gap between paper powers and political will. The practical answer came later and in pieces, usually after a scare. The Cuban missile crisis of 1962 was followed by the Partial Test Ban Treaty of 1963 and the Moscow-Washington hotline. The spread of nuclear technology in the 1960s was followed by the #NPT in 1968. The realisation that biological weapons were militarily unreliable but morally catastrophic produced the #BWC in 1972. Fear of an unlimited missile defence race produced the Anti-Ballistic Missile Treaty in the same year. By the late Cold War, arms control had become a professional discipline with its own vocabulary of throw-weight, telemetry and on-site inspection. Its finest hour came at the end. The 1987 Intermediate-Range Nuclear Forces Treaty eliminated an entire class of missiles, with intrusive verification. The 1990 Treaty on Conventional Armed Forces in Europe capped tanks, artillery and aircraft across a continent. The 1991 START agreement began deep cuts in strategic forces. 4.4 The humanitarian turn, 1992 to 2010 The end of the Cold War opened space for a different kind of disarmament. Instead of two superpowers negotiating in private about strategic balance, coalitions of medium and small states worked with non-governmental organisations to ban weapons on the ground of their effects on civilians. The #CWC of 1993 belongs partly to this period. It bans an entire category of weapon, requires destruction of stocks, and created a permanent inspectorate. It remains the most complete disarmament treaty ever built. The clearest case of the new model is the 1997 #Ottawa_Convention on anti-personnel mines. It was negotiated outside the consensus-bound Conference on Disarmament, driven by the International Campaign to Ban Landmines, and adopted in about a year. The campaign received the #Nobel_Peace_Prize in 1997. The same method produced the 2008 Convention on Cluster Munitions. This approach is called #humanitarian_disarmament, and its logic is simple: if a weapon cannot be used without unacceptable civilian harm, the weapon must go, whatever military convenience it offers. 4.5 The stalling and reversal, 2011 to 2026 The last fifteen years have run in the other direction. The Anti-Ballistic Missile Treaty had already been abandoned in 2002. The Intermediate-Range Nuclear Forces Treaty ended in 2019 after mutual accusations of violation. The Treaty on Open Skies unravelled in 2020 and 2021. Russia suspended its participation in New START in 2023 and revoked its ratification of the test ban treaty in the same year. The conventional forces treaty in Europe is defunct. The one clear counter-current is the #TPNW, adopted in 2017 and in force since 22 January 2021. It shows that new prohibition law can still be made. It also shows the limits of that law, because no state that possesses nuclear weapons has joined it or intends to. 5. The weapons of mass destruction regimes 5.1 The Treaty on the Non-Proliferation of Nuclear Weapons The NPT was opened for signature in 1968 and entered into force in 1970. With 191 states parties, it is the most widely joined disarmament agreement in existence. Only a handful of states remain outside it, and each of those is nuclear armed or nuclear capable. The treaty rests on a bargain with three parts, often called the three pillars. The first pillar is #non_proliferation. Under Articles I and II, the five states that had tested a nuclear device before 1 January 1967 undertake not to transfer nuclear weapons, and all other parties undertake not to acquire them. The second pillar is #disarmament. Article VI obliges all parties to pursue negotiations in good faith on effective measures relating to nuclear disarmament. The wording is vague, deliberately so, and the gap between what non-nuclear states believe it requires and what nuclear-armed states admit it requires is the central political wound of the regime. The third pillar is peaceful use. Article IV recognises an inalienable right to develop nuclear energy for peaceful purposes, subject to safeguards. The verification system is run by the International Atomic Energy Agency. Comprehensive #safeguards agreements give the #IAEA the right to account for declared nuclear material. After the discovery of Iraq's undeclared programme in 1991, states added the Additional Protocol, which gives inspectors far wider access. The Additional Protocol is voluntary, and its uneven adoption is a structural weakness. What has the NPT achieved? Predictions in the 1960s suggested twenty to thirty nuclear-armed states within a generation. There are nine today. Several states that had programmes gave them up. South Africa built weapons and then dismantled them before joining the treaty in 1991. Belarus, Kazakhstan and Ukraine transferred Soviet warheads to Russia and joined as non-nuclear states. Argentina and Brazil abandoned competing programmes and built a joint inspection agency. Libya gave up its programme in 2003. These are not small results. The failures are equally clear. India, Israel and Pakistan never joined and are nuclear armed. North Korea announced withdrawal in 2003, tested weapons from 2006 and has expanded since. Iran's programme has repeatedly produced crises, agreements and collapses. And the disarmament pillar has not been delivered to the satisfaction of the majority. Global arsenals fell steeply after 1986, but that fall has stopped. Roughly twelve thousand warheads remain in the world, of which about nine thousand six hundred sit in military stockpiles, and every possessor state is modernising. The review cycle now shows the strain. The 2015 and 2022 review conferences produced no consensus document. The Eleventh Review Conference, held in New York from 27 April to 22 May 2026 under the presidency of Ambassador Do Hung Viet of Viet Nam, tried a new procedural route by circulating an early draft text rather than relying only on committee reports. It still ended without agreement. Sixteen years have now passed since the review process last produced an agreed outcome. The treaty is not dead, since its core obligations continue to bind, but its political machinery is jammed. 5.2 The Comprehensive Nuclear-Test-Ban Treaty The #CTBT was adopted in 1996. It bans any nuclear explosion, anywhere, for any purpose. It has not entered into force, because entry into force requires ratification by a specific list of forty-four states with nuclear technology, and nine of them have not ratified: China, the Democratic People's Republic of Korea, Egypt, India, Iran, Israel, Pakistan, Russia and the United States. Russia revoked its earlier ratification in 2023, aligning its position with that of the United States. Even so, the treaty is one of the more successful instruments in the field, which is a strange thing to say about a text that is not in force. Two reasons explain this. First, the norm holds. Since 1996, apart from North Korea, no state has conducted a nuclear explosive test. A near-universal practice of restraint has developed around an unratified treaty. That is a powerful demonstration of how #norms can bite even without formal legal force. Second, the machinery works. The Preparatory Commission has built an International Monitoring System of more than three hundred facilities using seismic, hydroacoustic, infrasound and radionuclide sensors. It detected each North Korean test. The data are also used for tsunami warning and for tracking radioactive releases, which gives the system a civilian value that helps sustain its funding. The risk in the current period is that a return to explosive #nuclear_testing by one major power would trigger others, and the norm that has held for thirty years would collapse quickly. 5.3 The Treaty on the Prohibition of Nuclear Weapons The TPNW was adopted on 7 July 2017 by a vote of 122 states and entered into force on 22 January 2021. It is the first treaty to ban #nuclear_weapons in the same comprehensive way that chemical and biological weapons were banned. Parties undertake never to develop, test, produce, acquire, possess, stockpile, transfer, use or threaten to use nuclear weapons, or to allow them to be stationed on their territory. Its origin lies in the humanitarian initiative of 2013 to 2014, when a series of conferences in Oslo, Nayarit and Vienna examined what would actually happen if a nuclear weapon were detonated in a city. The testimony of the #hibakusha, the survivors of Hiroshima and Nagasaki, and of people affected by test programmes in the Pacific, Kazakhstan and Algeria, shifted the debate from abstract strategy to concrete consequence. The International Campaign to Abolish Nuclear Weapons, #ICAN, coordinated the civil society side of that effort and received the Nobel Peace Prize in 2017. As of October 2025 the treaty had 95 signatories and 74 states parties, and the first review conference is scheduled for New York from 30 November to 4 December 2026 under South African presidency. The critique of the treaty is straightforward. No nuclear-armed state has joined, none of their allies who shelter under extended deterrence have joined, and therefore the treaty prohibits nothing that its parties were going to do anyway. Critics argue it may even harm the NPT by splitting the non-nuclear majority from the nuclear minority. The defence is equally clear. Law does not only constrain by direct application. It also changes what counts as normal. The treaty makes possession legally deviant, gives financial institutions grounds to divest, gives parliamentarians a text to campaign on, and creates an alternative forum where the humanitarian and environmental consequences of nuclear weapons are on the record. This is the process of #stigmatisation, and its results are slow and cumulative rather than immediate. The treaty also does something the NPT never did. Articles 6 and 7 create obligations of #victim_assistance and environmental remediation for people and places harmed by nuclear use and testing. That is a real, operational agenda, and it explains why states with testing legacies, such as Kazakhstan and Kiribati, have been active parties. 5.4 The Chemical Weapons Convention The #CWC entered into force in 1997 and is the strongest of all disarmament treaties in design. It bans development, production, stockpiling, transfer and use of #chemical_weapons; it requires destruction of existing stocks and production facilities under international supervision; and it created a permanent implementing body, the #OPCW, with a professional inspectorate and the right to conduct challenge inspections anywhere at short notice. It has 193 states parties, which is close to universal. The results are measurable in tonnes. All chemical weapons stockpiles declared by states parties have been verifiably destroyed, a process completed in 2023 when the last declared United States stockpile was eliminated. Tens of thousands of tonnes of nerve and blister agent no longer exist. The OPCW received the Nobel Peace Prize in 2013. The problems are equally instructive. Verified destruction of declared stocks is not the same as the end of chemical warfare. Chemical weapons were used repeatedly in the Syrian conflict from 2013. Nerve agents of the Novichok family were used in assassination attempts in the United Kingdom in 2018 and in Russia in 2020. In response, states parties in 2018 gave the OPCW a mandate not only to say that an attack occurred but to identify who was responsible, a step that was pushed through by vote rather than consensus and that remains politically contested. Three lessons follow. First, near-universal membership and strong verification still cannot stop a determined violator; they can only expose it. Second, attribution is a political act and creating institutions that attribute blame will divide an organisation. Third, the frontier is moving. Central nervous system acting chemicals and advances in chemical synthesis and automation raise questions the drafters of 1993 did not face. 5.5 The Biological Weapons Convention The #BWC opened for signature in 1972 and entered into force in 1975. It was the first treaty to ban an entire category of weapon. It prohibits the development, production and stockpiling of #biological_weapons, and it has around 190 states parties. Its defect is famous. It has no #verification mechanism. There is no inspectorate, no declaration system with teeth, and no equivalent of the OPCW. An attempt to negotiate a verification protocol ran from 1995 to 2001 and collapsed when the United States rejected the draft, arguing that it would not catch cheaters and would expose commercial and biodefence secrets. Efforts to strengthen the convention have continued through review conferences, and the Ninth Review Conference in 2022 established a working group on strengthening the treaty, including on verification, but progress has been slow. This gap has become more serious, not less, as biotechnology has advanced. Gene synthesis is cheap. Design tools using #artificial_intelligence can propose novel toxic compounds. The knowledge and equipment needed are widely distributed and overwhelmingly used for good purposes, which makes any inspection system intrusive, expensive and hard to design. The COVID-19 pandemic, whatever its origin, showed how badly the world manages large biological events even when no one intends harm. For students, the BWC is the best available case study in what a treaty without verification can and cannot do. It sets a norm that is close to universal and rarely challenged in public. It cannot, on its own, provide assurance. 5.6 Nuclear-weapon-free zones Regional treaties have taken entire continents out of the nuclear competition. The Treaty of Tlatelolco, #Tlatelolco, covers Latin America and the Caribbean and was the first. Rarotonga covers the South Pacific, Bangkok covers Southeast Asia, #Pelindaba covers Africa, and Semipalatinsk covers Central Asia. Mongolia has a recognised single-state status. Together these #nuclear_weapon_free_zones cover the whole southern hemisphere and much of the north. They matter for three reasons. They provide legally binding assurance among neighbours, which reduces regional insecurity. They give non-nuclear states a collective voice in global forums. And they demonstrate that #regional_security arrangements can go further than global consensus permits, which is a useful model when global negotiation is blocked. The unfinished case is the Middle East. A zone free of nuclear and other weapons of mass destruction in that region has been on the agenda since 1974 and was part of the bargain that secured indefinite extension of the NPT in 1995. A conference process on the #Middle_East_WMD_Free_Zone now meets annually in New York, but Israel does not participate and the United States has been sceptical, so progress is limited. 6. Bilateral nuclear arms control and the end of New START The bilateral track between Washington and Moscow was always the load-bearing wall of the nuclear order. It is now gone, and students should understand exactly what was lost. New START was signed in 2010 and entered into force on 5 February 2011. It limited each side to 1,550 deployed strategic warheads, 700 deployed delivery vehicles, and 800 deployed and non-deployed launchers. Crucially, it was not only a set of numbers. It contained data exchanges twice a year, notifications of movements, exhibitions of new systems, up to eighteen on-site inspections a year, a commitment not to interfere with the other side's national technical means of verification, and a standing consultative body to handle disputes. The decline was gradual. On-site inspections were suspended in 2020 because of the pandemic and never resumed. In February 2023, after the full-scale invasion of Ukraine and the deterioration of relations, Russia suspended its participation, while stating it would keep observing the numerical ceilings. The treaty allowed only one extension, which had been used in 2021, so 5 February 2026 was always the final date. In September 2025 President Putin proposed that both sides continue to observe the central limits for one year after expiry, without verification. The United States did not formally accept the proposal, and the treaty lapsed on schedule. For the first time since 1972, there is no legally binding limit on the strategic nuclear forces of the two states that hold most of the world's warheads. Three consequences follow. The first is the loss of information. Without inspections and data exchanges, each side must rely on its own intelligence to estimate the other's forces. Uncertainty leads planners to assume the worst, and worst-case assumptions drive procurement. This is the classic mechanism of the #arms_race, and it operates even when neither side wants a race. The second is the loss of a ceiling. Both the United States and Russia have the technical ability to upload additional warheads onto existing missiles relatively quickly, because reductions under the treaty were achieved partly by removing warheads rather than destroying them. The third is complexity. Any future agreement will have to reckon with China, whose arsenal is growing faster than any other, with non-strategic nuclear weapons, with missile defence, and with new systems such as hypersonic glide vehicles and long-range conventional strike. Washington has said future talks should include Beijing. Beijing has so far declined, arguing that its arsenal is far smaller. Moscow has argued that British and French forces should then be counted too. Each of these positions is understandable and together they make agreement very hard. The realistic near-term agenda is therefore not a grand treaty. It is risk reduction: keeping military-to-military channels open, agreeing rules on incidents at sea and in the air, maintaining launch notification arrangements, and building #confidence_building measures around new technologies. These are modest steps. In the current climate they would still be significant achievements. 7. Conventional weapons and humanitarian disarmament Nuclear weapons dominate the headlines. Conventional weapons dominate the casualty statistics. Almost everyone killed in armed conflict in the past eighty years was killed by a conventional weapon, and most of them by weapons that are cheap, portable and widely traded. 7.1 The Convention on Certain Conventional Weapons The #CCW of 1980 is a framework treaty with five protocols. Protocol I bans weapons that injure by fragments not detectable by X-ray. Protocol II, amended in 1996, regulates mines, booby traps and other devices. Protocol III restricts incendiary weapons. Protocol IV, adopted in 1995, bans blinding laser weapons before they were ever deployed, which is the only true case of pre-emptive prohibition in the modern law of weapons. Protocol V, adopted in 2003, deals with explosive remnants of war. The CCW works by consensus. That gives every state a veto, which is why the convention regulates rather than bans, and why frustrated states left it in the 1990s to negotiate the mine ban treaty elsewhere. The same dynamic is visible today in the debate on autonomous weapons. 7.2 The Anti-Personnel Mine Ban Convention The Ottawa Convention was adopted in 1997 and entered into force in 1999. It bans the use, production, stockpiling and transfer of anti-personnel #landmines, requires destruction of stockpiles, requires clearance of contaminated areas, and requires assistance to victims. Before the recent withdrawals it had around 165 states parties. Its results were among the best in the whole field. Tens of millions of stockpiled mines were destroyed. Large areas of land were cleared and returned to use in Mozambique, Cambodia, Colombia, Bosnia and elsewhere. Annual casualties fell steeply from the levels of the early 1990s. Trade in anti-personnel mines effectively stopped. Even states outside the treaty, including the United States for long periods, adjusted their policies in its shadow. The Landmine Monitor, produced each year by civil society, made the regime measurable and therefore accountable. Then came the reversal. On 18 March 2025 the defence ministers of Estonia, Latvia, Lithuania and Poland jointly announced that they would leave the treaty, citing a fundamental deterioration of security in their region since Russia's invasion of Ukraine. Finland announced the same intention on 1 April 2025. All five deposited instruments of withdrawal between June and August 2025. Under Article 20, withdrawal takes effect six months after deposit. The withdrawals of Estonia, Latvia and Lithuania took effect on 27 December 2025, Finland's on 10 January 2026, and Poland's on 20 February 2026. Finland, Lithuania and Poland have signalled that they intend to resume production of anti-personnel mines. Ukraine, which is a state party, has also announced that it will stop implementing the treaty, although the text provides that withdrawal does not take effect while the withdrawing state is engaged in armed conflict. This is the largest set of exits from any humanitarian disarmament treaty. At the annual meeting of states parties in Geneva in December 2025, other states recorded their regret and warned that the withdrawals damage the goal of #universalisation. Students should think carefully about this case, because it is a live test of a theoretical question. If a treaty is only obeyed when it is cheap, is it law or is it convenience? The withdrawing states did not break the treaty. They used its own exit clause, in the open, with parliamentary approval. That is lawful behaviour, and it is also a serious blow to the norm. The lesson is uncomfortable: the humanitarian argument was strong enough to win when the perceived military value of mines was low, and it was not strong enough to hold when that perception changed. There is a counter-argument that deserves a fair hearing. States on NATO's eastern flank face an adversary that uses mines freely, is not party to the treaty, and has invaded a neighbour. They argue that a defensive barrier along a threatened border is a proportionate response, that modern mines can be made self-deactivating, and that their own populations would bear the risk they are choosing to accept. Against this, the International Committee of the Red Cross has argued that studies of past conflicts do not support claims that anti-personnel mines are of high military value, and that the long-term burden of contamination falls on civilians for decades after the fighting ends. Both arguments should be understood before either is judged. 7.3 The Convention on Cluster Munitions The Convention on Cluster Munitions was adopted in 2008 and entered into force in 2010, following the same model as the mine ban treaty. It bans #cluster_munitions, which scatter submunitions over a wide area and leave large numbers of unexploded bomblets behind. It has around 110 states parties, and it too was driven by a coalition of affected states and civil society. It has faced the same reversal. Lithuania gave notice in September 2024 and its withdrawal took effect on 6 March 2025, the first withdrawal from the convention. It cited Russian use of cluster munitions in Ukraine, interoperability with allies, and improvements in the technology. Meanwhile, cluster munitions have been used extensively in Ukraine by both sides, and transfers of such weapons by a state that is not party to the convention have exposed the limits of a prohibition regime that the largest producers have never joined. 7.4 The Arms Trade Treaty The #Arms_Trade_Treaty, adopted in 2013 and in force since December 2014, is different in kind. It does not ban a weapon. It regulates the international transfer of conventional arms, from small arms to battle tanks and warships. Parties must assess whether an export would be used to commit or facilitate genocide, crimes against humanity, grave breaches of the Geneva Conventions, or serious violations of international humanitarian or human rights law, and must not authorise the transfer if there is an overriding risk of such consequences. It has more than 110 states parties, but several major exporters are outside it or have unsigned it, and reporting by parties is uneven. Its central weakness is that the risk assessment is carried out by the exporting state itself, and there is no international body that can overturn a national decision. Critics point to continued exports to parties in conflicts where serious violations are documented. Supporters answer that the treaty has created a common language of risk assessment, improved national licensing systems, and given parliaments and courts a legal standard against which to test government decisions. 7.5 Small arms and light weapons Most people killed in conflict are killed with #small_arms. The international framework here is weaker than elsewhere: a politically binding Programme of Action adopted in 2001, an instrument on marking and tracing adopted in 2005, and the Firearms Protocol to the convention against transnational organised crime. There is no global treaty banning or capping small arms, and there will not be one, because these weapons are legitimate tools of policing and defence. The realistic agenda is therefore control rather than prohibition: marking and record keeping so that weapons can be traced, physical security of stockpiles so that they are not looted, and action against #illicit_trafficking. This work is unglamorous and it saves a great many lives. It also links disarmament directly to #sustainable_development, because armed violence destroys development gains faster than aid can build them. 8. The new frontiers: autonomous weapons, outer space and cyber operations Every generation of disarmament law has been built after the weapon arrived. The Hague rules came after the machine gun. The Geneva Protocol came after gas. Nuclear arms control came after Hiroshima. The current generation faces a choice: whether to keep following the same pattern, or to regulate before the harm. 8.1 Autonomous weapons systems An autonomous weapon system is one that, once activated, can select and engage targets without further human intervention. The debate is not about science fiction robots. It is about a design decision that is technically available now: whether a machine may apply lethal force on the basis of sensor data and software processing, rather than on a specific human decision about a specific target. The objections fall into four groups. Legal objections concern the law of armed conflict. Distinction, proportionality and precaution are judgements about context. A commander asks whether a person carrying a rifle is a combatant or a farmer protecting livestock, whether an attack will cause excessive harm relative to the military advantage, whether circumstances have changed since the order was given. It is contested whether software can make these judgements reliably in unstructured environments. Moral objections concern the value of human life. Many argue that a decision to kill a person must be taken by a person, and that delegating it to an algorithm treats human beings as data points to be processed. Accountability objections concern responsibility. If an autonomous system commits an unlawful killing, who is answerable? The commander who deployed it, the programmer, the manufacturer, the procurement officer, or nobody? A responsibility gap is not acceptable in a legal system. Security objections concern escalation. Autonomous systems interacting at machine speed could produce escalation faster than humans can intervene, in the same way that automated trading has produced flash crashes in financial markets. The negotiations sit in two places. Inside the CCW, a Group of Governmental Experts on emerging technologies in the area of lethal #autonomous_weapons has met since 2017 and continued through sessions in March and August 2026, working on a set of elements for a possible instrument. Because the CCW works by consensus, a small number of states that are developing these systems can block a binding outcome. Outside the CCW, the General Assembly has taken up the issue directly. In late 2025 it adopted a resolution on autonomous weapons systems by 164 votes to 6, with 7 abstentions, calling for work towards the elements of an instrument with a view to future negotiation. The Secretary-General and the President of the International Committee of the Red Cross have jointly called for a legally binding instrument that prohibits systems which cannot be used with #meaningful_human_control and regulates all others. That two-tier approach, prohibitions plus regulation, is now the central proposal on the table. The parallel with the mine ban process is obvious, and it may repeat. If the consensus forum stays blocked, a coalition of states may take the issue to the General Assembly and negotiate a treaty outside it. Such a treaty would not bind the main developers, at least not at first. Whether it would still shape behaviour is exactly the question raised by the nuclear prohibition treaty. 8.2 Outer space The 1967 Outer Space Treaty is the foundation. It bans placing nuclear weapons or other weapons of mass destruction in orbit, on celestial bodies or otherwise stationing them in #outer_space, and it reserves the Moon and other bodies for peaceful purposes. It does not ban conventional weapons in orbit, and it does not address anti-satellite systems based on Earth. The gap is now dangerous. Modern societies depend on satellites for navigation, timing, banking, agriculture, weather and communications, and modern militaries depend on them for early warning and command. Destructive anti-satellite tests conducted by several states have generated debris fields that threaten every user of low Earth orbit, including the testing state itself. Efforts to close the gap have taken two forms. One is a proposed treaty on the prevention of the placement of weapons in outer space, promoted by Russia and China, which Western states criticise because it lacks verification and says nothing about ground-based anti-satellite weapons. The other is a bottom-up approach based on norms of responsible behaviour, including a growing political commitment by a number of states not to conduct destructive direct-ascent anti-satellite missile tests. Neither approach has produced binding universal law. #Space_security remains the clearest example of a domain where dependence has grown much faster than regulation. 8.3 Cyber operations There is no cyber disarmament treaty, and there probably cannot be one in the classic sense, because the weapon is code and code cannot be counted, stockpiled or destroyed in a verifiable way. What exists instead is a slow process of norm building. Groups of governmental experts and an open-ended working group at the United Nations have agreed, in political rather than binding terms, that international law including the Charter applies in cyberspace, that states should not knowingly allow their territory to be used for internationally wrongful acts, and that states should not attack critical infrastructure such as hospitals and power grids in peacetime. The difficulties are severe. Attribution is technically hard and politically contested. Capabilities are held by criminals and private actors as well as states. The same tool can be used for espionage, which is broadly tolerated, and for sabotage, which is not. #Cyber_security law is therefore likely to develop through the interpretation of existing rules and through practice, rather than through a grand convention. 9. Compliance, verification and enforcement A treaty is a promise. The interesting question is what happens when the promise is broken. 9.1 The verification ladder Disarmament instruments can be ranked by how far they go in checking behaviour. At the bottom sit treaties with no verification at all, such as the Biological Weapons Convention. Trust is the only mechanism. Above them sit treaties with declarations and reporting, such as the Arms Trade Treaty and the mine ban convention, where states describe their own behaviour and other states and civil society check the plausibility of what they say. Higher still sit treaties with technical monitoring by an international body, such as the IAEA safeguards system and the test ban monitoring network. At the top sit treaties with intrusive on-site inspection, including challenge inspection at short notice. The Chemical Weapons Convention is the leading example, and New START's inspection regime was another until it stopped. The clear pattern is that the strength of verification tracks the political will that existed when the treaty was made. Verification is not a technical add-on. It is the price states are willing to pay for assurance, and it can be withdrawn when the politics change. 9.2 Enforcement and its limits #Enforcement is the weakest part of the architecture. The Security Council can act under Chapter VII of the Charter, but the five permanent members hold vetoes, and several of the hardest #compliance cases involve those members or their close partners. #Sanctions can raise the cost of violation, as they have in the case of North Korea, but they have not reversed a nuclear programme once it matured. International criminal law adds another layer. The use of chemical weapons, the use of poison, and attacks directed at civilians are war crimes, and individuals can in principle be prosecuted. In practice, prosecutions are rare and slow. This leaves reputational and relational costs as the main enforcement mechanism in most cases. States that violate arms control treaties lose credibility in other negotiations, face export restrictions, and give their rivals a justification for building up. Those costs are real, and they explain a great deal of compliance. They are also clearly insufficient when a state believes its survival is at stake. 9.3 The compliance puzzle Why do states obey these treaties at all, given weak enforcement? The literature offers four answers, and all four are probably partly right. Rational calculation. States obey when the benefits of a stable, predictable environment exceed the value of the prohibited capability. This explains why most states never wanted chemical or biological weapons in the first place, and why nuclear #modernisation continues among those who believe deterrence is essential. Reciprocity. States obey because the treaty gives them reciprocal restraint from others. This is why a violation by one party can unravel an entire regime. Norms and identity. States obey because a rule has become part of what a decent state does. This is the mechanism behind the taboo on chemical weapons and the near-taboo on nuclear testing. Norms are slow to build and can be damaged quickly. Domestic politics. States obey because parliaments, courts, media and civil society hold governments to their promises. Where those institutions are strong, treaties are implemented. Where they are weak, treaties become paper. This is why #national_implementation, meaning the domestic laws and licensing systems that give effect to a convention, matters far more than most students expect. 10. Regional pictures 10.1 The Middle East The region contains one undeclared nuclear-armed state, one state whose nuclear programme has repeatedly been the subject of crisis and negotiation, several states with past chemical weapons programmes, and a long record of missile proliferation. It is also the region where the proposed zone free of weapons of mass destruction has been discussed for fifty years without result. Any serious progress would require confidence-building between states that do not recognise each other, which is why almost every proposal has failed at the first step. 10.2 The Korean Peninsula North Korea is the only state to have joined the NPT, announced withdrawal, and then tested nuclear weapons. It has conducted six nuclear tests, developed intercontinental-range missiles, and enshrined its nuclear status in law. Diplomacy has passed through agreed frameworks, six-party talks and summit meetings without producing denuclearisation. The #Korean_Peninsula is the clearest evidence that a non-proliferation regime cannot reverse a determined programme once weapons exist. It also raises the hardest policy question in the field: whether to keep insisting on complete denuclearisation as a precondition, or to pursue risk reduction and arms control with a state whose weapons are not going away. 10.3 South Asia India and Pakistan both tested nuclear weapons in 1998 and neither joined the NPT. They face each other across a disputed border, with short missile flight times, conventional imbalance, cross-border militancy and repeated crises. #South_Asia is where the risk of nuclear use through escalation is arguably highest, and where formal arms control is thinnest. Existing measures are modest: an agreement not to attack each other's nuclear facilities, notification of missile tests, and some hotlines. Building on those modest measures is more realistic than pursuing membership of treaties that neither state will join. 10.4 Africa and the global south #Africa has been an important actor in disarmament, not a passive object of it. South Africa dismantled its arsenal. The Pelindaba Treaty made the continent a nuclear-weapon-free zone. African states have been among the strongest supporters of humanitarian disarmament, because they carry a heavy burden of landmine contamination, explosive remnants of war and small arms violence. This points to a general truth that is often missed in the strategic literature. For most of the world's states, disarmament is not an abstract debate about deterrence. It is a development issue, a public health issue and a governance issue. 11. Civil society, victims and the human dimension Three features of modern disarmament law would surprise a diplomat from 1950. The first is the role of #civil_society. The mine ban treaty, the cluster munitions convention and the nuclear prohibition treaty were all produced by partnerships between small and medium states and organised campaigns. Two of those campaigns, the International Campaign to Ban Landmines and ICAN, received Nobel Peace Prizes. The #ICRC has played a distinctive role as a guardian of international humanitarian law with access to both governments and battlefields. Monitoring reports produced by non-governmental organisations are now the main public source of data on implementation. The second is the presence of survivors. Victim assistance provisions in the mine, cluster munitions and nuclear prohibition treaties create obligations towards people already harmed: medical care, rehabilitation, psychological support, economic inclusion, and participation in decisions that affect them. Survivors are treated as rights holders, not as objects of charity. This is a genuine innovation in the law of weapons, and it connects disarmament to disability rights and to #human_rights law more generally. The third is attention to #gender. Recent instruments recognise that armed violence affects women, men, girls and boys differently, that mine casualties are overwhelmingly male in some contexts because of the gendered division of labour, and that survivors' needs differ by sex. Security Council resolution 1325 and the wider women, peace and security agenda have also pushed for the participation of women in disarmament negotiations, where they remain underrepresented. 12. Discussion: what the evidence shows 12.1 The regimes work better than their reputation Public debate treats disarmament law as a failure because nuclear weapons still exist. Measured properly, the record is more complicated. Chemical weapons stockpiles declared by states parties have been destroyed. Nuclear warheads have fallen from roughly seventy thousand at the peak of the Cold War to around twelve thousand today. Nuclear testing has almost stopped. Blinding lasers were banned before use. Tens of millions of landmines have been destroyed, and land is being returned to communities every year. The number of states with nuclear weapons is nine, not thirty. None of this happened by accident. It happened because agreements were negotiated, verified and implemented. 12.2 The crisis is political, not legal Nothing in the legal architecture broke by itself. New START did not fail; it expired on a date fixed at signature, and it was not replaced because relations between the parties collapsed. The mine ban convention was not violated by the five European withdrawals; its exit clause was used exactly as written. The NPT review process has not produced a consensus document because states disagree, not because the procedure is defective. This is important for students to grasp, because it points to where repair work must be done. Better drafting will not fix a problem of political will. The trend of #treaty_withdrawal, in a period of rising #military_spending that reached roughly 2.7 trillion United States dollars worldwide in 2024, reflects a change in how states value restraint, not a change in the quality of the law. 12.3 The consensus forums are blocked The #Conference_on_Disarmament in Geneva, formally the world's single multilateral disarmament negotiating body, has not concluded a major treaty since the CTBT in 1996. It works by consensus, and consensus in a divided world means paralysis. The CCW is similarly stuck on autonomous weapons. The response, repeatedly, has been to move negotiations to forums where a majority can act: the General Assembly and its #First_Committee, or stand-alone diplomatic conferences. This produces treaties, but treaties without the states that matter most. That is the central trade-off of contemporary disarmament diplomacy, and there is no way around it. Either you accept a veto and get nothing, or you accept a partial treaty and hope the norm spreads. 12.4 Technology is outrunning law Autonomy, artificial intelligence, cyber capability, hypersonic delivery, dual-use biotechnology and the militarisation of space are all developing faster than the diplomatic process that would regulate them. The historical pattern suggests law will arrive after the first catastrophe. The purpose of preventive regulation, such as the ban on blinding lasers, is to break that pattern. It has been done once. It can be done again, but only if states decide to act before rather than after. 12.5 Norms decay if they are not maintained Perhaps the most important finding is that the moral achievements of disarmament are not permanent. The taboo on landmines took thirty years to build and is being weakened within a decade. The taboo on chemical weapons was strained by repeated use in Syria and by nerve agent assassinations. A norm is not a ratchet. It is a structure that requires maintenance, and maintenance means political cost, public pressure, #education and consistent behaviour by leading states. 13. Limitations of this study Three limits should be acknowledged. The article relies on open sources. Classified information about arsenals, compliance and covert programmes is not available, and estimates of warhead numbers and stockpiles carry uncertainty. The article is a synthesis, not new empirical #research. It does not test hypotheses statistically, and its causal claims about why states comply or defect are drawn from existing scholarship rather than from original data. The article is written at a moment of rapid change. Events after mid-2026, including the outcome of any talks following the expiry of New START, the first review conference of the nuclear prohibition treaty, and the CCW review conference on autonomous weapons, may change the picture significantly. 14. Recommendations The following proposals follow directly from the analysis. They are deliberately modest, because grand proposals in this field have a poor record. Rebuild the bilateral floor. The immediate priority is not a comprehensive treaty between Washington and Moscow but the restoration of a minimum: data exchanges, launch notifications, no interference with national technical means, and a working consultative channel. These measures are cheap, they do not require anyone to reduce anything, and they reduce the risk of catastrophic misreading. Separate risk reduction from disarmament. Progress towards elimination is blocked. Progress on reducing the risk of accidental or mistaken use is not, and it should be pursued on its own track, including agreements on human control over nuclear launch decisions and restraint in the use of artificial intelligence in nuclear command and control. Defend the exit clauses. Withdrawal from humanitarian treaties should be made politically expensive. States parties should require detailed justification, independent review of the claimed military necessity, and a standing obligation to report on the humanitarian consequences of any resumed use. Lawful exit should not be a quiet exit. Finish the autonomous weapons instrument. States should complete the elements of an instrument and move to negotiation, using the two-tier approach of prohibiting systems that cannot be used with meaningful human control and regulating the rest. If the consensus forum blocks this, a majority-based process should proceed. Give the Biological Weapons Convention institutional muscle. Even without a full verification protocol, the convention needs a properly funded implementation support unit, a mechanism for peer review of national biosafety and biosecurity, and a strengthened process for investigating alleged use. Invest in the boring parts. Stockpile security, weapons marking and tracing, national licensing systems, customs training and clearance operations save more lives per dollar than any summit. They also build the administrative capacity on which every future treaty will depend. Protect the evidence base. The Landmine Monitor, the Cluster Munition Monitor, the SIPRI databases and independent warhead estimates make the regimes measurable. Funding them is a form of arms control. Teach the subject seriously. Disarmament #education is treated in several treaties as an obligation, not a luxury. The generation now in university will inherit a weaker architecture than the one their teachers inherited, and they will need to know how it was built in order to rebuild it. 15. Conclusion The conventions on peace, security and disarmament are one of the more remarkable inventions of the modern international system. They took the most jealously guarded area of state sovereignty, the right to arm, and made it a subject of negotiated rules, international inspection and public accountability. They did so imperfectly, unevenly and with many failures. They also destroyed chemical arsenals, cut nuclear warhead numbers by more than eighty per cent from their peak, ended nuclear testing in practice, cleared land for farmers, and made the possession of certain weapons a mark of shame rather than of strength. The middle of the 2020s is the hardest test this system has faced. The last bilateral nuclear limits have lapsed. States are leaving humanitarian treaties for the first time. The main review process for the nuclear order has failed again. Military spending is at record levels, and new technologies are arriving faster than the rules to govern them. None of that means the project has failed. It means the project is unfinished and currently unfunded in political terms. The law has not disappeared. The treaties are still there, with their texts, their institutions, their inspectors and their monitoring systems. What has weakened is the belief among powerful states that restraint serves their interest. Beliefs can change again. They changed after 1945, after 1962, after 1986, and after the images of mine victims reached the public in the 1990s. Change came each time from a combination of fear, evidence, leadership and public pressure, and each time it was the patient work of #multilateralism, not any single dramatic event, that turned that pressure into binding rules. The task for the next generation is not to invent a new system. It is to defend a good one that is being neglected, to finish the parts that were left incomplete, and to extend it to the weapons of the coming decades before those weapons write their own history in casualties. #Conflict_prevention through law is slow, unglamorous and often frustrating work. It is also the only method humanity has found that has ever measurably reduced the number of weapons in the world. 16. Study questions for students Distinguish arms control from disarmament, and identify which standard should be used to evaluate the Non-Proliferation Treaty. The five European withdrawals from the mine ban convention were lawful. Does lawful withdrawal damage a norm more or less than an unlawful violation? Argue both sides. The Comprehensive Nuclear-Test-Ban Treaty is not in force, yet testing has almost stopped. What does this tell you about the relationship between legal force and normative effect? Should a treaty on autonomous weapons be negotiated in a consensus forum that includes the main developers, or in a majority forum that excludes them? Defend your choice with reference to the mine ban and nuclear prohibition precedents. If you were advising a small state with no military power, what disarmament priorities would you pursue in the next five years, and why? References Acheson, R. (2021). Banning the Bomb, Smashing the Patriarchy. Lanham: Rowman and Littlefield. Arms Control Association (2026). Countries Leave Mine Ban Treaty. Arms Control Today, January 2026. Washington DC. Arms Control Association (2026). New START at a Glance. Fact Sheet. Washington DC. Bode, I. and Huelss, H. (2022). Autonomous Weapons Systems and International Norms. Montreal: McGill-Queen's University Press. Boulanin, V., Bruun, L. and Goussac, N. (2021). Autonomous Weapon Systems and International Humanitarian Law: Identifying Limits and the Required Type and Degree of Human-Machine Interaction. Stockholm: Stockholm International Peace Research Institute. Bulletin of the Atomic Scientists (2026). Doomsday Clock Statement. Chicago: Bulletin of the Atomic Scientists. Congressional Research Service (2026). Extension of New START Central Limits: Overview of the Expert Debate. CRS Insight IN12640. Washington DC: Library of Congress. Council on Foreign Relations (2026). Nukes Without Limits: A New Era After the End of New START. New York: Council on Foreign Relations. Federation of American Scientists (2026). The Expiration of New START and What It Means For Us All. Washington DC: Federation of American Scientists. Futter, A. (2021). The Politics of Nuclear Weapons. Second edition. Bristol: Bristol University Press. Gibbons, R. D. (2022). The Hegemon's Tool Kit: US Leadership and the Politics of the Nuclear Nonproliferation Regime. Ithaca: Cornell University Press. House of Commons Library (2026). The Ottawa Treaty and the Convention on Cluster Munitions. Research Briefing CBP-10195. London: UK Parliament. Human Rights Watch (2025). Five European States Withdraw from Mine Ban Treaty. New York: Human Rights Watch. Human Rights Watch and International Human Rights Clinic, Harvard Law School (2025). A Hazard to Human Rights: Autonomous Weapons Systems and Digital Decision-Making. New York: Human Rights Watch. International Campaign to Abolish Nuclear Weapons (2024). Surge: 2023 Global Nuclear Weapons Spending. Geneva: ICAN. International Campaign to Ban Landmines and Cluster Munition Coalition (2024). Landmine Monitor 2024. Geneva: ICBL-CMC. International Campaign to Ban Landmines and Cluster Munition Coalition (2025). Cluster Munition Monitor 2025. Geneva: ICBL-CMC. International Committee of the Red Cross (2021). ICRC Position on Autonomous Weapon Systems. Geneva: ICRC. International Atomic Energy Agency (2024). Safeguards Statement and Annual Report. Vienna: IAEA. Kmentt, A. (2021). The Treaty Prohibiting Nuclear Weapons: How It Was Achieved and Why It Matters. Abingdon: Routledge. Kristensen, H. M., Korda, M., Johns, E. and Knight, M. (2025). Nuclear Notebook. Bulletin of the Atomic Scientists, volume 81. Chicago. Library of Parliament (2026). The Future of the Ottawa Treaty. Publication 2025-20-E. Ottawa: Parliament of Canada. Lieber Institute for Law and Warfare (2026). Assessing the Ottawa Anti-Personnel Mine Convention Withdrawals. West Point: United States Military Academy. Nuclear Threat Initiative (2026). The End of New START: From Limits to Looming Risks. Washington DC: NTI. Onderco, M. (2021). Networked Nonproliferation: Making the NPT Permanent. Stanford: Stanford University Press. Organisation for the Prohibition of Chemical Weapons (2024). Report of the OPCW on the Implementation of the Chemical Weapons Convention. The Hague: OPCW. Small Arms Survey (2024). Small Arms Survey 2024. Geneva: Graduate Institute of International and Development Studies. Stockholm International Peace Research Institute (2024). SIPRI Yearbook 2024: Armaments, Disarmament and International Security. Oxford: Oxford University Press. Stockholm International Peace Research Institute (2025). SIPRI Yearbook 2025: Armaments, Disarmament and International Security. Oxford: Oxford University Press. Stockholm International Peace Research Institute (2025). Trends in World Military Expenditure, 2024. SIPRI Fact Sheet. Stockholm: SIPRI. Stop Killer Robots (2025). Autonomous Weapons Systems: Key Issues and the Path to a Treaty. Policy Brief. Geneva: Stop Killer Robots. United Nations (2023). A New Agenda for Peace. Our Common Agenda Policy Brief 9. New York: United Nations. United Nations General Assembly (2024). Pact for the Future. Resolution A/RES/79/1. New York: United Nations. United Nations General Assembly (2025). Lethal Autonomous Weapons Systems. Draft Resolution A/C.1/80/L.41. New York: United Nations. United Nations Office for Disarmament Affairs (2026). Documents of the Eleventh Review Conference of the Parties to the Treaty on the Non-Proliferation of Nuclear Weapons. New York: UNODA. Vienna Center for Disarmament and Non-Proliferation (2026). End of New START: Short and Medium Term Options. Vienna: VCDNP. Hashtags #PeaceAndSecurity #DisarmamentConventions #ArmsControlLaw #NuclearDisarmament #NonProliferation #HumanitarianDisarmament #InternationalLawStudies #GlobalSecurityStudies #TreatyCompliance #WeaponsOfMassDestruction #PeaceStudies #ConflictPrevention #DisarmamentEducation #SecurityStudies #STULIB

The Convention against Discrimination in Education, adopted by the General Conference of #UNESCO in Paris on 14 December 1960 and in force since 22 May 1962, is the first legally binding international treaty devoted entirely to the #right_to_education. It is also the first universal instrument to state clearly that separating learners into different schools or systems on grounds such as race, colour, sex, language, religion, political opinion, national or social origin, economic condition or birth is, as a rule, unlawful. This article offers a structured, student-facing but scholarly examination of the Convention. It traces the historical conditions that produced the treaty, reconstructs its legal architecture article by article, and explains the difficult compromises embedded in Article 2, which permits certain separate systems under strict conditions. The article then situates the Convention within the wider architecture of international human rights law, including the International Covenant on Economic, Social and Cultural Rights, the Convention on the Rights of the Child and the Convention on the Rights of Persons with Disabilities. It reviews the treaty monitoring machinery, in particular the periodic consultations of Member States, the most recent of which covered the period 2021 to 2024 and attracted a record number of national reports. Finally, the article tests the Convention against six contemporary fault lines: disability and the persistence of #segregation, the exclusion of girls and women, the schooling of national and linguistic minorities, the education of refugees and displaced learners, the growth of private and commercial schooling, and the emerging inequalities produced by digital technology and climate displacement. The analysis concludes that the Convention remains conceptually powerful and legally sound, but that its enforcement architecture is weak, its coverage of disability is a historical gap, and its promise of #equality_of_opportunity is repeatedly defeated at the point of implementation. A set of reform proposals is offered, addressed to states, to UNESCO, to national courts and to the research community. Keywords: right to education; non-discrimination; segregation; inclusive education; UNESCO; international human rights law; equality of opportunity; Sustainable Development Goal 4 1. Introduction Few sentences in international law are as short and as demanding as the one that opens Article 26 of the Universal Declaration of Human Rights. Everyone has the right to education. The sentence is easy to read and extremely hard to honour. It says nothing about who pays, who teaches, who decides what counts as education, or what happens when a state provides schools for some children and not for others. It says nothing about whether a state may build one school for one group and a different school for a different group, and then call the arrangement equal. The Convention against Discrimination in Education, usually shortened in the literature to the acronym CADE, was drafted precisely to answer those harder questions. Adopted at the eleventh session of the General Conference of UNESCO in Paris in December 1960, and entering into force in May 1962, it was the first treaty in history dedicated wholly and exclusively to education as a legal entitlement rather than as a policy aspiration. It is, for that reason alone, a foundational document. It is also, for reasons this article will explore, a document that is frequently praised and rarely enforced. The purpose of this article is threefold. First, it seeks to explain, in language accessible to students of law, education, development studies and public policy, what the Convention actually says and why it says it in the way that it does. Textbooks tend to summarise the treaty in a single paragraph, which does the instrument an injustice, because its most interesting provisions are precisely the ones that resist summary. Second, the article seeks to place the Convention in its historical and political context, since a treaty adopted in 1960 cannot be understood without reference to #decolonisation, to the American civil rights struggle, and to the international campaign against #apartheid. Third, and most importantly, the article seeks to assess whether the Convention still works. Sixty-five years after its adoption, hundreds of millions of children remain outside school, girls are barred from secondary education in at least one country by state policy, Roma children in Europe continue to be placed in separate and inferior classrooms, and learners with disabilities are still routinely educated in settings designed to keep them away from their peers. A treaty that prohibits all of this, and yet coexists with all of this, invites a serious question about the relationship between #international_law and social reality. The article proceeds as follows. Section 2 sets out the historical background. Section 3 explains the methodology and the limits of the study. Section 4 reconstructs the legal architecture of the Convention. Section 5 focuses specifically on the treatment of #segregation, which is the analytical heart of the instrument. Section 6 situates the Convention within the wider human rights system. Section 7 examines the monitoring machinery. Section 8 tests the Convention against six contemporary challenges. Section 9 discusses the findings. Section 10 acknowledges limitations. Section 11 offers recommendations. Section 12 concludes. A word on why this matters for students in particular. It is tempting to treat #treaty_obligations as remote technical matters, the business of diplomats and specialist lawyers. That view is mistaken. The Convention is one of the few instruments in international law that speaks directly to the daily experience of the classroom, to who sits where, to who is admitted and who is turned away, to which language a child may speak and which curriculum a child may follow. Understanding it is a form of literacy about one of the most consequential institutions in modern life. 2. Historical and Political Background 2.1 The post-war settlement and the invention of educational rights The Convention did not emerge from nowhere. It was the product of a specific moment in which three currents converged. The first current was the post-war construction of the international human rights system. The Charter of the United Nations, signed in 1945, committed member states to promoting universal respect for #human_rights without distinction as to race, sex, language or religion. The Universal Declaration of Human Rights, adopted in 1948, translated that commitment into a catalogue of rights, of which education was one. UNESCO itself was founded in 1945 with a constitutional mandate that included advancing the ideal of equality of educational opportunity for all. The Convention of 1960 can be read as UNESCO discharging that constitutional mandate in binding legal form. The second current was #decolonisation. Between 1945 and 1960, the map of the world was redrawn. In 1960 alone, seventeen African states became independent. The new states arrived at the United Nations with a set of grievances that were, in significant part, educational. Colonial education systems had been explicitly stratified. They had provided one kind of schooling for settler populations and another, shorter and more limited, for colonised populations. They had frequently suppressed local languages. They had produced literacy rates that were, in many territories, catastrophically low. For the new states, a treaty prohibiting educational discrimination was not an abstract principle. It was a description of the injustice they had just escaped. The third current was the domestic struggle against racial #segregation, above all in the United States and in South Africa. In 1954, the Supreme Court of the United States decided Brown v. Board of Education, holding that separate educational facilities are inherently unequal. The reasoning of that judgment, that separation carries a message of inferiority which damages the learner regardless of the material quality of the facilities provided, entered international legal discourse with remarkable speed. At the same time, the South African government was constructing the Bantu Education system, a deliberate architecture of #apartheid schooling designed to prepare Black South Africans for subordinate positions. The international community was, in short, watching two contrasting demonstrations of what educational separation meant in practice. 2.2 The drafting process Work on the instrument began at the tenth session of the UNESCO General Conference, which decided that the question of discrimination in education should be addressed both by an international convention and by a recommendation to Member States. This dual approach is important and often misunderstood. UNESCO adopted, on the same day in December 1960, two texts with almost identical substantive content: the Convention, which is a binding treaty for states that ratify it, and the Recommendation against Discrimination in Education, which is a non-binding instrument carrying political and moral force. The dual approach was designed to accommodate states which, for constitutional or federal reasons, could not immediately assume binding obligations, while still drawing them into the normative framework and, critically, into the reporting system. The drafting itself involved a series of hard bargains. Delegations disagreed about whether religious schools should be permitted, whether single-sex schooling was discriminatory, whether private schools should be allowed to exist alongside public ones, and whether minority communities had a right to run their own educational institutions. The compromises reached on these questions are recorded in Article 2 and Article 5, and they explain why the Convention is not the simple prohibition that its title might suggest. 2.3 Entry into force and subsequent development The Convention entered into force on 22 May 1962. A Protocol Instituting a Conciliation and Good Offices Commission, intended to provide a mechanism for resolving disputes between States Parties, was adopted on 10 December 1962 and entered into force on 24 October 1968. By the middle of the 2020s, around 110 states had become parties to the Convention, a substantial figure but far short of universal #ratification, and notably short of the near-universal adherence enjoyed by the Convention on the Rights of the Child. The Convention has since been repeatedly reaffirmed. #UNESCO describes it as the cornerstone of the #Education_2030 agenda and as a central legal tool for advancing Sustainable Development Goal 4, which commits states to ensuring inclusive and equitable quality education and promoting lifelong learning opportunities for all. The instrument has also been drawn into the more recent conversation about what education law should look like in the twenty-first century, a conversation that UNESCO itself launched through its Initiative on the Evolving Right to Education, and which has produced a substantial body of reflection on whether the treaty framework of the 1960s is adequate to present conditions (UNESCO, 2023b; Beiter, 2023). 3. Methodology This article adopts a doctrinal and analytical approach, supplemented by policy analysis. The doctrinal component consists of a close reading of the text of the Convention and of the associated Recommendation, together with the interpretive materials produced by UNESCO and by United Nations human rights treaty bodies. The policy component draws on UNESCO monitoring reports, in particular the synthesis reports of the tenth and eleventh periodic consultations of Member States, and on the Global Education Monitoring Report series, which provides the most authoritative statistical picture of educational exclusion available. The comparative component draws on the jurisprudence of the European Court of Human Rights concerning the schooling of Roma children, which constitutes the richest body of judicial reasoning anywhere in the world on the specific question of whether de facto educational separation amounts to prohibited discrimination. Although the European Court does not apply the UNESCO Convention directly, its reasoning under Article 14 of the European Convention on Human Rights, read together with Article 2 of Protocol No. 1, addresses the same conceptual problem and provides a useful test of how the underlying principle operates when courts are asked to enforce it. Three limitations should be stated at the outset. First, this is not an empirical study. It does not generate new data on educational outcomes. Second, the analysis relies on state self-reporting for much of its evidence about implementation, and self-reporting is systematically optimistic. Third, the article is written in English and draws principally on scholarship available in English, which introduces a linguistic and regional bias that the author acknowledges without being able to eliminate. These limitations are revisited in Section 10. 4. The Legal Architecture of the Convention 4.1 Article 1: the definition of discrimination Everything in the Convention flows from its opening definition. Article 1 provides that the term discrimination includes any distinction, exclusion, limitation or preference which, being based on race, colour, sex, language, religion, political or other opinion, national or social origin, economic condition or birth, has the purpose or effect of nullifying or impairing equality of treatment in education. Four features of this definition deserve close attention, because they carry the analytical weight of the entire treaty. The list of prohibited grounds is broad but closed. The Convention names eleven grounds. This was, in 1960, an unusually generous list. It notably included economic condition, which many contemporary instruments omitted, and birth, which reaches illegitimacy and caste-adjacent forms of exclusion. But the list is exhaustive rather than illustrative. It does not include disability. It does not include sexual orientation or gender identity. It does not include migration status as such, although national origin covers a great deal of the same territory. This closed list is one of the most significant weaknesses of the instrument, and Section 8 returns to it. The definition captures purpose or effect. This is the single most important phrase in the treaty. A measure need not be adopted with discriminatory intent to be unlawful. It is enough that the measure has a discriminatory effect. This means the Convention prohibits what modern lawyers call indirect discrimination: a rule that appears neutral on its face but that operates disproportionately against a protected group. A requirement that pupils present a birth certificate to enrol may be neutral in form, but if it excludes undocumented children in practice, it is caught by Article 1. This is a remarkably modern piece of drafting for 1960, and it anticipates by decades the reasoning that European courts would later adopt in the Roma education cases. The definition enumerates specific forms of discrimination. Article 1 continues by listing four particular manifestations. These are: depriving any person or group of access to education of any type or at any level; limiting any person or group to education of an inferior standard; establishing or maintaining separate educational systems or institutions for persons or groups, subject to the exceptions in Article 2; and inflicting on any person or group conditions which are incompatible with the dignity of the human person. The third item is the anti-segregation clause. The fourth is a dignity clause of considerable but underused potential. Discrimination is defined in relation to equality of treatment. The reference point is not formal identity of treatment but equality of treatment properly understood. This distinction, which the Convention leaves implicit, is the foundation of the modern doctrine of substantive #equality. Treating a deaf child exactly as one treats a hearing child, by offering identical instruction in spoken language, is formally equal and substantively discriminatory. 4.2 Article 2: the permitted separations Article 2 is where the Convention becomes complicated, and where students most often lose their footing. Having declared in Article 1 that establishing or maintaining separate educational systems is a form of discrimination, the treaty immediately identifies three situations in which separation is not deemed to constitute discrimination. Single-sex systems. Separate educational systems or institutions for pupils of the two sexes are permitted if they offer equivalent access to education, provide teaching staff with qualifications of the same standard, provide school premises and equipment of the same quality, and afford the opportunity to take the same or equivalent courses of study. Religious or linguistic systems. Separate systems or institutions may be maintained for religious or linguistic reasons, offering an education which is in keeping with the wishes of the pupils' parents or legal guardians, provided that participation in such systems or attendance at such institutions is optional and that the education provided conforms to the standards laid down or approved by the competent authorities. Private institutions. The establishment or maintenance of private educational institutions is permitted, provided the object is not to secure the exclusion of any group but to provide educational facilities in addition to those provided by the public authorities, and provided the institutions are conducted in accordance with that object and conform to approved standards. The logic of Article 2 is best understood as a set of guardrails rather than as a set of loopholes. Each permitted separation is subject to conditions of equivalence, voluntariness or supplementarity. Separation is tolerated when it is chosen, when it is genuinely equal in resources and outcomes, and when it does not have exclusion as its object. Separation is prohibited when it is imposed, when it delivers inferior provision, or when its purpose or effect is to keep a group out. That said, Article 2 has been the source of persistent difficulty. Three problems recur. The first is that the equivalence conditions attached to single-sex education are demanding on paper and rarely audited in practice. Where girls' schools are systematically under-resourced, where they lack science laboratories, or where they are staffed by less qualified teachers, the conditions of Article 2 are not met, and the separation ceases to be lawful under the Convention. Yet this analysis is almost never performed by national authorities. The second is that the voluntariness condition attached to religious and linguistic separation is easily hollowed out. Formally optional systems may be practically compulsory where social pressure, geography or the absence of an alternative leaves families with no meaningful choice. A choice that exists only on paper is not a choice. The third is that the private education exception, drafted in an era of small, charitable and religious foundations, now has to be applied to a global market in commercial schooling of a kind the drafters did not contemplate. This is one of the most contested issues in contemporary education law and is discussed in Section 8.5. 4.3 Articles 3 and 4: the two kinds of state obligation The Convention imposes two analytically distinct sets of duties, and it is worth keeping them apart. Article 3 imposes negative and immediate duties. States Parties undertake to abrogate statutory provisions and administrative instructions and to discontinue administrative practices which involve discrimination in education. They undertake to ensure, by legislation where necessary, that there is no discrimination in the admission of pupils. They undertake not to permit differences of treatment between nationals in the matter of school fees, scholarships and other assistance, except on the basis of merit or need. They undertake not to permit restrictions or preferences in public assistance to educational institutions based solely on the ground that pupils belong to a particular group. And they undertake to give foreign nationals resident in their territory the same access to education as their own nationals. These are obligations of result, and they are immediate. A state cannot plead lack of resources as a justification for maintaining a discriminatory admissions rule, because repealing a discriminatory rule costs nothing. Article 4 imposes positive and progressive duties. States Parties undertake to formulate, develop and apply a national policy which will tend to promote #equality_of_opportunity and of treatment in education. That policy must, in particular, make primary education free and compulsory, make secondary education in its different forms generally available and accessible to all, make higher education equally accessible to all on the basis of individual capacity, assure compliance with compulsory attendance requirements, ensure that standards of education are equivalent across public institutions of the same level, encourage and intensify the education of persons who have not received or completed primary education, and provide training for the teaching profession without discrimination. Article 4 is the more ambitious provision, and it is inherently programmatic. It presupposes resources, planning and time. The distinction between the immediate obligations of Article 3 and the progressive obligations of Article 4 mirrors the structure later adopted in the International Covenant on Economic, Social and Cultural Rights, and it is central to any honest assessment of what the Convention requires of a low-income state. It is worth pausing on the phrase #free_primary_education. The obligation to make primary education free and #compulsory_education is unambiguous and immediate in its formulation. There is no corresponding obligation to make secondary or pre-primary education free. This gap has become one of the most active areas of reform advocacy in contemporary right-to-education scholarship, and it is the subject of a sustained argument by Sheppard (2022), who contends that the omission is an artefact of the resource constraints of the 1950s rather than a principled limit, and that an optional protocol to the Convention on the Rights of the Child should extend the free-education guarantee upward and downward. That argument has since found institutional traction, with the United Nations Human Rights Council establishing an open-ended intergovernmental working group to consider precisely such an instrument (UN Human Rights Council, 2024). 4.4 Article 5: the purposes of education and the rights of parents and minorities Article 5 is the values clause of the Convention, and it contains three distinct commitments. First, it declares that education shall be directed to the full development of the human personality and to the strengthening of respect for human rights and fundamental freedoms, and that it shall promote understanding, tolerance and friendship among all nations, racial or religious groups. Education, in other words, is not merely a service to be distributed. It has a purpose, and that purpose is itself a matter of legal obligation. This provision is the ancestor of every subsequent commitment to human rights education and global citizenship education. Second, it protects the liberty of parents to choose for their children institutions other than those maintained by the public authorities, and to ensure the religious and moral education of their children in conformity with their own convictions. This is a genuine right, but it is not unlimited: the chosen institutions must conform to minimum standards, and no person or group may be compelled to receive religious instruction inconsistent with their convictions. Third, and most significantly for the purposes of this article, Article 5 recognises the right of members of national minorities to carry on their own educational activities, including the maintenance of schools and, depending on the educational policy of each state, the use or teaching of their own language. This right, however, is hedged with three conditions. It must not be exercised in a manner which prevents members of minorities from understanding the culture and language of the community as a whole. The standard of education must not be lower than the general standard. And attendance at such schools must be optional. The tension in this provision is evident and instructive. #minority_rights in education pull in two directions. A minority community may reasonably wish its children to be educated in its own language and its own cultural tradition. Yet a separate minority school system may also function as an instrument of exclusion, leaving minority children without the linguistic and social capital needed to participate in the wider society, and leaving majority children without any experience of their minority compatriots. Article 5 attempts to hold both concerns at once. Whether it succeeds is a question that Section 8.3 revisits in the context of #mother_tongue_education. 4.5 Articles 6 to 19: implementation, reporting and dispute settlement The remaining provisions are procedural, and they are, from an enforcement perspective, the weakest part of the treaty. Article 6 commits States Parties to pay attention to any recommendations that the General Conference of UNESCO may adopt on the subject. Article 7, which is the core #monitoring provision, requires States Parties to give information in periodic reports to the General Conference on the legislative and administrative provisions they have adopted and other action taken to apply the Convention, including the formulation and development of the national policy required by Article 4, and including the results achieved and the obstacles encountered. Article 8 provides that disputes between States Parties concerning interpretation or application, if not settled by negotiation, shall at the request of the parties be referred to the International Court of Justice, failing other means of settlement. Two structural observations follow. First, there is no individual complaints mechanism. A child excluded from school on discriminatory grounds cannot bring a claim under this Convention. This is a profound limitation. The Convention creates rights but does not, in itself, create remedies for the rights-holder. It relies on the reporting system, on inter-state dispute settlement which has never been used in practice, and on the Protocol of 1962, whose Conciliation and Good Offices Commission has been almost entirely dormant. Second, the Convention admits no reservations. This is unusual and important. Unlike many human rights treaties, where states routinely ratify while carving out the provisions they dislike, a state that ratifies CADE accepts the whole instrument. The prohibition of reservations is a genuine strength, and it means that the ratifying state cannot lawfully claim that some part of the treaty does not apply to it. The corollary, of course, is that states which find the treaty inconvenient simply do not ratify it, which is one plausible explanation for the fact that adherence has stalled well below universal levels. 5. Segregation as the Analytical Core 5.1 What the Convention actually prohibits The heart of the Convention is the proposition, stated in Article 1 and qualified in Article 2, that establishing or maintaining separate educational systems or institutions for persons or groups is discrimination. It is worth being precise about what this proposition does and does not assert. It does not assert that all separation is unlawful. Article 2 makes that clear. What it asserts is that separation is presumptively discriminatory, and that the burden falls on the state to bring the separation within one of the defined exceptions by demonstrating equivalence, voluntariness or supplementarity. This is a burden-shifting rule, and it is a powerful one. In the absence of Article 1, a state defending a separate school system would only have to show that it had no discriminatory intent. Under the Convention, it must show that the separation satisfies the conditions of Article 2. 5.2 The separate but equal problem The Convention was drafted in the shadow of a doctrine that the international community was in the process of rejecting. The doctrine of separate but equal held that separation was permissible provided the separate facilities were equal. The American Supreme Court had repudiated it in 1954, on the ground that in the field of public education the doctrine of separate but equal has no place, because separation itself generates a feeling of inferiority that affects motivation to learn in a way unlikely ever to be undone. The UNESCO Convention occupies a more ambivalent position. It does not adopt the full logic of Brown, which would have required an absolute prohibition on separation by protected characteristic. Instead, it retains a conditional permission for separation in the three cases identified in Article 2. The compromise was politically necessary. Many states, including many democratic states, operated single-sex schools and faith schools and had no intention of abolishing them. A treaty that outlawed them would not have been adopted. But the compromise carries a cost, and the cost is analytical. Once a treaty admits that separation can be equal, it must then police the conditions of equality, and policing those conditions requires exactly the kind of resourced, independent, adversarial scrutiny that the Convention's monitoring machinery does not provide. The result is a familiar pattern: separation is maintained, equality is asserted, and nobody with the power to compel disclosure ever checks. 5.3 De jure and de facto segregation A further distinction, only implicit in the text but essential in practice, is between separation created by law and separation created by circumstance. De jure segregation is the easier case. Where a statute or regulation assigns children of a particular group to particular schools, Article 3 requires its immediate abrogation. There is no room for progressive realisation and no resource-based excuse. De facto segregation is far harder. Where a school system is formally open to all but produces racially or ethnically homogeneous schools because of residential patterns, catchment rules, testing regimes, informal selection, parental choice or the flight of majority families from mixed schools, the causal chain between state action and educational separation is long and contested. Yet the Convention's purpose or effect formula is designed exactly for this case. If an apparently neutral admissions or assessment rule has the effect of producing separate and unequal schooling for a protected group, it is prohibited, whatever the intentions behind it. The most developed judicial elaboration of this reasoning is found not in UNESCO practice but in the case law of the European Court of Human Rights. In D.H. and Others v. Czech Republic, decided by the Grand Chamber in 2007, the Court considered the placement of Roma children in special schools for pupils with intellectual disabilities. The children had been placed there on the basis of psychological tests. The tests were not, on their face, racial. But Roma children were disproportionately likely to be placed in such schools, and the schools followed a reduced curriculum which foreclosed further education. The Court held that where a general policy has disproportionately prejudicial effects on a particular group, it may be considered discriminatory notwithstanding the absence of discriminatory intent, and that statistical evidence of disparate impact may shift the burden of justification onto the state. That reasoning has been applied repeatedly since. In Horvath and Kiss v. Hungary, the Court addressed the misdiagnosis of Roma children as having intellectual disabilities and their consequent placement in remedial schools. In Elmazova and Others v. North Macedonia, decided in 2022, the Court addressed schools attended predominantly by Roma pupils and the operation of Roma-only classes. In Szolcsan v. Hungary, decided in 2023, the Court addressed the segregation of a Roma pupil in a state-run school. In each case, the Court found a violation of the prohibition of discrimination read together with the right to education. Three lessons follow for the interpretation of the UNESCO Convention. First, statistical disparity is evidence. Where a group is present in a separate or inferior educational setting at rates far exceeding its share of the population, that fact alone requires explanation. Second, an apparently objective assessment instrument can be a vehicle of discrimination. Psychological tests standardised on a majority population, administered in the majority language, to children whose home environment differs systematically from that assumed by the test, are not neutral. Third, consent obtained from parents who have not been informed of the consequences of their choice is not consent. Several of the European cases turned on parental signatures obtained without any explanation that the special school curriculum would preclude progression to secondary education. Each of these lessons maps directly onto the conditions in Article 2 of the UNESCO Convention. Equivalence, voluntariness and non-exclusionary purpose are precisely the questions the European Court has been asking, without ever citing the treaty. 6. The Convention within the Wider Human Rights System 6.1 Relationship with the International Covenant on Economic, Social and Cultural Rights The International Covenant on Economic, Social and Cultural Rights, adopted in 1966, contains in Articles 13 and 14 the most detailed general statement of the #right_to_education in United Nations treaty law. Its provisions closely track Article 4 of the UNESCO Convention, which is unsurprising, since the drafters of the Covenant had the earlier text before them. The Covenant's supervisory body, the Committee on Economic, Social and Cultural Rights, produced in 1999 a General Comment on the right to education which introduced the analytical framework now known as the four A scheme: availability, accessibility, acceptability and adaptability. Education must be available in sufficient quantity. It must be accessible to all without discrimination, physically and economically. It must be acceptable in form and substance, meaning relevant, culturally appropriate and of good quality. And it must be adaptable, meaning capable of responding to the changing needs of diverse learners in diverse settings. The four A scheme is not found in the UNESCO Convention, but it is an extremely useful lens through which to read it. Accessibility corresponds to Article 3. Availability corresponds to Article 4. Acceptability corresponds to Article 5. Adaptability is the dimension the Convention handles least well, and it is precisely the dimension that matters most for learners with disabilities. The practical significance of the relationship is that the Covenant has a treaty body which examines state reports, adopts concluding observations, and, for states which have accepted the Optional Protocol, receives individual communications. The UNESCO Convention has none of these things. In consequence, a great deal of the actual enforcement of educational non-discrimination norms happens through the Covenant machinery rather than the UNESCO machinery, even where the substantive norm originates in the 1960 treaty. Dorsi (2026) argues that this dependence on United Nations treaty bodies, and on the civil society organisations that feed them information, is now itself under threat from funding shortfalls and institutional strain, with serious consequences for the #accountability of states. 6.2 Relationship with the Convention on the Rights of the Child The Convention on the Rights of the Child, adopted in 1989 and now the most widely ratified human rights treaty in existence, addresses education in Articles 28 and 29. Article 28 concerns access and progression. Article 29 concerns the aims of education and reproduces, in updated language, the values commitment of Article 5 of the UNESCO Convention. The Child Rights Convention adds two things of substance. It adds the principle of the best interests of the child, which functions as an interpretive lens across the whole instrument. And it adds a strong participation principle, requiring that the views of the child be given due weight. Neither concept exists in the 1960 text, which is written in the language of state obligation rather than child agency. 6.3 Relationship with the Convention on the Rights of Persons with Disabilities This is the most consequential gap in the 1960 Convention, and it must be stated plainly. Disability is not among the prohibited grounds in Article 1. A state could, without violating the letter of the UNESCO Convention, operate an entirely separate school system for children with disabilities. The Convention on the Rights of Persons with Disabilities, adopted in 2006, closes that gap decisively. Article 24 requires States Parties to ensure an inclusive education system at all levels. The Committee on the Rights of Persons with Disabilities, in its General Comment No. 4 of 2016, drew a set of distinctions that have transformed the field. It distinguished exclusion, which occurs when learners are directly or indirectly prevented from accessing education in any form; segregation, which occurs when the education of learners with disabilities is provided in separate environments designed to respond to particular impairments, in isolation from learners without disabilities; integration, which occurs when learners with disabilities are placed in mainstream institutions without any structural change to the institution itself; and #inclusive_education, which requires a systemic transformation of culture, policy and practice so that the general system accommodates all learners. The Committee's central holding is that the right to non-discrimination includes the right not to be segregated. This is a stronger statement than anything in the 1960 Convention, and it should be read as the modern authoritative gloss on what educational separation means for disabled learners. Its practical implications, and the persistent gap between the norm and practice in national policy, remain a live subject of scholarly debate. 6.4 Relationship with other instruments The International Convention on the Elimination of All Forms of Racial Discrimination, adopted in 1965, prohibits racial segregation in education directly and provides an individual complaints mechanism for states which accept it. The Convention on the Elimination of All Forms of Discrimination against Women, adopted in 1979, addresses in Article 10 the elimination of stereotyped conceptions of the roles of men and women at all levels of education. Regional instruments, including Article 2 of Protocol No. 1 to the European Convention on Human Rights, the African Charter on Human and Peoples' Rights, and the Protocol of San Salvador in the inter-American system, provide further layers of protection, and in some cases, unlike the UNESCO Convention, provide courts. The picture that emerges is one of dense normative overlap and thin enforcement, at least at the level of the UNESCO instrument itself. The norms are stated many times. The remedies are stated once or twice, and usually somewhere else. 7. Monitoring and Implementation 7.1 The periodic consultation system The principal implementation mechanism of the Convention is the periodic consultation of Member States, conducted under Article 7 of the Convention and under the parallel provision of the Recommendation. Because the Recommendation applies to all UNESCO Member States, and not only to States Parties to the Convention, the consultation reaches a substantially wider group of countries than the ratification list would suggest. A state that has not ratified the Convention is still expected to report on its implementation of the Recommendation. This is an elegant piece of institutional design and it materially extends the reach of the normative framework. Consultations have been held at roughly four-year intervals since 1968. The tenth consultation, whose findings were published in 2023, covered the period from 2017 to 2020 and generated, among other outputs, a dedicated overview of state measures concerning the education of minorities (UNESCO, 2023a). The eleventh consultation was launched in February 2024, covered the reporting period from 2021 to 2024, and attracted national reports from a record number of countries, a substantial increase in participation that UNESCO has characterised as a demonstration of renewed political commitment to the rights-based approach (UNESCO, 2025a). The eleventh consultation is analytically interesting for a further reason. It was the first in which states were specifically asked to report on measures concerning learners displaced by conflict, disaster and climate change, reflecting the emergence of #climate_displacement as an educational rights issue. It also captured a period of profound disruption, encompassing the tail of the pandemic and the intensification of armed conflict in several regions. 7.2 What the consultations reveal Read carefully, the consultation reports reveal three patterns. The first is a broad convergence in formal legal frameworks. Most reporting states now have constitutional or statutory guarantees of non-discrimination in education. Most have adopted inclusive education policies at least on paper. Most report free and compulsory primary education. The formal architecture is, in a great many countries, sound. The second is a marked expansion of the concept of the right itself. States increasingly report on early childhood care and education, on adult and non-formal learning, and in some cases on #lifelong_learning entitlements. This reflects the influence of the UNESCO Initiative on the Evolving Right to Education and of the Tashkent Declaration on early childhood, and it suggests that the boundaries of the right recognised in 1960 are being pushed outward in practice even in the absence of formal treaty amendment. The third pattern is less encouraging. The reports consistently identify implementation gaps rather than normative gaps. States describe teacher shortages, weak data systems, inadequate financing, geographic disparities between rural and urban provision, and difficulties in reaching the most marginalised learners. The obstacle, in other words, is rarely that the law permits discrimination. The obstacle is that the state lacks the capacity, or the political will, to give the law effect. 7.3 The limits of self-reporting The consultation system depends on states reporting on themselves. This creates predictable distortions. States report their policies rather than their outcomes. They report the existence of a legal guarantee rather than the number of children who actually benefit from it. They rarely report the exclusions for which they are directly responsible. There is no independent body with the power to examine a state, to question its officials, to receive evidence from affected communities, or to reach a finding of non-compliance. There is no procedure by which a non-governmental organisation may file a shadow report that triggers scrutiny. The contrast with the United Nations treaty body system, imperfect as that system is, is stark. This is why the observation of Dorsi (2026), that the treaty bodies and the civil society organisations that supply them with information carry a disproportionate share of the real monitoring burden for the right to education, is so important. It is also why the weakening of those mechanisms, through underfunding and political pressure, is a threat to the enforcement of the UNESCO Convention itself, even though the Convention has no formal connection to them. 8. Contemporary Fault Lines 8.1 Disability and the persistence of segregation The exclusion of disability from Article 1 is the Convention's most conspicuous drafting failure, judged by contemporary standards. It reflects the medical model of disability that was dominant in 1960, under which disabled children were understood as requiring specialised institutional care rather than as rights-holders entitled to participate in ordinary schooling. The consequence is that the legal work in this area is now done almost entirely by the Convention on the Rights of Persons with Disabilities and by General Comment No. 4. The direction of travel is clear: #segregation of disabled learners is discrimination, and states are obliged to move toward a single inclusive system rather than to maintain a parallel special system. The reality on the ground is very different. Separate special schools remain the dominant model in many countries, including many wealthy ones. Where inclusive placement occurs, it frequently amounts to integration rather than inclusion, meaning that the child is physically present in a mainstream classroom without the support, the curriculum adaptation, the accessible materials or the trained teachers that would make the placement meaningful. The distinction between inclusion and its imitation has become a standing theme in the scholarly literature, and the persistence of segregated provision three decades after the Salamanca Statement is a sobering measure of the gap between norm and practice (UNESCO, 2024b). For the purposes of the UNESCO Convention, the analytically correct position is this. Even without disability as a named ground, a separate special school system that delivers an inferior curriculum, that forecloses progression to further education, and that is entered on the basis of assessments which the family does not understand, is caught by the dignity clause in Article 1 and, where the population of such schools is disproportionately drawn from an ethnic minority, by the racial discrimination provisions as well. That is exactly the reasoning the European Court adopted in the Roma cases. Disability and ethnicity, in the special school system, are frequently the same problem wearing two different labels. This is a clear instance where an #intersectionality lens is not an academic refinement but an evidentiary necessity. 8.2 Girls, women and gendered exclusion Sex is a named ground in Article 1, and single-sex education is permitted under Article 2 only on strict conditions of equivalence. The Convention is therefore well equipped, on paper, to address gendered exclusion. The record is mixed. Enormous progress has been made. Since the mid-1990s, tens of millions of additional girls have entered primary and secondary schooling, and women's participation in tertiary education has expanded dramatically, in many regions overtaking that of men. Gender parity in enrolment has been achieved or approached in a majority of countries. Yet parity in enrolment is not the same as #equality in education, and the Global Education Monitoring Report has repeatedly cautioned against reading a parity index as though it were an equality outcome. Girls remain underrepresented in some fields of study and overrepresented in others. Gendered assumptions persist in curricula and in teacher expectations. Where schooling is formally segregated by sex, women may be structurally excluded from leadership positions in the wider system, since eligibility for headship may be confined to schools of one's own sex. The most extreme contemporary case is Afghanistan, where state policy has barred girls from education beyond the primary level. The scale is difficult to overstate. Reporting by UNESCO and UNICEF indicates that millions of adolescent girls have been excluded from secondary schooling, that female enrolment in formal technical and vocational training has collapsed to a negligible figure, and that the effects have propagated across the whole system, including a substantial decline in male tertiary enrolment (UNESCO and UNICEF, 2025; UNESCO, 2026). Afghanistan also illustrates a methodological point of considerable importance. Administrative data systems may continue to record enrolments that no longer exist in reality, producing gender parity indices that appear to improve in a country that has implemented one of the most comprehensive exclusions of women from formal education in modern history. This is a warning against governing by indicator. A closing statistical gap and a disappearing population can look identical in a spreadsheet. Under the UNESCO Convention, a categorical prohibition on the schooling of girls beyond a given level is not a borderline case requiring careful analysis. It is discrimination in the plainest sense of Article 1: the deprivation of a group of persons of access to education of a type and at a level, on the ground of sex. No Article 2 exception is available, because the separation is neither equivalent nor optional. It is exclusion. 8.3 National, ethnic and linguistic minorities Article 5 grants minorities the right to carry on their own educational activities, subject to conditions. In practice, the schooling of minorities raises two opposite risks, and states routinely fall into one or the other. The first risk is assimilation. Where instruction is provided exclusively in the majority language, minority-language children start school at a systematic disadvantage. The evidence that early instruction in a language the child understands improves learning outcomes is robust and long-standing. Denying #mother_tongue_education is not merely a cultural loss. It is a mechanism that manufactures educational failure, which is then attributed to the child. The second risk is ghettoisation. Separate minority schools, particularly where they are the only realistic option, can become instruments of exclusion, delivering inferior provision to a captive population and severing minority children from the majority society. This is what the European Court found in the #Roma_education cases, and it is precisely the harm that Article 5's conditions were meant to prevent. The solution the Convention gestures toward, and which the evidence supports, is neither pure separation nor pure assimilation. It is bilingual and multilingual provision within a common system: instruction beginning in the child's home language, with structured transition to the language of the wider community, delivered in schools that all children attend. This satisfies every condition of Article 5. It is optional in the relevant sense, it does not prevent minority children from understanding the culture and language of the community as a whole, and it need not involve any lowering of standards. 8.4 Refugees, migrants and displaced learners The Convention obliges States Parties, in Article 3, to give foreign nationals resident within their territory the same access to education as their own nationals. This is a strikingly clear obligation, and it is one of the least honoured. Displaced learners face a familiar catalogue of barriers. Documentation requirements exclude children who cannot produce a birth certificate or residence permit. Language of instruction excludes children who do not speak it. Fees and hidden costs, including uniforms, transport and examination charges, exclude children whose families have no income because their parents are denied the right to work. Recognition of prior learning fails, so a child who completed six years of school in one country is placed with much younger children in another. Bullying and institutional hostility push children out of schools they have formally been admitted to. Every one of these barriers falls squarely within the purpose or effect formula of Article 1. A documentation requirement is neutral on its face and discriminatory in operation. The Convention prohibits it. The eleventh consultation's decision to ask states specifically about displaced learners is therefore welcome and overdue, and it reflects the reality that #refugee_education has become one of the defining equity questions of the period. 8.5 Privatisation and commercialisation Article 2 permits private educational institutions provided their object is not to secure the exclusion of any group, but to provide facilities in addition to those provided by the public authorities. That formulation assumed a particular model: a public system that provides for everyone, supplemented at the margins by private institutions chosen by families who want something different. The contemporary reality in many countries is close to the inverse. Low-fee private schools have proliferated in contexts where the public system has collapsed or was never adequately built. Public-private partnerships channel state funding to private operators. Commercial chains operate at scale. Household expenditure on education, including private tutoring, has become a major determinant of educational outcomes. The equality consequences are direct. A system in which the quality of a child's education depends on the fee the family can pay is a system stratified by #economic_condition, which is a prohibited ground under Article 1. The scale of the private sector need not, in itself, be unlawful. What is unlawful, on a straightforward reading of the Convention, is a private sector whose growth has the effect of nullifying or impairing equality of treatment in education. This is the terrain of the Abidjan Principles, adopted by an expert group in 2019, which restate states' existing obligations to provide free, quality, public education and to regulate private involvement in education so that it does not undermine the rights to equality and non-discrimination (Adamson et al., 2021). Beiter (2023) goes further, arguing that the whole configuration of the right to education in international law is a product of the post-war order and is inadequate to the neoliberal conditions in which education is now delivered, and that a reframing is necessary. Whether one accepts the full force of that critique or not, the observation that the 1960 drafters did not contemplate a global market in schooling is difficult to dispute. Shaheed (2026), writing from the mandate of the United Nations Special Rapporteur on the right to education, makes a related point: the risks that #privatisation_of_education, digitalisation and education in crisis pose to equity require robust governance and a renewed insistence on education as a public and common good. 8.6 Digital technology, data and the new divides The Convention makes no mention of technology, which is unsurprising. Yet the #digital_divide has become one of the most significant vectors of educational inequality, a fact that became impossible to ignore during the pandemic-era school closures, when a very large share of learners worldwide had no household computer and no internet connection at home. Three distinct issues arise, and they should be kept apart. The first is access. Where instruction, homework, assessment or communication with the school is mediated by devices and connectivity that a household does not possess, exclusion follows. Under the purpose or effect test, a policy of digital-first delivery in a context of unequal connectivity is discriminatory in its operation, whatever its intent. The second is the quality and governance of educational technology itself. The Global Education Monitoring Report devoted to technology in education raised sharp questions about the evidence base for large-scale technology deployment, about commercial capture of public education, and about the collection and use of children's data (UNESCO, 2023c). The third is the emergence of algorithmic decision-making in education, including in assessment and in the allocation of school places. An algorithm trained on historical data from a discriminatory system will reproduce that system's outputs while presenting them as objective. This is, structurally, the same problem the European Court identified with psychological testing in the Roma cases: an apparently neutral instrument that encodes and launders an existing pattern of disadvantage. The doctrinal answer is available in the Convention already. Purpose or effect. The practical answer, which requires audit, transparency and remedy, is not. 9. Discussion 9.1 What the Convention does well Three achievements should be recorded without qualification. Conceptual clarity. The definition in Article 1 is a fine piece of legal drafting. It is broad in its grounds, it captures indirect as well as direct discrimination, it names the specific harms of exclusion, inferior provision, separation and indignity, and it has aged remarkably well. Nothing in the definition needs to be rewritten. It needs only to be applied. Normative durability. The Convention has been reaffirmed by successive generations of instruments and remains, in UNESCO's own characterisation, the cornerstone of the #Education_2030 agenda and a central tool for #SDG4. Its main elements recur across the treaty landscape. That is a mark of a document that got the substance right. Institutional reach. The pairing of a binding Convention with a parallel Recommendation, coupled with a reporting obligation that attaches to all Member States rather than only to States Parties, is a clever design that draws non-ratifying states into the normative and monitoring framework. Few treaties manage this. 9.2 What the Convention does poorly Four weaknesses are equally clear. The enforcement deficit. There is no individual complaints procedure, no independent expert body with investigative powers, no adversarial procedure, and no meaningful sanction for non-compliance. The dispute settlement machinery has never been used. The reporting system depends on states describing their own conduct. This is the central problem, and no amount of normative reaffirmation compensates for it. The closed list of grounds. Disability is absent. So are sexual orientation, gender identity, and disability-adjacent grounds such as health status. The absence of disability in particular is a serious substantive gap, only partially remedied by later instruments to which not all states are party. The Article 2 compromise. By admitting that separation can be non-discriminatory if it satisfies certain conditions, the Convention created a category that requires vigilant policing, and then failed to provide the machinery to police it. In practice, the conditions of equivalence and voluntariness attached to permitted separations are asserted rather than tested. Silence on the modern political economy of education. The Convention has nothing to say about commercial schooling at scale, about household expenditure, about technology, about data, or about education in armed conflict and displacement. Some of these silences can be filled by interpretation. Others cannot. 9.3 The implementation gap and its causes The recurring finding of the consultation reports is that the problem is implementation rather than law. It is worth asking why, because the answer is not simply that states are insincere. Financing is a genuine constraint. A significant proportion of countries spend below the internationally recommended floor of public expenditure on education as a share of national income. Where the budget is insufficient, the obligations of Article 4 cannot be met, and states will tend to meet them for the populations that are easiest and cheapest to reach, which by definition are not the marginalised populations the Convention was designed to protect. #education_financing is therefore not a technical footnote to the right to education. It is a precondition of it. Capacity is a second constraint. Delivering inclusive education requires teachers who have been trained to teach diverse learners, and the global #teacher_workforce is under acute strain, with shortages reported almost universally in the eleventh consultation. Data is a third constraint. States cannot remedy exclusions they cannot see. Where administrative data systems record enrolments rather than attendance, or fail to disaggregate by disability, language, ethnicity or displacement status, the marginalised become statistically invisible, and their invisibility is then read as their absence. Political will is the fourth and hardest constraint. Some exclusions are not accidents of capacity. They are policies. A state that bars girls from secondary school is not experiencing an implementation problem. 9.4 Hard law, soft law and the question of reform There is an ongoing debate about how the framework should be strengthened. Broadly, three positions can be identified. The first holds that new binding instruments are needed. Sheppard (2022) argues for an optional protocol extending the free-education guarantee to pre-primary and secondary levels, and the Human Rights Council has now established a working group to explore precisely that (UN Human Rights Council, 2024). This position treats the gaps as gaps in #hard_law that only hard law can fill. The second holds that the more realistic route is authoritative reinterpretation through #soft_law: updated general comments, expert principles such as the Abidjan Principles, and the progressive development of jurisprudence in regional courts. Beiter (2023) inclines toward this view while acknowledging that in some instances new binding law will be unavoidable. The advantage is speed and flexibility. The disadvantage is that soft law does not bind, and that states may simply ignore it. The third holds that neither normative route addresses the actual bottleneck, which is enforcement and financing, and that effort should be redirected accordingly. On this view, the most useful reform to the UNESCO Convention would not be a change to its text at all. It would be the addition of a complaints mechanism and an independent expert committee. These positions are not mutually exclusive, and the most defensible conclusion is that all three are needed. But it is worth being honest about the sequencing. A new instrument that adds fine words to an already fine treaty, without adding a mechanism by which an excluded child can obtain a remedy, would represent activity rather than progress. 10. Limitations of This Study Four limitations should be acknowledged. First, this article is doctrinal and analytical rather than empirical. It does not test its claims against primary data on educational outcomes, and its assessment of the implementation gap therefore relies on secondary sources, principally UNESCO monitoring reports and the Global Education Monitoring Report series. Second, much of the evidence about implementation derives from state self-reporting, which is subject to well-understood biases. States report policies rather than outcomes and are unlikely to volunteer information about their own failures. The picture drawn in Section 7 should be read with that caveat firmly in mind. Third, the jurisprudential analysis in Section 5 relies heavily on the case law of a single regional court. The European Court of Human Rights has produced the most developed body of reasoning on educational segregation, but it operates in a specific legal and political environment, and its reasoning should not be assumed to transfer without adaptation to other regions. Fourth, the article is written in English and draws principally on English-language scholarship and on UNESCO documentation available in English. Significant bodies of work in other languages, and significant national jurisprudence in non-English-speaking jurisdictions, are therefore underrepresented. 11. Recommendations 11.1 For States Parties Ratify, if you have not. The Convention admits no reservations, which means that ratification is a serious step, but it also means that the resulting obligation is clean and unqualified. Around a hundred and ten states have taken this step. That leaves a great many that have not. Audit permitted separations against the conditions in Article 2. Where a state maintains single-sex, religious, linguistic or private educational institutions, it should be able to demonstrate, with evidence, that the equivalence, voluntariness and supplementarity conditions are actually satisfied. Most states have never conducted such an audit. They should. Disaggregate data. A right that cannot be measured cannot be enforced. National data systems should routinely disaggregate enrolment, attendance, completion and learning outcomes by sex, disability, language, ethnicity, migration and displacement status, and household economic condition. Repeal discriminatory law immediately. Article 3 obligations are immediate and cost nothing. There is no defence to a discriminatory statute. Legislate for #justiciability. The single most effective step a state can take is to incorporate the substance of the Convention into domestic law in a form that an individual can invoke before a national court. 11.2 For UNESCO Strengthen the consultation process. The reporting cycle should be supplemented by a mechanism for receiving information from national human rights institutions and civil society organisations, so that the picture presented is not exclusively the state's own account of itself. Consider an optional complaints procedure. An optional protocol establishing an individual communications procedure, available to states that choose to accept it, would transform the Convention from a statement of norms into an enforceable instrument for at least some rights-holders. Publish an updated authoritative commentary. The existing commentary on the Convention is now two decades old. A revised interpretive text addressing disability, privatisation, digitalisation, displacement and climate would be of enormous practical value to national courts and legislatures. 11.3 For national courts and human rights institutions Use the purpose or effect test. National courts adjudicating educational discrimination claims should apply the disparate impact reasoning that the Convention's text plainly authorises, and should follow the European Court in accepting statistical evidence as capable of shifting the burden of justification onto the state. Scrutinise assessment instruments. Testing regimes that channel children into separate or inferior educational tracks should be treated as suspect where their outcomes correlate with protected characteristics. Interrogate consent. Parental agreement to a placement is meaningless unless the parent was informed of its consequences. 11.4 For researchers and students Study outcomes, not enrolments. The most important research contribution in this field is work that goes behind the enrolment figure and asks who is actually learning, in what conditions, and with what consequences. Take #intersectionality seriously as a method. The Roma special school cases demonstrate that disability status and ethnic status can be causally entangled in ways that a single-ground analysis will miss entirely. Study the states that succeed. The literature is dominated by accounts of failure. Comparative work on jurisdictions that have successfully desegregated school systems, or that have built genuinely inclusive provision, would be of far greater practical use. 12. Conclusion The Convention against Discrimination in Education is, in its essentials, a good law. Its definition of discrimination is broad, precise and prescient. Its insistence that discrimination is to be assessed by purpose or effect anticipates by decades the doctrine of indirect discrimination that courts would later develop. Its treatment of separate educational systems as presumptively discriminatory, subject to demanding conditions of equivalence and voluntariness, correctly identifies #segregation as the central mechanism through which educational inequality is manufactured and reproduced. Its insistence that education has purposes, and that those purposes include the development of the human personality and respect for human rights, refuses the reduction of schooling to a mere delivery of skills. And yet. Sixty-five years after its adoption, hundreds of millions of children and young people remain out of school. Girls in at least one country are barred from secondary education by state decree. Roma children in Europe continue to be routed into inferior classrooms by tests that were never designed for them. Children with disabilities are still educated apart from their peers as a matter of routine in a large proportion of the world's school systems. Refugee children are excluded by paperwork. The quality of a child's schooling depends, in a growing number of countries, on what the family can pay. The gap between the norm and the reality is not primarily a gap in the law. It is a gap in enforcement, in financing, in capacity and, in some cases, in political will. The Convention prohibits nearly everything that is wrong. It simply provides almost no way for anyone to do anything about it. It creates rights without remedies, and a right without a remedy is, in the end, a hope with a legal citation attached. This is a solvable problem. The instruments that have made the greatest practical difference to educational rights in the past twenty years, the Convention on the Rights of Persons with Disabilities above all, are distinguished from their predecessors not by the elegance of their norms but by the machinery attached to them. The lesson is not that the 1960 Convention should be rewritten. It is that it should be equipped. For students encountering this material for the first time, the most useful thing to carry away is this. The prohibition of #segregation in education is not a historical artefact belonging to a particular struggle in a particular country in the middle of the last century. It is a live legal norm, binding on a substantial part of the international community, that speaks directly to how school places are allocated, how children are assessed, which language is spoken in the classroom, and who is told, by the architecture of the system itself, that they belong somewhere else. The task of the next generation of lawyers, educators and policymakers is not to invent that norm. It is to make it work. Hashtags #UNESCO #CADE_1960 #right_to_education #non_discrimination #segregation #inclusive_education #equality_of_opportunity #SDG4 #Education_2030 #international_human_rights_law #education_policy #educational_equity #minority_rights #disability_rights #girls_education #right_to_education_1960 #anti_discrimination_law #school_segregation #education_for_all #UNESCO_convention References Adamson, F., Aubry, S., de Koning, M., and Dorsi, D. (Eds.). (2021). Realizing the Abidjan Principles on the right to education: Human rights, public education, and the role of private actors in education. Cheltenham: Edward Elgar Publishing. Beiter, K. D. (2023). Why neoliberal ideology, privatisation, and other challenges make a reframing of the right to education in international law necessary. The International Journal of Human Rights, 27(3), 425 to 470. https://doi.org/10.1080/13642987.2022.2131773 Committee on Economic, Social and Cultural Rights. (1999). General comment No. 13: The right to education (Article 13 of the Covenant). UN Doc. E/C.12/1999/10. Geneva: United Nations. Committee on the Rights of Persons with Disabilities. (2016). General comment No. 4 (2016) on the right to inclusive education. UN Doc. CRPD/C/GC/4. Geneva: United Nations. Committee on the Rights of Persons with Disabilities. (2018). General comment No. 6 (2018) on equality and non-discrimination. UN Doc. CRPD/C/GC/6. Geneva: United Nations. Dorsi, D. (2026). Protecting and strengthening the essential contribution of UN human rights treaty bodies and civil society organizations to the implementation and evolution of the right to education: a necessity at a pivotal time. Frontiers in Education, 11, 1770144. https://doi.org/10.3389/feduc.2026.1770144 Shaheed, F. (2026). The right to education: progress, challenges and the road ahead. Frontiers in Education, 10, 1708336. https://doi.org/10.3389/feduc.2025.1708336 Sheppard, B. (2022). It is time to expand the right to education. Nordic Journal of Human Rights, 40(1). https://doi.org/10.1080/18918131.2022.2071401 UNESCO. (2021). Reimagining our futures together: A new social contract for education. Report of the International Commission on the Futures of Education. Paris: UNESCO. UNESCO. (2022a). Protect her rights, strengthen your laws: Her Atlas status report on girls' and women's right to education. Paris: UNESCO. UNESCO. (2022b). Tashkent Declaration and commitments to action for transforming early childhood care and education. Paris: UNESCO. UNESCO. (2023a). The right to education of minorities: Overview of States' measures reported in the 10th Consultation on the 1960 Convention and Recommendation against Discrimination in Education. Paris: UNESCO. UNESCO. (2023b). Initiative on the evolving right to education 2021 to 2025. Paris: UNESCO. UNESCO. (2023c). Global education monitoring report 2023: Technology in education. A tool on whose terms? Paris: UNESCO. UNESCO. (2024a). Global education monitoring report 2024/5: Leadership in education. Lead for learning. Paris: UNESCO. UNESCO. (2024b). The Salamanca Statement anniversary: 30 years of progress? Paris: UNESCO. UNESCO. (2025a). Monitoring education rights worldwide: Synthesis report of the 11th Consultation on the implementation of the 1960 Convention and Recommendation against Discrimination in Education. Paris: UNESCO. UNESCO. (2025b). The right to education: Past, present and future directions. Paris: UNESCO. UNESCO. (2025c). Global education monitoring report: Gender report 2025. Paris: UNESCO. UNESCO. (2026). Banned from education: A review of the right to education in Afghanistan. Paris: UNESCO. UNESCO and UNICEF. (2024). Global report on early childhood care and education: The right to a strong foundation. Paris and New York: UNESCO and UNICEF. UNESCO and UNICEF. (2025). Afghanistan education situation report 2025. Paris and New York: UNESCO and UNICEF. UN Human Rights Council. (2024). Resolution on an open-ended intergovernmental working group on an optional protocol to the Convention on the Rights of the Child on the rights to early childhood education, free pre-primary education and free secondary education. UN Doc. A/HRC/56/L.8/Rev.1. Geneva: United Nations. Legal Instruments and Cases Cited Instruments Abidjan Principles. (2019). Guiding principles on the human rights obligations of States to provide public education and to regulate private involvement in education. Adopted at Abidjan, Cote d'Ivoire. Convention on the Elimination of All Forms of Discrimination against Women. (1979). New York: United Nations. Convention on the Rights of Persons with Disabilities. (2006). New York: United Nations. Convention on the Rights of the Child. (1989). New York: United Nations. International Convention on the Elimination of All Forms of Racial Discrimination. (1965). New York: United Nations. International Covenant on Economic, Social and Cultural Rights. (1966). New York: United Nations. Protocol Instituting a Conciliation and Good Offices Commission to be Responsible for Seeking the Settlement of any Disputes which may Arise between States Parties to the Convention against Discrimination in Education. (1962). Paris: UNESCO. Recommendation against Discrimination in Education. (1960). Paris: UNESCO. UNESCO Convention against Discrimination in Education. (1960). Adopted 14 December 1960, entered into force 22 May 1962. Paris: UNESCO. Universal Declaration of Human Rights. (1948). Paris: United Nations General Assembly. Cases Brown v. Board of Education of Topeka, 347 U.S. 483 (1954), Supreme Court of the United States. D.H. and Others v. the Czech Republic, European Court of Human Rights, Grand Chamber, judgment of 13 November 2007. Elmazova and Others v. North Macedonia, European Court of Human Rights, judgment of 13 December 2022. G.L. v. Italy, European Court of Human Rights, judgment of 10 September 2020. Horvath and Kiss v. Hungary, European Court of Human Rights, judgment of 29 January 2013. S. v. the Czech Republic, European Court of Human Rights, judgment of 2024. Szolcsan v. Hungary, European Court of Human Rights, judgment of 30 March 2023.

This article examines the Convention on the Nationality of Married Women, adopted by the United Nations General Assembly on 29 January 1957 and in force since 11 August 1958. The treaty was the first universal instrument to state plainly that marriage to a foreigner, the ending of such a marriage, or a change in a husband's nationality during marriage should not, by itself, alter a wife's #nationality. The article traces the long campaign that produced the Convention, explains its legal architecture article by article, and assesses what it achieved and where it fell short. It then places the treaty inside the wider system of #international_law, including the Universal Declaration of Human Rights, the two statelessness conventions of 1954 and 1961, the International Covenant on Civil and Political Rights, the Convention on the Rights of the Child, and above all Article 9 of the Convention on the Elimination of All Forms of Discrimination against Women. Drawing on doctrinal analysis and on comparative material from Malaysia, Nepal, the Gulf states, and sub Saharan Africa, the article argues that the 1957 Convention remains conceptually important but institutionally weak. It secured the negative right of a wife not to lose her nationality automatically, yet it did not secure the positive right of a mother to transmit nationality to her children, and it left the machinery of enforcement almost entirely to states. The article concludes that the treaty is best understood not as a finished settlement but as a first draft of a norm that later instruments completed and that domestic courts and civil society movements are still working to make real. Recommendations are offered for legislators, treaty bodies, and researchers. Keywords: nationality of married women; gender discriminatory nationality laws; statelessness; CEDAW Article 9; dependent nationality; equal citizenship; international human rights law 1. Introduction For most of the modern history of the nation state, a woman's legal membership in a political community was treated as a function of the man she married. She did not hold her #citizenship in her own name so much as borrow it from a husband. If she married a foreigner, many legal systems stripped her of her original status without asking her opinion. If her husband later changed his nationality, hers changed with his. If the marriage ended through death or divorce, she could find herself holding a nationality that no longer had any social meaning for her, or holding none at all. This arrangement had a name in legal scholarship: #dependent_nationality. It was not an accident or an oversight. It was a deliberate design, rooted in the doctrine of marital unity that common law systems called #coverture and that civil law systems expressed through the principle of family unity of nationality. The idea was that a married couple formed a single legal person and that the husband was the head of that person. One family, one nationality, one head. The rule was tidy for administrators. It was disastrous for women. The Convention on the Nationality of Married Women was the first universal treaty to attack that rule directly. Opened for signature by General Assembly resolution 1040 (XI) of 29 January 1957 and entering into force on 11 August 1958, it declared that neither the celebration nor the dissolution of a marriage between one of its nationals and an alien, nor the change of nationality by the husband during marriage, shall automatically affect the nationality of the wife. It is a short instrument. It has eleven articles, only three of which contain substantive obligations. It has never attracted mass ratification. As of the mid 2020s it counts roughly seventy five states parties, a modest figure by the standards of later #human_rights treaties. Yet the Convention matters, and it matters in ways that are easy to miss if one measures treaties only by counting signatures. First, it broke a conceptual monopoly. Before 1957, the dominant international response to the problem of women's nationality had been to reduce the harm caused by dependent nationality rather than to abolish the principle itself. The 1930 Hague Convention on Certain Questions Relating to the Conflict of Nationality Laws, produced by the #League_of_Nations, tried to prevent married women from becoming stateless while leaving intact the assumption that a wife's status should follow her husband's. The 1957 Convention rejected that compromise. It said that the wife's nationality is her own. Second, it supplied the drafting template that later, more powerful instruments would inherit. Article 9(1) of the Convention on the Elimination of All Forms of Discrimination against Women, adopted in 1979, is in substance an expanded restatement of the 1957 text. Anyone reading #CEDAW Article 9 without knowing the earlier treaty will miss the fact that they are reading a second edition. Third, it exposed the limits of a purely negative conception of equality. The 1957 Convention protects a woman from losing her nationality. It says nothing meaningful about her right to pass her nationality to her children, and it deals with the naturalisation of foreign wives in a way that, read today, is itself asymmetrical. Those omissions are not merely technical. They are the reason that #childhood_statelessness caused by #gender_discriminatory_nationality_laws persists in the 2020s in roughly two dozen countries, even though the automatic loss of a wife's nationality on marriage has become rare. 1.1 Research problem The scholarly literature on nationality and #statelessness has expanded rapidly over the past decade. Yet within that literature the 1957 Convention occupies an odd position. It is routinely cited in footnotes as a historical milestone and then set aside. Analysis moves quickly to CEDAW, to the 1961 Convention on the Reduction of Statelessness, and to contemporary campaigns. The result is that the treaty is remembered but not examined. Students in particular encounter it as a date and a title rather than as a legal text with a structure, a drafting history, a set of compromises, and a measurable record of implementation. This article treats the 1957 Convention as a subject worth studying in its own right, and asks what it can still teach. 1.2 Research questions The article addresses five questions. What historical conditions produced the Convention, and what alternatives were rejected in the process? What exactly does the treaty oblige states to do, and what does it deliberately leave alone? How does it relate to the instruments that came before and after it, and does it retain any independent legal function today? What does state practice, including #ratification patterns, reservations, litigation, and legislative reform, reveal about the treaty's effectiveness? What lessons does the Convention offer for the present campaign to eliminate remaining gender discrimination in #nationality laws? 1.3 Contribution and structure The contribution of the article is threefold. It offers a consolidated doctrinal reading of a treaty that is rarely read closely. It links that reading to recent empirical and comparative scholarship on statelessness and gendered citizenship. And it draws out practical implications for law reform. Section 2 sets out the methodology. Section 3 reconstructs the historical background. Section 4 examines the drafting and adoption of the Convention. Section 5 analyses the substantive articles. Section 6 identifies the treaty's silences and structural weaknesses. Section 7 situates it within the broader normative framework. Section 8 examines regional and national practice. Section 9 assesses the human consequences of the gaps that remain. Section 10 explains why the Convention was overtaken by CEDAW and asks whether it still does any work. Section 11 analyses the obstacles to reform. Section 12 offers recommendations. Section 13 notes limitations and directions for further research. Section 14 concludes. 2. Methodology This is a doctrinal and comparative legal study, supported by secondary empirical literature. 2.1 Approach The core method is textual and contextual analysis of primary legal instruments. The 1957 Convention is read alongside its immediate predecessors, its successors, and the interpretive output of #treaty_body institutions, in particular General Recommendation No. 32 of the CEDAW Committee. The analysis follows the interpretive rules that are now codified in the Vienna Convention on the Law of Treaties, giving terms their ordinary meaning in context and in light of the object and purpose of the instrument, and using preparatory materials where the ordinary meaning is unclear. The doctrinal analysis is supplemented by a comparative survey of national practice. Jurisdictions were selected on three criteria: the presence of documented gender discrimination in nationality law, the availability of recent scholarly or institutional analysis, and regional diversity. This yields a set that includes Malaysia, Nepal, Iran, Jordan, Lebanon, several Gulf Cooperation Council states, and a group of African states including Eswatini, Madagascar, Mauritania, and Sierra Leone. Finally, the article draws on the secondary literature of the past five years on statelessness, gendered citizenship, and treaty compliance, in order to connect legal doctrine to observed outcomes. 2.2 Scope and limits Three limits should be stated at the outset. First, this is not a quantitative study. Where numerical claims appear, for example about the number of states that still restrict women's ability to confer nationality on their children, they are drawn from published institutional surveys and are presented as such. Second, the article concentrates on nationality law as it affects women in marriage. It touches only briefly on racial and ethnic discrimination in nationality law, on #denationalization as a counterterrorism instrument, and on statelessness arising from state succession, even though all three intersect with the subject. Third, the article is written for a student readership. Technical vocabulary is explained rather than assumed. Some nuances of #conflict_of_laws are simplified. Readers pursuing the topic further are directed to the sources listed at the end. 2.3 Terminology Three terms require clarification. Nationality and citizenship. In international law, nationality describes the legal bond between an individual and a state, a bond that gives the state standing to protect that individual internationally. Citizenship, in most domestic systems, describes the bundle of rights and duties that flow from that bond. Many legal systems use the words interchangeably. This article follows the practice of using nationality when speaking of the international bond and citizenship when speaking of domestic membership, but does not treat the distinction as rigid. Statelessness. A stateless person is one who is not considered as a national by any state under the operation of its law. This is the definition in Article 1 of the 1954 Convention Relating to the Status of Stateless Persons, and it is now widely accepted as customary. Gender discriminatory nationality laws. This phrase covers any legal provision that treats men and women differently in the acquisition, change, retention, or transmission of nationality. It includes both direct discrimination, where the law says so on its face, and indirect discrimination, where a formally neutral rule produces unequal outcomes because of underlying social conditions such as low rates of #birth_registration among unmarried mothers. 3. Historical Background: How Wives Lost Their Nationality 3.1 The doctrine of marital unity The rule that a wife takes her husband's nationality was not ancient. It hardened in the nineteenth century, in step with the consolidation of the modern state and the modern passport. Before that, membership was often a matter of allegiance to a sovereign, residence in a territory, or membership of a corporate body, and it was not always neatly gendered. What changed was the arrival of comprehensive nationality statutes. Once states began codifying who belonged and who did not, they had to decide what to do about mixed marriages. Almost everywhere, they decided in favour of unity under the husband. The British Naturalization Act of 1870 provided that a British woman who married an alien would be deemed an alien. Comparable rules were adopted across Europe, in the United States through the Expatriation Act of 1907, and in colonial territories where metropolitan codes were transplanted with little modification. The justification offered was practical rather than principled. A family, it was said, should not be divided in its loyalties. Diplomatic protection would be simpler if one state spoke for the whole household. Children would have a clear status. Behind these arguments sat the older assumption of #coverture: that on marriage the legal personality of the wife merged into that of the husband, who represented the family in public life. The consequences were severe and were not hypothetical. A woman who married a foreigner might lose her nationality without acquiring her husband's, because his state applied different rules. She became stateless immediately, through no act of her own beyond marriage. A woman who acquired her husband's nationality and was then widowed or divorced might find that the acquired nationality lapsed while the original one did not revive. She lost the vote, lost the right to hold certain jobs, lost the right to own land in some jurisdictions, and could be deported from the country of her birth. During the two world wars, women who had married enemy aliens were interned or dispossessed on the basis of a nationality they had never chosen and often did not want. 3.2 The interwar campaigns The first sustained international challenge to dependent nationality came from organised feminist movements between the wars. This is a history that has been recovered in recent scholarship and that deserves emphasis, because it corrects a common assumption that campaigning against gender discriminatory nationality laws began only in the twenty first century. The International Alliance of Women, then known as the International Alliance of Women for Suffrage and Equal Citizenship, made nationality a central plank of its programme. National women's organisations petitioned their governments. Lawyers such as Chrystal Macmillan in Scotland, Maria Verone in France, and Betsy Bakker Nort in the Netherlands built a coordinated legal case. In the Americas, the Inter American Commission of Women, under the direction of Doris Stevens, carried out a multi year comparative study of nationality legislation and presented its findings to states. The immediate target was the Conference for the Codification of International Law convened by the League of Nations at The Hague in 1930. Campaigners understood that if a global code of nationality law was to be written, the moment to insist on equality was before the ink dried. They were largely excluded from official delegations. They attended anyway, as unofficial observers, and staged public demonstrations that irritated the conference presidency to the point of police intervention. The outcome was a compromise that campaigners rejected. The 1930 Hague Convention on Certain Questions Relating to the Conflict of Nationality Laws addressed the position of married women only to the extent of reducing statelessness and dual nationality. It provided, in effect, that a woman should not lose her nationality on marriage unless she acquired her husband's. That is a statelessness safeguard, not an equality guarantee. It preserved the underlying idea that a married woman's status was properly conditional on her husband's. The campaigners saw this clearly and said so. Their refusal to accept a statelessness safeguard as a substitute for equality is one of the most instructive episodes in the history of the field, and it prefigures a debate that continues today. 3.3 The Montevideo breakthrough The regional level moved faster than the global one. At the Seventh International Conference of American States, held at Montevideo in 1933, the Inter American system adopted the Convention on the Nationality of Women. Its operative provision was a single sentence to the effect that there shall be no distinction based on sex as regards nationality in legislation or in practice. Adopted alongside a companion instrument on equality of rights between men and women, the #Montevideo_1933 Convention is generally regarded as the first international treaty devoted to women's rights, and the first to make sexual equality a principle to be written into domestic legislation. Its practical reach was limited to the Americas and its ratification was uneven, but its normative significance was large. It established that the equality position was not utopian. It could be drafted. It could be adopted. States could live with it. 3.4 From the League to the United Nations The Second World War produced a vast stateless population and made the connection between nationality and human dignity impossible to ignore. When the United Nations set about constructing a human rights framework, the right to a nationality was included in the Universal Declaration of Human Rights of 1948. Article 15 states that everyone has the right to a nationality and that no one shall be arbitrarily deprived of their nationality nor denied the right to change it. The 1957 Convention's preamble expressly recalls this provision, which tells us something about how its drafters understood their task. They were not inventing a new right. They were applying an existing one to a group of people from whom it had been withheld. Two institutional developments then made a dedicated treaty possible. The Commission on the Status of Women, established in 1946, took up nationality as a priority and began a systematic study of national legislation from 1948. And the International Law Commission, working on statelessness, produced draft materials that kept the technical questions of conflict of laws in view. The #Commission_on_the_Status_of_Women supplied the political will. The technical bodies supplied the drafting competence. The result was the treaty adopted in 1957. 4. Drafting and Adoption 4.1 The negotiating problem The drafters of the 1957 Convention faced a problem that recurs throughout #treaty_law. The states most in need of reform were the states least willing to accept a strong obligation. A treaty drafted at the level of the most progressive states would be ratified by the most progressive states and by nobody else. A treaty drafted at the level of the most reluctant would be worthless. The solution adopted was characteristic of the period. The Convention states a firm principle in Article 1, then softens the surrounding provisions with procedural language, and then leaves enforcement almost entirely to domestic law. It secures the core and concedes the periphery. 4.2 Adoption The Convention was adopted by the #General_Assembly at its 647th plenary meeting on 29 January 1957, by resolution 1040 (XI), and annexed to that resolution. It was opened for signature and ratification at the close of the eleventh session. It entered into force on 11 August 1958, in accordance with its Article 6, once the required number of instruments of ratification or accession had been deposited. The preamble is worth reading carefully, because it frames everything that follows. It recognises that conflicts in law and in practice with reference to nationality arise as a result of provisions concerning the loss or acquisition of nationality by women as a result of marriage, of its dissolution, or of the change of nationality by the husband during marriage. It recalls Article 15 of the Universal Declaration. And it expresses the desire to cooperate with the #United_Nations in promoting universal respect for human rights without distinction as to sex. Three things follow from that framing. The Convention is presented as a solution to a #conflict_of_laws problem as much as to a discrimination problem. The mischief identified is the clash between legal systems that produces stateless or involuntarily denationalised women. The Convention is anchored in an existing right rather than a new one. This gives it interpretive weight beyond the number of its parties, because it can be read as a specification of what Article 15 of the #UDHR requires in a particular situation. The Convention explicitly invokes non discrimination on grounds of sex. It is, in that sense, a women's rights treaty and not merely a technical instrument of private international law. 5. The Legal Architecture of the Convention The Convention contains eleven articles. Articles 1 to 3 carry the substance. Articles 4 to 11 deal with signature, ratification, entry into force, territorial application, dispute settlement, denunciation, and related matters. This section analyses the substantive core. 5.1 Article 1: the automatic effects rule Article 1 is the heart of the instrument. Each contracting state agrees that neither the celebration nor the dissolution of a marriage between one of its nationals and an alien, nor the change of nationality by the husband during marriage, shall automatically affect the nationality of the wife. Four features of this provision repay attention. It is an obligation of result, not of effort. The state agrees that these events shall not have the stated consequence. There is no qualifying language about progressive realisation or available resources. Once a state is bound, its law must not produce the prohibited automatic effect. It covers three distinct triggers. Marriage. Dissolution of marriage. Change of the husband's nationality during marriage. Each of these had, historically, been a moment at which a woman's status was altered without her consent. The Article closes all three doors. The operative word is automatically. This is the hinge on which the whole treaty turns, and it is both the source of the Convention's strength and the source of its weakness. What Article 1 prohibits is the automatic, operation of law consequence. It does not prohibit a woman from voluntarily changing her nationality on marriage. It does not prohibit a state from offering her that option. What it prohibits is the state deciding for her. The Convention therefore protects choice rather than imposing an outcome. This is, on reflection, the correct approach. A rule that forbade a woman from taking her husband's nationality would simply substitute one form of paternalism for another. But the word also creates room for evasion, discussed below. It applies to the wife, not to spouses generally. The Convention is drafted in the vocabulary of its time. It addresses the harm as it was then experienced, which was a harm to women. A modern drafter would use gender neutral language and would extend the same protection to a husband whose wife changes her nationality. The asymmetry of the text is a historical artefact, but it is an artefact with consequences, because it makes the Convention harder to read as a general principle of #equal_protection and easier to read as a special protective measure. 5.2 Article 2: statelessness through the husband's acts Article 2 provides that neither the voluntary acquisition of the nationality of another state nor the renunciation of its nationality by one of its nationals shall prevent the retention of its nationality by the wife of such national. This provision addresses a specific and cruel scenario. A husband emigrates, naturalises elsewhere, or renounces his nationality. Under the family unity principle, his wife's nationality would evaporate along with his, whether or not she had ever left the country, whether or not she consented, and whether or not she acquired anything in exchange. Article 2 forbids this. Her nationality survives his decisions. Read together, Articles 1 and 2 constitute the Convention's central achievement. They sever the legal link between a woman's nationality and her husband's biography. After 1957, in states bound by the treaty, a wife's nationality is a fact about her, not a reflection of him. 5.3 Article 3: the naturalisation of alien wives Article 3 is the most complicated and the most criticised provision. It has two paragraphs. Paragraph 1 provides that each contracting state agrees that the alien wife of one of its nationals may, at her request, acquire the nationality of her husband through specially privileged #naturalization procedures, and that the grant of such nationality may be subject to such limitations as may be imposed in the interests of national security or public policy. Paragraph 2 provides that the Convention shall not be construed as affecting any legislation or judicial practice by which the alien wife of one of its nationals may, at her request, acquire her husband's nationality as a matter of right. Two observations follow. First, the provision is asymmetrical. It creates a facilitated route to nationality for the foreign wife of a male national. It says nothing at all about the foreign husband of a female national. A state could comply fully with Article 3 while offering fast track naturalisation to foreign wives and ordinary, slow, discretionary naturalisation to foreign husbands. This is precisely the pattern that persists today in a number of jurisdictions, and it is a live source of hardship for #mixed_marriage families, since it determines whether a father can lawfully reside with his children. Second, the safeguard clause is broad. The privileged procedure may be limited in the interests of national security or #public_policy. Those are elastic categories. Article 3 thus offers a benefit with one hand and a wide discretion to withhold it with the other. Article 3 is best understood as a product of its era. In 1957, the assumption that migration in marriage flowed overwhelmingly from wife to husband's country was empirically defensible and normatively unexamined. The drafters were trying to ensure that a woman who moved to her husband's country would not be left in permanent legal limbo. They did not think to ask the mirror question. That failure of imagination is the Convention's clearest limitation. 5.4 The procedural articles Articles 4 to 11 are conventional final clauses. The Convention is open to signature and ratification by member states and by other states invited by the General Assembly. It enters into force on the ninetieth day following the deposit of the sixth instrument of ratification or #accession. It may be extended to non metropolitan territories, a provision that reflects the colonial architecture still in place in 1957. It contains a dispute settlement clause referring unresolved disputes over interpretation or application to the International Court of Justice at the request of any party to the dispute. It may be denounced by written notification. Two features of these clauses matter for the assessment that follows. There is no monitoring body. The Convention creates no committee, requires no periodic reports, and establishes no complaints mechanism. Compliance depends on the good faith of states and, in theory, on the willingness of another state party to take a dispute to the International Court of Justice. No such case has ever been brought. In practice, therefore, the Convention has no international #monitoring machinery at all. The reservations position is governed by general treaty law rather than by an express clause tailored to the instrument. This has allowed states to accede while carving out precisely the obligations that would have required them to change their law. 6. What the Convention Does Not Do A treaty is defined as much by its silences as by its commands. Four silences shape the legacy of the 1957 Convention. 6.1 Silence on the transmission of nationality to children This is the decisive gap. The Convention protects a woman's own nationality. It says nothing about her capacity to pass that nationality to her children. The omission is not a small one. It is, in fact, the source of nearly all of the statelessness that gender discriminatory nationality laws produce today. Consider the arithmetic. A country may fully comply with Articles 1 and 2. Its women keep their nationality when they marry foreigners. But if that country's law permits only fathers to transmit nationality by descent, then the child of a national mother and a foreign father acquires nationality from neither parent whenever the father is stateless, unknown, absent, unable to prove his status, or a national of a state that itself confers nationality only through the mother, or only on children born on its territory. The child is stateless from birth. The mother, who is fully protected in her own status, watches her child grow up without one. Under a strict rule of #jus_sanguinis through the paternal line, the mother's nationality is legally invisible to her own child. This is why later instruments had to go further. Article 9(2) of CEDAW addresses exactly this point, granting women equal rights with men with respect to the nationality of their children. It is a provision the 1957 Convention lacks, and its absence explains why the earlier treaty could never have solved the problem on its own. 6.2 Silence on the foreign husband As noted, Article 3 provides for facilitated naturalisation of alien wives without providing anything for alien husbands. Where domestic law follows the same logic, the consequences are practical and severe. A foreign husband may be unable to obtain residence, may be barred from formal employment, may be excluded from public health and education systems, and may face repeated visa renewals for decades. Families are separated. Women are forced to choose between their country and their marriage. Recent qualitative research on cross national families in the Gulf, conducted during the diplomatic rupture of 2017 to 2021, documented this pattern in detail, showing how women married to non nationals were denied the practical benefits of full citizenship and how the resulting insecurity damaged both socioeconomic position and psychological wellbeing (Abu Ras et al., 2022). 6.3 Silence on remedies The Convention prohibits an outcome. It does not say what happens to the woman who has already suffered that outcome. It contains no provision on restoration of nationality wrongly lost, no rule on retroactivity, and no requirement that states provide an effective remedy. This matters enormously in practice. When a state finally reforms a discriminatory nationality law, the question that immediately arises is whether the reform reaches backwards. The Malaysian constitutional amendment of 2024, discussed below, granted Malaysian mothers the right to confer automatic citizenship on children born overseas, but did so prospectively only, leaving the children born before the change to a separate and more onerous registration route. United Nations experts observed that non retroactive application fails to remedy past violations and allows discriminatory treatment to continue for those already affected. The 1957 Convention gives no textual foothold for arguing otherwise. 6.4 Silence on indirect discrimination The Convention addresses laws that discriminate on their face. It does not address rules that are neutral in wording but discriminatory in effect. Yet much contemporary exclusion operates in exactly that way. Requirements that a child's birth be registered with documentation the mother cannot obtain alone, evidentiary rules that presume paternity within marriage and demand proof outside it, and administrative discretion exercised by officials with fixed views about respectable motherhood all produce unequal outcomes without ever mentioning sex. Modern equality law has developed the concept of #indirect_discrimination to reach such rules. The 1957 text has no such concept. 7. The Convention Within the International Framework The 1957 Convention did not remain alone. It is one node in a network of instruments, and its meaning today depends on that network. 7.1 The Universal Declaration and the right to a nationality Article 15 of the UDHR is the normative parent of the Convention. It establishes the #right_to_a_nationality and prohibits arbitrary deprivation. Although the Declaration is formally a resolution rather than a treaty, its central provisions are widely regarded as reflecting customary international law, and the right to a nationality has been treated as such by the African Court on Human and Peoples' Rights in its 2018 judgment in Anudo v United Republic of Tanzania. The relationship runs in both directions. The UDHR gives the 1957 Convention its normative authority. The 1957 Convention gives the UDHR concrete content, by specifying that a deprivation triggered automatically by marriage is an #arbitrary_deprivation. 7.2 The statelessness conventions of 1954 and 1961 The Convention Relating to the Status of Stateless Persons of 1954 defines statelessness and establishes a protection regime for those already stateless. The Convention on the Reduction of Statelessness of 1961 imposes obligations designed to prevent statelessness from arising, including a duty to grant nationality to persons born on the territory who would otherwise be stateless, and rules restricting loss and deprivation of nationality. The 1961 Convention is the closest structural relative of the 1957 instrument, and the comparison is instructive. The 1961 Convention approaches gender in the manner of the #Hague_Convention_1930: it seeks to prevent statelessness, including statelessness arising from marriage, without requiring equality as such. A state can prevent statelessness while maintaining a thoroughly patriarchal transmission rule, simply by adding a safety net for the residual cases. Reform advocates have long argued, correctly, that a safety net is not equality, and that safety nets are precisely the provisions most likely to fail in practice because they depend on discretionary administration. 7.3 The International Covenant on Civil and Political Rights The Covenant, adopted in 1966, contains a general guarantee of equality before the law and equal protection without discrimination, and a specific provision in Article 24(3) that every child has the right to acquire a nationality. Read together, these support the argument that a nationality law which distinguishes between mothers and fathers violates the Covenant even where the state has not ratified CEDAW or has entered a reservation to Article 9. 7.4 The Convention on the Rights of the Child Article 7 of the CRC provides that a child shall be registered immediately after birth and shall have the right from birth to a name and the right to acquire a nationality. Article 8 protects the child's right to preserve their identity, including nationality. With near universal ratification, the CRC is in practice the most widely binding statement of the child's right to nationality, and it has become a central tool for advocates challenging discriminatory transmission rules, since it allows the argument to be framed as a children's rights issue rather than a women's rights issue, which is sometimes politically easier. 7.5 CEDAW Article 9 and General Recommendation 32 Article 9 of CEDAW is the instrument that completed what the 1957 Convention began. Article 9(1) restates the 1957 rule and strengthens it. States parties shall grant women equal rights with men to acquire, change, or retain their nationality. They shall ensure in particular that neither marriage to an alien nor change of nationality by the husband during marriage shall automatically change the nationality of the wife, render her stateless, or force upon her the nationality of the husband. Note the additions. The 1957 Convention addressed only automatic change. Article 9(1) also prohibits rendering the wife stateless and forcing the husband's nationality upon her. And it opens with a general equality guarantee covering acquisition, change, and retention, rather than a narrow prohibition on specified automatic effects. Article 9(2) supplies the missing piece. States parties shall grant women equal rights with men with respect to the nationality of their children. General Recommendation No. 32, adopted by the CEDAW Committee in 2014, elaborates the gender related dimensions of refugee status, asylum, nationality, and statelessness of women. It treats sex and gender based discrimination in nationality laws as having a significant and detrimental impact on the enjoyment by women and their children of their human rights, and it frames such discrimination as something to be eliminated rather than accommodated. The #general_recommendation_32 framework is now the primary interpretive reference for advocates and treaty bodies. The practical consequence is that CEDAW has largely absorbed the function of the 1957 Convention. With one hundred and eighty nine states parties, CEDAW binds almost the entire international community, and its committee produces a continuous stream of concluding observations naming specific states and specific provisions. The 1957 Convention, with roughly seventy five parties and no #treaty_body, cannot compete. 7.6 The problem of reservations CEDAW's near universality comes at a price. It is among the most heavily reserved of the core human rights treaties, and Article 9 is one of the most frequently reserved provisions. Recent quantitative work confirms what advocates have long suspected. States that ratify CEDAW with reservations show smaller improvements in women's rights than states that ratify without them, and states with weaker records tend to place their reservations on the articles most central to the treaty's purpose rather than on procedural provisions (Kreutzer and Mitchell, 2024). Complementary work on the withdrawal of reservations suggests that sustained international social pressure can move states to remove them, which offers some ground for optimism (Boyes et al., 2024). A dedicated dataset on CEDAW reservations and objections now makes it possible to study these patterns systematically (Kreutzer, 2024). Institutional surveys of nationality legislation show the shape of the problem clearly. Of the states whose nationality laws still contain gender discriminatory provisions in the conferral of nationality on children, the large majority are parties to CEDAW, and a substantial minority of those have entered #reservations to Article 9(2) or to Article 9 in its entirety. A further group of states with discriminatory laws are parties to CEDAW without any relevant reservation at all, which means that they are, on the face of it, in continuing breach. This is the central paradox of the modern framework. The obligation is nearly universal. The #compliance is not. 8. Implementation and State Practice 8.1 Ratification patterns The 1957 Convention has attracted around seventy five states parties. This is a low figure, and the reasons are instructive. Some states did not ratify because they saw no need to. Their domestic law already complied, and they judged that a treaty adds administrative cost without adding protection. This is the standard argument of the confident common law state, and it is not entirely unreasonable, though it ignores the value of a treaty as a signal to others. Some states did not ratify because the Convention would have required legal change they were unwilling to make. This is the honest refusal. Some states did not ratify because by the time they were ready to act, CEDAW existed and offered a more comprehensive and more prestigious vehicle. This is the most common explanation for the treaty's stagnation after 1979. A state that joins CEDAW acquires the 1957 obligations and more. Joining the older Convention as well begins to look redundant. The result is that the 1957 Convention is now, in practical terms, a closed instrument. It is not growing. It is not being enforced through its own machinery. Its provisions live on through their descendants. 8.2 The global picture today The situation regarding a wife's own nationality has improved dramatically since 1957. Automatic loss of nationality on marriage to a foreigner, once the global norm, is now rare. On that specific question the Convention and its successors can be said to have largely won. The situation regarding the transmission of nationality is very different. Institutional surveys conducted jointly by the office of the United Nations High Commissioner for Refugees and UN Women continue to identify roughly two dozen states in which women do not enjoy equal rights with men to confer nationality on their children, and a larger group, historically counted at around fifty, in which nationality laws contain some form of gender discriminatory provision affecting women's ability to acquire, change, retain, or transmit nationality, or to confer it on a spouse. These states are spread across every region, though they cluster in the Middle East and North Africa and in parts of sub Saharan Africa and South and Southeast Asia. The direction of travel is positive. Over the past two decades a series of states have reformed, and the pace has, if anything, quickened. But the remaining cases are the hard ones, and the reforms that are achieved are often partial. 8.3 Case study: Malaysia Malaysia offers the clearest recent illustration of both progress and its limits, and it deserves extended treatment because it shows how #constitutional_litigation, legislative reform, and civil society pressure interact. Under the Federal Constitution as it stood, a child born outside Malaysia acquired citizenship by operation of law if the father was a Malaysian citizen at the time of birth. A child born outside Malaysia to a Malaysian mother and a foreign father did not. Such a child could apply for citizenship by registration under a separate constitutional provision, but registration was discretionary, slow, and frequently refused without reasons. In December 2020, the organisation Family Frontiers and six affected Malaysian mothers filed a constitutional challenge. In September 2021 the High Court held that Malaysian mothers' overseas born children were entitled to citizenship automatically. In August 2022 the Court of Appeal overturned that decision by a majority. In December 2022 the Federal Court permitted the appellants to pursue a further appeal. While the litigation was pending, the government moved legislatively. A Constitution Amendment Bill was tabled in March 2024, passed by the lower house in October 2024 and by the upper house in December 2024. It provides that a child born abroad to either a Malaysian father or a Malaysian mother acquires citizenship by operation of law. On the central question of maternal transmission, #Malaysia has now equalised its law. Three qualifications matter. The amendment is not retroactive. Children born abroad to Malaysian mothers before the amendment takes effect do not benefit automatically. Following a settlement recorded by consent in the Federal Court in March 2025, such children under the age of eighteen may apply for registration under the existing constitutional route on defined terms, which improves their position but does not equate it with that of children born after the change. Implementation of the automatic rule was reported to be scheduled for the middle of 2026. The amendment was bundled with provisions widely regarded as regressive. These included the removal of automatic citizenship for children born in Malaysia to stateless permanent residents, a reduction in the age limit for certain applications, and a provision permitting revocation of citizenship from foreign wives whose marriages are dissolved within two years of the grant. That last provision is a direct echo of the very logic the 1957 Convention was written to abolish: a woman's status made contingent on the survival of her marriage. The reform did not extend equality to the conferral of nationality on a foreign spouse. A Malaysian man may sponsor a foreign wife more easily than a Malaysian woman may sponsor a foreign husband. Malaysia therefore illustrates a pattern seen elsewhere. Reform arrives, it is real, and it is incomplete. The equality achieved on one axis is partly offset by restrictions introduced on another. 8.4 Case study: Nepal Nepal presents a different configuration. Its constitutional and statutory provisions on citizenship by descent have long made it difficult for a mother to transmit citizenship without establishing the father's status or his absence, with the practical result that the children of Nepali women married to foreign men, and the children of women who cannot or will not identify a father, face acute difficulty. Grassroots campaigning, including the movement known as Citizenship in the Name of the Mother, has kept the issue politically alive across successive governments and made it an election issue. Nepal is a party to CEDAW without a reservation to Article 9(2), which places it squarely within the category of states whose domestic law is difficult to reconcile with an obligation they have freely accepted. The Nepali case shows the value and the limits of #civil_society mobilisation. Sustained #advocacy has produced repeated legislative attempts, considerable public sympathy, and a shift in the terms of national debate. It has not yet produced full equality. 8.5 Case study: the Gulf states Several Gulf Cooperation Council states maintain nationality laws under which a female national married to a non national cannot confer nationality on her children, or can do so only in narrowly defined circumstances and at the discretion of the executive, and cannot confer nationality on her husband at all. The scholarship on this region is now substantial. Qualitative research in Qatar during and after the diplomatic blockade of 2017 documented how the interaction of gendered nationality rules with a sudden political rupture left children of cross national marriages stateless or undocumented and left their mothers stripped of the practical content of citizenship even while retaining its formal label (Abu Ras et al., 2022). Comparable analyses have examined recent legislative developments in Oman and elsewhere in the region (Kanchana, 2023). What the literature shows is that in high income states with generous welfare entitlements tied to nationality, the material stakes of transmission rules are unusually high, and the demographic and fiscal arguments against reform are correspondingly loud. It should be noted that the position is not uniform across the region and that several states have introduced partial measures, such as facilitated residence, access to public education, or discretionary naturalisation pathways for the children of national mothers. These measures relieve hardship. They do not confer equality, because they leave the underlying entitlement in the gift of the state. 8.6 The African picture The African continent presents the widest variation. A number of states have achieved full gender equality in their nationality laws, in some cases through constitutional reform following democratic transition. Others retain provisions of colonial origin that have never been revisited. The comparative literature is well developed, most notably in Bronwen Manby's study of the law of belonging across the continent, which demonstrates how colonial nationality codes were transplanted, how they interacted with ethnic and territorial politics after independence, and how gender discrimination in nationality law sits alongside, and often reinforces, discrimination on grounds of ethnicity and race (Manby, 2021). Recent analysis by Beninger and Manjoo (2023) surveys the causes and consequences of gender discrimination in nationality law across the continent and evaluates the legal responses available, including through the African regional human rights system and the Protocol to the African Charter on the Rights of Women, commonly known as the #Maputo_Protocol, whose Article 6 addresses the nationality rights of married women directly. Sierra Leone's reform of its citizenship legislation, and comparable movement elsewhere, shows that change is possible even in states with limited administrative capacity. Eswatini's constitutional provision restricting transmission to fathers, criticised repeatedly by treaty bodies, shows how a constitutional entrenchment can make reform far harder than a statutory rule would be. 8.7 Litigation as an engine of reform Where legislatures stall, courts have sometimes moved. The judgment of the Botswana Court of Appeal in the case brought by #Unity_Dow, decided in 1992, remains the most cited example. Dow, a Botswanan woman married to a foreign man, challenged citizenship legislation under which her children could not acquire her nationality. The court held that the law was inconsistent with constitutional guarantees of equality, and the legislation was subsequently amended. The case demonstrated that domestic constitutional equality guarantees can do the work that international treaty obligations, lacking enforcement machinery, cannot. Regional courts have added to this. The European Court of Human Rights, in Genovese v Malta, held that although the European Convention on Human Rights contains no free standing right to nationality, an arbitrary denial of citizenship may engage the right to respect for private life read with the prohibition of discrimination. The Inter American Court of Human Rights, in the case of the Yean and Bosico children against the Dominican Republic, addressed the denial of birth registration and nationality to children of Haitian descent and articulated obligations of non discrimination in access to nationality. The African Court, in Anudo, affirmed the right to a nationality as a matter of customary international law and placed the burden on the state to prove that an individual is not its national. Taken together, these decisions constitute an emerging body of #judicial_review of nationality decisions. They do not yet amount to a coherent doctrine, and their reach depends on the willingness of domestic systems to give effect to them. But they show that the sovereignty of the state over nationality, once treated as absolute, is now constrained. 9. Consequences of the Remaining Gaps The point of studying nationality law is that nationality determines almost everything else. This section sets out what is at stake. 9.1 Statelessness The most severe consequence is statelessness. A stateless person is not merely a person without a passport. They are a person whom no state is obliged to admit, protect, or claim. They may be unable to register a birth, enrol in school, sit an examination, obtain a formal job, open a bank account, marry lawfully, own property, access public healthcare, travel, or vote. They are exposed to arbitrary detention because there is nowhere to deport them to, and to indefinite detention for the same reason. Gender discrimination in nationality law is one of the leading causes of statelessness worldwide, and the causal mechanism is direct. Where only fathers may transmit, any failure of paternal transmission produces a stateless child (Krishan and Sharma, 2023; Beninger and Manjoo, 2023). 9.2 Intergenerational effects Statelessness reproduces itself. A stateless mother cannot transmit a nationality she does not have. A stateless child who reaches adulthood without regularisation will very likely have stateless children. The literature refers to this as #intergenerational_harm, and it is the reason that reforms which are not retroactive leave a permanent residue. Malaysia's non retroactive amendment, for example, does not merely disadvantage the individuals born before the cut off. It creates a cohort whose own children may inherit the disadvantage. 9.3 Socioeconomic and psychological harm Recent empirical work has begun to document the human cost with more precision. Women who cannot confer nationality on their children report guilt, anxiety, insomnia, and depression. Families are separated across borders for years. Mothers postpone or forgo travel, career opportunities, and even medical treatment because of the fragility of their family's legal position. Children internalise their exclusion. In states where public healthcare and education are tied to nationality, the exclusion has measurable effects on health and educational attainment (Abu Ras et al., 2022). There is also a labour market dimension. Where a foreign spouse cannot obtain a work permit, the national wife becomes the sole earner in circumstances where she may face her own labour market disadvantages. Gendered nationality law thus interacts with gendered economic structures to compound disadvantage. 9.4 The wider damage to the rule of law There is a further, less often noted consequence. A legal system that treats women as partial citizens teaches everyone in that system, including its officials, that women are partial citizens. The discriminatory rule is not confined to the nationality file. It radiates. It informs how a registrar treats an unmarried mother, how a court weighs a woman's testimony about her own child's paternity, and how a family regards the value of a daughter. Nationality law is a statement about who counts. When it is written unequally, it teaches inequality. 10. Assessing the Convention: Overtaken but Not Obsolete 10.1 The case that the Convention is obsolete The case is easy to state. The 1957 Convention has few parties. It has no monitoring body. Its dispute settlement clause has never been used. Every obligation it contains is contained, in stronger form, in CEDAW, which binds almost every state on earth. Its Article 3 is asymmetrical in a way that a modern instrument would not tolerate. Its language is dated. Its silence on the transmission of nationality means that it cannot address the principal harm that gender discriminatory nationality laws now cause. On this view, the Convention is a museum piece. It should be taught as history and otherwise ignored. 10.2 The case that it still matters Three arguments run the other way. It remains in force and it still binds. For the states that are parties, the obligations in Articles 1 and 2 are live legal obligations. Where a state party is also a state that has entered a reservation to CEDAW Article 9, the 1957 Convention may supply an obligation that the reservation was designed to exclude. This is not a hypothetical point. Reservations are instrument specific. A reservation to CEDAW does not affect obligations under a different treaty to which the state is separately bound. Advocates and litigators should check whether the older Convention offers an unreserved route to the same result. It supports the interpretation of other instruments. The 1957 Convention is evidence of what states understood the right to a nationality in Article 15 of the UDHR to require. Where a court or treaty body is asked whether automatic loss of nationality on marriage is arbitrary, the existence of a dedicated universal treaty saying that it is provides strong support. This interpretive function does not depend on the number of parties. It teaches a lesson about the architecture of equality. The most valuable thing about the Convention today may be pedagogical. It shows, with unusual clarity, the difference between a negative right and a positive one, and it shows what happens when a reform movement secures the first and postpones the second. The 1957 Convention gave women the right not to lose their nationality. It did not give them the right to give it. Sixty nine years later, that is still the fault line. 10.3 A note on measuring treaty success It is tempting to judge a treaty by counting ratifications. The literature on treaty compliance suggests this is a poor measure. What matters is whether a treaty changes behaviour, and behaviour change depends on domestic legal mobilisation, on the existence of institutions willing to enforce, and on the political costs of non compliance (Comstock, 2021; Comstock, 2024; Mulesky, Sandholtz and Zvobgo, 2024). By that standard, the 1957 Convention was moderately successful in the domain it actually addressed. Automatic loss of nationality on marriage has become rare. The treaty was one of several causes. It was not nothing. 11. Obstacles to Completing the Reform Why do roughly two dozen states still deny women equal rights to confer nationality on their children, decades after accepting an international obligation to grant them? Five obstacles recur. 11.1 Sovereignty and the political economy of membership Nationality is the last redoubt of #state_sovereignty. States accept international constraints on trade, on the environment, on the treatment of prisoners. They resist constraints on who belongs. The reason is that membership rules determine the composition of the political community, and therefore the distribution of political power and public resources. Reforming nationality law is never merely a technical exercise. It is a decision about who gets to be counted. In states where nationality carries substantial material entitlements, such as subsidised housing, healthcare, education, land rights, or a share of resource revenues, the fiscal argument against expansion is made openly. In states with contested demographic balances between ethnic or religious groups, the argument is made quietly and is often more powerful. 11.2 Personal status law and the family In many jurisdictions, nationality rules are entangled with #personal_status_law governing marriage, divorce, guardianship, and inheritance, and that body of law is frequently treated as beyond the reach of ordinary legislative revision because of its religious or customary foundations. Reservations to CEDAW Articles 9, 15, and 16 are often framed in exactly these terms. The argument is that equality in nationality would disturb an integrated family law system in which the father is the legal head of household and the transmitter of lineage. This argument deserves engagement rather than dismissal. It is worth noting, first, that the states which invoke it are not uniform in their practice, and that several states with similar legal and religious traditions have reformed their nationality laws without disturbing personal status codes. Nationality is a public law relationship between individual and state. Lineage within a family is a private law relationship. The two can be separated, and in a number of jurisdictions they have been. Second, the practical effect of the argument is to make the child bear the cost of a doctrinal position held by adults. 11.3 Security framing Since the early 2000s, nationality has been increasingly discussed in the vocabulary of security. Where reform is framed as opening a channel for undesirable persons to acquire citizenship, it becomes politically expensive. The Malaysian amendment illustrates the dynamic: the progressive reform for mothers was bundled with restrictive provisions aimed at other groups, in what appears to have been a political trade. 11.4 The reservations regime The law of reservations permits a state to accept a treaty while excluding the provisions it dislikes, provided the reservation is not incompatible with the object and purpose of the treaty. In practice, the incompatibility test has proved difficult to enforce. Objections by other states have limited legal effect. The result is that a state may present itself as a party to CEDAW while having excluded the very article that would require it to change its nationality law. Recent scholarship shows that this is not a formality: reserving states show systematically weaker improvement in women's rights outcomes (Kreutzer and Mitchell, 2024). 11.5 Administrative practice Finally, even where the law is equal, the administration may not be. Officials with discretion exercise it. Documentary requirements that are formally neutral may be impossible for some women to satisfy. Reform of the statute is a necessary condition of equality. It is not a sufficient one. Attention to #birth_registration systems, to #legal_identity documentation, and to the training of front line officials is indispensable, and it is frequently neglected in the drafting of reform packages. 12. Recommendations The following recommendations follow from the analysis. 12.1 For legislatures Legislate symmetry, not exceptions. A reform that allows a mother to transmit nationality only where the father is unknown, stateless, or absent is not equality. It is a safety net, and it depends on discretionary determinations that are frequently made badly. The legislative goal should be a rule under which either parent transmits nationality on identical terms. Extend equality to the spouse. Reform should address the position of the foreign husband as well as the foreign wife. Facilitated naturalisation and residence should not depend on the sex of the national spouse. Make reform retroactive, or provide a full remedy. Prospective only reform creates a permanently disadvantaged cohort. Where full retroactivity is politically impossible, the state should at minimum establish a simple, free, time bound registration procedure for those born before the reform, with written reasons for refusal and a right of appeal. Do not bundle progress with regression. The practice of pairing a gender equality reform with restrictions on other vulnerable groups converts a human rights advance into a political transaction and undermines its legitimacy. Legislate for administration. Reform statutes should be accompanied by implementing regulations, published procedures, staff training, and accessible birth registration, especially for unmarried mothers and for births occurring outside health facilities. 12.2 For treaty bodies and international institutions Press for the withdrawal of reservations to Article 9. The evidence that reservations depress outcomes is now reasonably strong, and the evidence that sustained pressure produces withdrawals is encouraging (Boyes et al., 2024; Kreutzer and Mitchell, 2024). Use the older Convention where it helps. Treaty bodies and litigants should recall that a state party to the 1957 Convention has accepted obligations that a CEDAW reservation does not touch. Integrate nationality into the sustainable development agenda. The commitments on legal identity and on gender equality provide a policy vehicle for nationality reform, and framing the issue in developmental terms sometimes succeeds where the rights framing has failed (#SDG_5, #SDG_16). 12.3 For civil society Litigate. The Botswana, Malaysia, and Nepal experiences show that constitutional litigation can force movement, both directly, through judgments, and indirectly, by creating a deadline that concentrates legislative attention. Centre the affected. The most effective campaigns of the past decade have been led by the women and children directly affected, and the historical record of the interwar campaigns shows the same pattern (Brennan, 2020). Research and advocacy that speaks about stateless people without them reproduces the exclusion it seeks to end (Boone, 2023). Guard against partial victories being declared final. When a state equalises transmission to children but leaves spousal rules untouched, the campaign is not over. 12.4 For educators and researchers Teach the 1957 Convention as a text, not a date. Read Article 3 alongside Article 1 and ask students why the drafters could see one asymmetry and not the other. It is a good exercise in the sociology of law, and it prepares students to notice the asymmetries in the instruments being drafted today. 13. Limitations and Directions for Future Research This article has three main limitations. It is doctrinal and comparative rather than empirical. It does not present original data on outcomes, and its comparative case studies are drawn from published sources rather than from fieldwork. Its coverage is uneven. Some regions, notably Latin America and the post Soviet space, receive less attention than they merit, and the treatment of statelessness arising from state succession, from #denationalization for security reasons, and from racial or ethnic discrimination is deliberately brief even though these phenomena intersect with the subject (Foster and Baker, 2021; Jain, 2022; Kenny, 2020). It writes about statelessness largely from the outside. The field has been criticised, rightly, for a power imbalance in which those who study statelessness are rarely those who experience it (Boone, 2023). Four lines of #future_research seem especially promising. First, quantitative work testing whether ratification of the 1957 Convention, controlling for CEDAW status and other factors, has any independent association with the timing of domestic reform. To the author's knowledge this has not been done. Second, implementation studies of recent reforms. The Malaysian amendment, whose automatic transmission rule was reported to be taking effect in 2026, offers a natural experiment. What proportion of eligible children actually obtain documentation? How long does it take? Who is refused? Third, work on the interaction between nationality law and digital identity systems. As states move to biometric registration and digital civil registries, the risk is that the excluded become more precisely, and more permanently, excluded. Fourth, comparative work on the political conditions under which nationality reform succeeds. The existing literature is rich in doctrine and thin in political explanation. 14. Conclusion The Convention on the Nationality of Married Women did something that had never been done at the universal level. It said that a woman's #nationality is hers. Not her husband's to lend, not her husband's to withdraw, not her marriage's to determine. Marriage, divorce, and a husband's naturalisation were declared legally irrelevant to a wife's status. In 1957 this was a radical proposition. Twenty seven years earlier, at The Hague, the international community had refused to say it. The Convention was, at the same time, a partial instrument. It secured a negative right and left the positive one for later. It protected the woman and forgot the child. It built a privileged route to nationality for the foreign wife and none for the foreign husband. It created no institution to watch over its own implementation. It has been ratified by fewer than half the states of the world, and it has been superseded, in practice, by a treaty that says everything it says and more. The right conclusion is not that the Convention failed. It is that it did what a first instrument can do. It established that #dependent_nationality was not a natural feature of the legal order but a policy choice, and that the choice could be made differently. Everything since, including Article 9 of CEDAW, General Recommendation 32, the Maputo Protocol, the judgments of regional courts, and the constitutional litigation now proceeding in national courts across Asia and Africa, has been built on that foundation. The unfinished part of the project is easy to state and hard to complete. Roughly two dozen states still do not permit a mother to give her child her nationality on the same terms as a father. Roughly fifty maintain some form of gender discrimination in their nationality law. Every one of those states has, in one form or another, accepted an international obligation not to do so. The gap between what states have promised and what their statute books say is where the work now lies. The lesson of 1957 is that the gap closes when three things happen together. A clear legal norm exists. Domestic institutions, especially courts, are willing to give it effect. And the people affected organise and refuse to accept a safety net in place of #equal_protection. The women who were thrown out of the Peace Palace in 1930 understood that. It took twenty seven more years for the international community to agree with them. It is taking rather longer to finish the job. Hashtags #Convention_on_the_Nationality_of_Married_Women #NationalityOfMarriedWomen1957 #EqualNationalityRights #WomensRightToNationality #GenderEqualCitizenship #EndStatelessness #CEDAW_Article_9 #NationalityLawReform #StatelessnessPrevention #WomenAndCitizenship #InternationalWomensLaw #LegalBelonging #GenderJusticeInLaw #HumanRightsResearch #STULIB References Abdulkadir, M. I. (2021). The role of law clinics in the fight against statelessness by the United Nations High Commissioner for Refugees (UNHCR) in Nigeria. International Journal of Clinical Legal Education, 28(1), 119 to 153. Abu Ras, W., Elzamzamy, K., Burghul, M. M., Al Merri, N. H., Alajrad, M., and Kharbanda, V. A. (2022). Gendered citizenship, inequality, and well being: The experience of cross national families in Qatar during the Gulf Cooperation Council crisis (2017 to 2021). International Journal of Environmental Research and Public Health, 19(11), 6638. https://doi.org/10.3390/ijerph19116638 Almutawa, A. (2021). The Arab Court of Human Rights and the enforcement of the Arab Charter on Human Rights. Human Rights Law Review, 21(3), 506 to 532. Beninger, C., and Manjoo, R. (2023). The impact of gender discrimination on statelessness: Causes, consequences and legal responses. African Human Mobility Review, 8(3). https://doi.org/10.14426/ahmr.v8i3.1125 Bloom, T. (2023). Noncitizen Power: Agency and Politics in the Age of Migration. London: Bloomsbury. Bloom, T., and Kingston, L. N. (Eds.) (2021). Statelessness, Governance, and the Problem of Citizenship. Manchester: Manchester University Press. Boone, J. (2023). A power imbalance in academic scholarship on statelessness: A thematic analysis of the academic literature on statelessness from 2014 onwards. Statelessness and Citizenship Review, 5(1), 76. Boyes, C., Eldredge, C. D., Shannon, M., and Zvobgo, K. (2024). Social pressure in the international human rights regime: Why states withdraw treaty reservations. British Journal of Political Science, 54(1). Brennan, D. (2020). Feminist foresight in statelessness: Century old citizenship equality campaigns. Statelessness and Citizenship Review, 2(1), 43 to 63. Comstock, A. L. (2021). Committed to Rights: UN Human Rights Treaties and Legal Paths for Commitment and Compliance. Cambridge: Cambridge University Press. Comstock, A. L. (2024). Signing CEDAW and women's rights: Human rights treaty signature and legal mobilization. Law and Social Inquiry, 49(2), 1222 to 1255. Foster, M., and Baker, T. R. (2021). Racial discrimination in nationality laws: A doctrinal blind spot of international law? Columbia Journal of Race and Law, 11(1), 83 to 142. Gunther, J., and Hinterhuber, E. M. (2024). Editorial: Citizenship and democratization, perspectives from different gender theoretical approaches. Frontiers in Sociology, 8. https://doi.org/10.3389/fsoc.2023.1343538 Jain, N. (2022). Manufacturing statelessness. American Journal of International Law, 116(2), 237 to 288. Kanchana, R. (2023). Oman: Recent developments in the protection of the rights of migrant workers and of women. Yearbook of Islamic and Middle Eastern Law Online, 22(1), 298 to 303. Kenny, C. (2020). Legislated out of existence: Mass arbitrary deprivation of nationality resulting in statelessness as an international crime. International Criminal Law Review, 20(6), 1026 to 1067. Kreutzer, W. (2024). Introducing the CEDAW ROCS dataset: A dataset on the United Nations treaty the Convention on the Elimination of All Forms of Discrimination against Women, reservations and objections of committed states. Foreign Policy Analysis, 21(1). Kreutzer, W., and Mitchell, S. M. (2024). The three R's of CEDAW commitment: Ratification, reservation, and rejection. International Interactions, 50(3), 418 to 447. https://doi.org/10.1080/03050629.2024.2327996 Krishan, R., and Sharma, R. (2023). Proliferation of statelessness and gender discriminatory nationality laws. International Journal of Law Management and Humanities, 6(2), 524 to 532. https://doij.org/10.10000/IJLMH.114392 Manby, B. (2021). Citizenship in Africa: The Law of Belonging. Oxford: Hart Publishing. Massari, M., Gatta, G., Miceli, S., and Cabras, F. (2025). Embodied experiences of lived citizenship beyond the European privilege: The case study of Romanian women farmworkers in eastern Sicily. Frontiers in Sociology, 10. https://doi.org/10.3389/fsoc.2025.1558541 Mulesky, S., Sandholtz, W., and Zvobgo, K. (2024). Do human rights treaty obligations matter for ratification? Journal of Human Rights, 23(1), 1 to 18. Roberts, J. (2022). Review of Mira L. Siegelberg, Statelessness: A Modern History. European Journal of International Law, 33(2), 695 to 720. Siegelberg, M. L. (2020). Statelessness: A Modern History. Cambridge, Massachusetts: Harvard University Press. Sperfeldt, C. (2022). The perpetual foreigner: Statelessness among the Vietnamese minority in Cambodia. Georgetown Journal of International Affairs. von Rutte, B. (2022). The Human Right to Citizenship: Situating the Right to Citizenship within International and Regional Human Rights Law. Leiden: Brill Nijhoff. Primary legal instruments and interpretive materials Committee on the Elimination of Discrimination against Women, General Recommendation No. 32 on the gender related dimensions of refugee status, asylum, nationality and statelessness of women (2014). Convention on Certain Questions Relating to the Conflict of Nationality Laws, The Hague, 12 April 1930. Convention on the Elimination of All Forms of Discrimination against Women, 18 December 1979, 1249 UNTS 13. Convention on the Nationality of Married Women, New York, 20 February 1957, opened for signature by General Assembly resolution 1040 (XI) of 29 January 1957, entered into force 11 August 1958, 309 UNTS 65. Convention on the Nationality of Women, Montevideo, 26 December 1933. Convention on the Reduction of Statelessness, 30 August 1961, 989 UNTS 175. Convention on the Rights of the Child, 20 November 1989, 1577 UNTS 3. Convention Relating to the Status of Stateless Persons, 28 September 1954, 360 UNTS 117. International Covenant on Civil and Political Rights, 16 December 1966, 999 UNTS 171. Protocol to the African Charter on Human and Peoples' Rights on the Rights of Women in Africa, 11 July 2003. United Nations High Commissioner for Refugees and UN Women, Background Note on Gender Equality, Nationality Laws and Statelessness (2025). Universal Declaration of Human Rights, General Assembly resolution 217 A (III), 10 December 1948. Cases cited Anudo Ochieng Anudo v United Republic of Tanzania, African Court on Human and Peoples' Rights, Application No. 012/2015, judgment of 22 March 2018. Attorney General of Botswana v Unity Dow, Botswana Court of Appeal, 1992. Case of the Girls Yean and Bosico v Dominican Republic, Inter American Court of Human Rights, judgment of 8 September 2005. Genovese v Malta, European Court of Human Rights, Application No. 53124/09, judgment of 11 October 2011. Suriani Kempe and Others v Government of Malaysia, Federal Court of Malaysia, proceedings concluded by consent order, March 2025.

The Convention for the Suppression of the Traffic in Persons and of the Exploitation of the Prostitution of Others, approved by the United Nations General Assembly on 2 December 1949 and in force since 25 July 1951, is widely described as the first genuinely global treaty devoted to the suppression of #human_trafficking and #forced_prostitution. This article offers a detailed legal and historical assessment of that instrument for a student and early researcher audience. It traces the treaty back to the four #League_of_Nations era agreements of 1904, 1910, 1921 and 1933, examines the drafting process that unfolded within the young #United_Nations between 1946 and 1949, and then works carefully through the twenty eight articles of the Convention itself. The analysis shows that the treaty did three things at once. It criminalised #procuring and #brothel_keeping regardless of the #consent of the person exploited. It required states to dismantle #state_regulated_prostitution, including registration systems and special identity documents for women in prostitution. And it obliged states to build prevention, protection and #repatriation services. The article then sets out the weaknesses that limited the treaty's influence: the absence of any definition of trafficking, the narrow focus on sexual exploitation to the exclusion of #forced_labour, the deference to #domestic_law, the lack of a treaty body or reporting mechanism with teeth, and a modest ratification record of roughly eighty two states. The final sections situate the 1949 instrument in relation to the 2000 #Palermo_Protocol, the 2005 Council of Europe Convention, the European Union anti trafficking directives and the Sustainable Development Goals, and argue that the 1949 text is neither dead law nor a usable modern framework, but a foundational document whose moral vocabulary continues to shape contemporary policy debates about #abolitionism, sex work and the limits of criminal law. The article concludes with reform pathways and research directions. Keywords: human trafficking, forced prostitution, 1949 Convention, international law, abolitionism, Palermo Protocol, exploitation, United Nations, sexual exploitation, treaty law 1. Introduction Human trafficking is often described as a modern problem. It is not. Long before the words "trafficking in persons" appeared in a criminal code, states were already negotiating treaties about the movement of women and girls across borders for the purpose of commercial sex. The story of #international_law on trafficking begins in the early twentieth century, and its first properly global chapter closes with a single document: the Convention for the Suppression of the Traffic in Persons and of the Exploitation of the Prostitution of Others. That Convention was approved by the #General_Assembly of the United Nations on 2 December 1949 through #Resolution_317 (IV). It was opened for signature in March 1950 and entered into force on 25 July 1951. It is a short treaty, only about two and a half thousand words spread across twenty eight articles and a Final Protocol, yet its ambitions were very large. It sought to end the traffic in persons for prostitution everywhere in the world, to punish everyone who profited from the prostitution of another human being, and to sweep away the licensing and registration systems through which many states had, for decades, quietly organised and taxed commercial sex. For students of international law, the 1949 Convention is important for several reasons that go well beyond its own operative provisions. First, it is a genuine milestone in institutional history. The four earlier agreements on this subject were negotiated under European leadership and later under the #League_of_Nations. The 1949 text was the first to be produced by the United Nations, a body with far broader membership, and it consolidated the earlier instruments into a single consolidated framework. As between parties, it superseded the agreements of 1904, 1910, 1921 and 1933. Second, it marks a decisive shift in language and in moral framing. The earlier treaties spoke of the "white slave traffic," a phrase soaked in racial anxiety and in the imagery of the innocent European girl carried away to a foreign brothel. The 1949 Convention dropped that phrase. It used #gender_neutral_language, speaking simply of "persons," and it applied whether or not a border was crossed. In doing so it quietly severed the conceptual link between trafficking and international movement, a link that would later be reintroduced, in practice if not in law, by the enforcement culture built around the 2000 Protocol. Third, and most controversially, it took a side in a long running argument about how the state should treat prostitution. The Convention's preamble declares that prostitution and the accompanying evil of the traffic in persons for the purpose of prostitution are incompatible with the dignity and worth of the human person and endanger the welfare of the individual, the family and the community. This is not a neutral sentence. It embeds a specific ideology, usually called #abolitionism, into the fabric of a universal treaty. Yet the Convention did not criminalise the sale of sex itself and did not criminalise the purchaser. It criminalised the third party who organises, profits from or facilitates the prostitution of another. That combination, punishing the exploiter while refusing to punish the exploited and refusing to license the trade, is what historians of the field call regulation abolitionism. Fourth, the treaty is a case study in how international law can be simultaneously influential and ineffective. Roughly eighty two states became parties, a number that looks small beside the near universal ratification of the later trafficking protocol. The Convention created no committee, no periodic review, no complaints procedure and no interpretive body. Its main enforcement idea was that states should file reports with the Secretary General and settle disputes through the #International_Court_of_Justice, an option that many parties promptly disclaimed by way of #reservations. And yet its normative vocabulary, especially the claim that the consent of the exploited person is legally irrelevant, travelled far beyond its formal membership and continues to shape national law in states that never ratified it at all. This article asks a simple question with a complicated answer. What exactly did the 1949 Convention achieve, what did it fail to achieve, and what remains of it today? The argument advanced here has three parts. The Convention was a serious and in some ways admirable attempt to shift international attention from the policing of women to the punishment of those who exploit them. It was, at the same time, structurally incapable of delivering that shift, because it lacked a definition, lacked a monitoring architecture and lacked any conception of trafficking outside the sexual sphere. And its most enduring legacy is not doctrinal but discursive: the moral grammar of the 1949 text, its language of dignity, its suspicion of consent and its hostility to the organised sex trade, still structures the way governments, campaigners and scholars argue with each other in 2026. 2. Research Problem, Questions and Method 2.1 The research problem Most contemporary teaching on trafficking law begins in the year 2000. Students learn the three element definition of the Palermo Protocol, they learn the "3P" framework of prevention, protection and prosecution, and they move quickly to national implementation. The period before 2000 is compressed into a sentence or two about "earlier and less effective instruments." This creates a distorted picture. It suggests that the international community had no concept of trafficking before the Protocol, which is false, and it hides the ideological choices that were made in 1949 and that continue to shape law today. The research problem addressed here is therefore the underexamination of the 1949 Convention as a legal instrument in its own right, rather than as a footnote to the Palermo Protocol. 2.2 Research questions The article is organised around five questions. What historical developments produced the 1949 Convention, and what problems were its drafters trying to solve? What obligations does the Convention actually impose, read article by article? What theoretical model of prostitution and exploitation does the treaty adopt, and what internal tensions does that model contain? Why did the Convention attract limited #ratification and limited compliance, and what does this teach us about treaty design? What is the legal status and practical relevance of the Convention in the current international framework, and what lessons does it offer for reform? 2.3 Method The method is doctrinal and historical, supplemented by a critical reading of recent socio legal scholarship. Doctrinal analysis involves close reading of the primary legal texts: the 1949 Convention and its Final Protocol, the four predecessor agreements of 1904, 1910, 1921 and 1933, the 1926 Slavery Convention and its 1956 Supplementary Convention, the International Labour Organization instruments on #forced_labour, the 1979 Convention on the Elimination of All Forms of Discrimination against Women, the 1989 Convention on the Rights of the Child, the 2000 Protocol to Prevent, Suppress and Punish Trafficking in Persons, the 2005 Council of Europe Convention on Action against Trafficking in Human Beings and the European Union directives of 2011 and 2024. Historical analysis draws on published work on the negotiation of the Convention, in particular the archival research showing how civil society organisations, United Nations secretariat officials and a small number of influential states shaped the final text during the late 1940s. Critical analysis engages with the scholarly literature on trafficking, sex work and #modern_slavery, including work that is sharply critical of the abolitionist tradition. The article does not attempt to resolve the underlying normative dispute about commercial sex. It reports the arguments fairly and identifies where the 1949 text takes a position. 2.4 Structure Sections 3 and 4 set out the history. Section 5 provides the article by article legal analysis. Section 6 examines the abolitionist paradigm and its tensions. Sections 7 and 8 assess achievements and failures. Section 9 examines ratification patterns. Section 10 traces the path from 1949 to the present framework. Sections 11 to 13 discuss continuing relevance, contemporary challenges and reform. Section 14 states the limitations of the study and Section 15 concludes. 3. Historical Background: The Road from 1904 to 1937 3.1 The invention of the "white slave traffic" In the last decades of the nineteenth century, European and North American publics became gripped by a story. The story held that innocent young women, usually poor, usually white, were being lured or abducted from their home towns and delivered into foreign brothels by organised networks of procurers. Newspapers ran sensational accounts. Reform societies were founded. Parliaments held inquiries. Historians have long argued that the reality was far more complicated than the story. Much of what was described as abduction was in fact migration, often economically driven, by women who moved to cities or across oceans and who then entered commercial sex because it paid better than the alternatives available to them. The panic nevertheless had real effects. It created a transnational movement, it created a vocabulary and it created political pressure for treaties. Contemporary scholarship treats this episode as a classic #moral_panic, one whose racial framing, expressed in the term #white_slave_traffic, mattered enormously to who was seen as a victim worth rescuing. Two organisations founded in this period matter for the story that follows. The International Abolitionist Federation, established in 1875, campaigned against state licensing of brothels and against the compulsory medical inspection of women in prostitution. The International Bureau for the Suppression of the Traffic in Persons, established in 1899, coordinated national committees. Both bodies would be present, and influential, when the United Nations came to draft the 1949 text. 3.2 The first four treaties The 1904 International Agreement for the Suppression of the White Slave Traffic was essentially administrative. It did not create crimes. It required states to appoint an authority to gather information about the procurement of women and girls for immoral purposes abroad, to watch railway stations and ports, to take statements from foreign women found in prostitution and to arrange repatriation where the woman wished to return home. It was, in effect, a treaty about information sharing and consular assistance. The 1910 International Convention for the Suppression of the White Slave Traffic went further and required criminalisation. States agreed to punish anyone who procured, enticed or led away a woman or girl under age for immoral purposes, even with her consent, and anyone who did the same to a woman of full age by fraud, violence, threats or abuse of authority. The seeds of the later law are visible here. Consent is already irrelevant where a minor is concerned. For adults, consent still matters, and the offence requires coercive or deceptive means. The 1921 International Convention for the Suppression of the Traffic in Women and Children, negotiated under the auspices of the newly created League of Nations, extended protection to boys as well as girls, raised the age of protection and provided for extradition. It also established the machinery that would define the interwar period: an advisory committee on the traffic in women and children, reporting to the League, which gathered national reports and commissioned studies. The 1933 International Convention for the Suppression of the Traffic in Women of Full Age took the decisive doctrinal step. It required states to punish anyone who procured, enticed or led away a woman of full age for immoral purposes in another country, even with her consent. The coercion requirement for adults was gone, at least for cross border cases. The abolitionist position that the #consent of an adult woman cannot cure the wrong of procurement had entered treaty law. 3.3 The unfinished business of 1937 By the late 1930s, the League's committee had prepared a draft convention that would have gone further still, extending the criminalisation of procurement to purely domestic cases, that is, to #internal_trafficking within a single state. This was radical. It meant that international law would reach into the way a state organised commercial sex inside its own borders. The draft was never adopted. The Second World War intervened, the League collapsed and the project was suspended. This unfinished 1937 draft is the direct ancestor of the 1949 Convention. When the United Nations inherited the League's social functions, it inherited this file as well. Understanding that lineage explains why the 1949 text reads the way it does. It is not a fresh start. It is the completion of a project that had been interrupted. 3.4 What the pre 1949 regime had achieved and had not By 1945 the international regime had four treaties, a vocabulary, a set of national committees and an established practice of reporting. What it did not have was a definition of trafficking, a rule against #state_regulated_prostitution, protection for men, any conception of exploitation outside commercial sex, or any meaningful enforcement power. Every one of these gaps would be either addressed or perpetuated by the 1949 Convention, and it is worth keeping the list in mind while reading Section 5. 4. Drafting and Adoption, 1946 to 1949 4.1 Institutional setting When the United Nations was created, responsibility for the League's social treaties passed to the Economic and Social Council and, through it, to the Commission on the Status of Women and the Social Commission. A 1947 protocol transferred the League's functions under the earlier trafficking agreements to the United Nations. The Secretariat then prepared a consolidated draft convention, drawing directly on the four earlier treaties and on the unfinished 1937 draft. 4.2 The actors Recent archival scholarship has shown that the drafting was shaped by three sets of actors. The first were non governmental organisations, above all the International Abolitionist Federation and its allies, who had been campaigning for the abolition of licensed brothels since the 1870s and who saw the new organisation as an opportunity to universalise their programme. For them, 1949 was a triumph. The Federation's own commemorative writing at the time treated the Convention as the finest achievement of its history. The second were officials of the United Nations Secretariat, who drafted, redrafted and mediated, and who were themselves strongly committed to the abolitionist reading of the problem. The third were states, and here the key participants were France and the United Kingdom, together with the Soviet bloc. France had abolished its licensed brothel system in 1946, and this domestic reform gave French delegates both credibility and an interest in exporting the model. The United Kingdom, which had never operated a national licensing system in the metropole but did so in parts of its empire, was more cautious about provisions that would bind colonial territories. 4.3 The Cold War and the colonial clause The negotiation took place at the beginning of the #Cold_War, and the ideological fault lines are visible in the record. On the desirability of abolishing state licensed prostitution there was, strikingly, broad agreement across the East West divide. Socialist delegations and Western delegations alike condemned prostitution in moral terms, though they explained it differently, the former as a product of capitalist exploitation and the latter as a matter of individual and social morality. The sharper disagreement concerned jurisdiction. The Soviet Union attacked the draft's #colonial_clause, the standard provision that allowed a metropolitan state to extend or withhold the application of a treaty in respect of non self governing territories. Soviet delegates argued that such a clause allowed imperial powers to exempt precisely those territories where the exploitation was worst. The argument fused anti imperial politics with moral reform, and it gave newly independent and soon to be independent states a language in which to criticise the colonial powers. The critique did not stop the Convention, but it shaped the debate and it foreshadowed the way that anti trafficking discourse would repeatedly be entangled with geopolitics. 4.4 Adoption The text was approved by the General Assembly on 2 December 1949. The vote was not unanimous, and a significant bloc of states declined to support it. Objections came from several directions. Some states objected to the compulsory jurisdiction of the International Court of Justice in Article 22. Some, with regulated systems, objected to Article 6. Some common law states objected to provisions on extradition and letters of request that sat awkwardly with their procedural traditions. The Convention entered into force on 25 July 1951, ninety days after the second instrument of ratification was deposited, in accordance with Article 24. 5. The Convention Analysed: Structure and Obligations The Convention can be divided into five functional blocks: the preamble and its normative premise; the criminal provisions; the abolition of regulation; the machinery of #international_cooperation; and the social provisions on prevention, protection and repatriation. A sixth block covers final clauses. This section walks through each. 5.1 The preamble and its normative premise The preamble is the ideological heart of the treaty. It states that prostitution and the accompanying evil of the traffic in persons for the purpose of prostitution are incompatible with the dignity and worth of the human person and endanger the welfare of the individual, the family and the community. Three features deserve attention. First, prostitution and trafficking are joined together. Trafficking is described as an evil that accompanies prostitution. There is no space in this sentence for the idea that a person might sell sex without being trafficked, or that trafficking might occur for purposes unconnected with sex. The conflation is deliberate. Second, the standard invoked is #human_dignity, not autonomy, not liberty, not equality and not labour rights. Dignity in this usage is an objective standard. It describes a condition that is degrading whether or not the individual concerned regards it as degrading. This is precisely why individual consent will later be declared irrelevant. Third, the harm is described as social as well as individual. Prostitution endangers the welfare of the family and the community. This communitarian framing gave states a public order justification for intervention and it is one reason why the treaty attracted support from governments that were not otherwise interested in women's rights. 5.2 The criminal provisions: Articles 1 to 4 Article 1 requires each party to punish any person who, to gratify the passions of another, procures, entices or leads away another person for purposes of prostitution, even with the consent of that person, or who exploits the prostitution of another person, even with the consent of that person. Four points follow. The subject of the offence is the third party, not the person selling sex. The phrase "to gratify the passions of another" makes this explicit. The Convention is aimed at the #procuring intermediary. The offence covers both recruitment and ongoing exploitation. Exploiting the prostitution of another is an independent offence and does not require any prior act of recruitment. Consent is expressly excluded as a defence. This is the single most important sentence in the treaty and it is repeated twice for emphasis. It represents the culmination of the doctrinal movement that began in 1910 and continued in 1933. There is no requirement of force, fraud, coercion or crossing a border. The offence is complete on the facts of procurement or exploitation alone. Article 2 requires punishment of any person who keeps or manages, or knowingly finances or takes part in the financing of, a brothel, or who knowingly lets or rents a building or other place, or any part of it, for the purpose of the prostitution of others. This is the provision against #brothel_keeping, and it also reaches landlords and financiers. Its practical effect, in states that implemented it faithfully, was to push commercial sex out of managed premises and into the street or into informal arrangements. Whether this made sellers of sex safer or more vulnerable is one of the most contested empirical questions in the field, and Section 8 returns to it. Article 3 requires that, to the extent permitted by domestic law, attempts to commit the Article 1 and 2 offences, and acts preparatory to their commission, also be punished. Article 4 requires that intentional participation in those offences be punishable, and provides that acts of participation shall be treated as separate offences to the extent required by domestic law to prevent impunity. Note the recurring qualification: "to the extent permitted by domestic law." This phrase, and the parallel provision in Article 12, is one of the treaty's central weaknesses. It converts what look like firm obligations into obligations conditioned on the state's own legal arrangements. 5.3 The abolition of regulation: Article 6 Article 6 is short and radical. Each party agrees to take all necessary measures to repeal or abolish any existing law, regulation or administrative provision by virtue of which persons who engage in, or are suspected of engaging in, prostitution are subject to special registration, to the possession of a special document, or to any exceptional requirements for supervision or notification. This provision attacked a system that was, at the time, standard practice across much of continental Europe, Latin America and colonial Asia and Africa. Under that system, women in prostitution were registered by the police, issued with a card, subjected to compulsory medical inspection and confined to designated houses or districts. The system was defended on public health and public order grounds. Article 6 declared the whole apparatus unlawful for parties to the Convention. Its rationale was twofold and the two rationales are in tension. The first rationale was libertarian in a narrow sense. Registration and surveillance created a legally marked class of women, stripped of ordinary civil status, permanently at the mercy of the police, and Article 6 was meant to free them from that #stigma and from that discretionary power. The second rationale was suppressive. Licensed brothels were understood as the demand infrastructure of the traffic, and closing them was expected to dry up the market. The provision is therefore simultaneously a human rights guarantee for people in prostitution and an instrument of #criminalisation directed at the organised trade around them. Recent historical scholarship has described exactly this combination as the "tensions of abolitionism," and has argued that in practice the suppressive logic tended to defeat the emancipatory one. Removing registration did not remove police attention. It relocated it, often into vagrancy, public nuisance, immigration and public morality offences that continued to fall on the seller of sex. 5.4 Cooperation, jurisdiction and procedure: Articles 5, 7 to 15 Article 5 provides that where injured persons are entitled under domestic law to be parties to criminal proceedings, aliens shall be entitled to do so on the same terms as nationals. This is an early, if modest, gesture towards #access_to_justice for foreign victims. Article 7 provides that previous convictions in foreign states for the Convention's offences shall, to the extent permitted by domestic law, be taken into account for establishing recidivism and for disqualifying the offender from civil rights. Article 8 deems the offences in Articles 1 and 2 to be extraditable offences in any existing or future extradition treaty between parties. Article 9 provides that where a state does not permit the #extradition of its own nationals, a national who has returned home after committing such an offence abroad shall be prosecuted and punished in the courts of his own state. Article 10 protects against double jeopardy. Article 11 preserves the position of states that do not make extradition conditional on a treaty. Article 12 states that the Convention does not affect the principle that the offences to which it refers shall in each state be defined, prosecuted and punished in conformity with its domestic law. This is the sovereignty saving clause, and it is very broad. Article 13 governs #letters_of_request, setting out three permitted channels of transmission: direct communication between judicial authorities, communication between ministries of justice, or transmission through diplomatic or consular representatives. It provides that execution of letters of request shall not give rise to claims for reimbursement other than expert expenses, and that nothing in the article requires a state to adopt methods of proof contrary to its own law. This is a workable but pre modern model of #mutual_legal_assistance, far less developed than the machinery later created under the United Nations Convention against Transnational Organized Crime. Article 14 requires each party to establish or maintain a service charged with the coordination and centralisation of the results of investigations into the Convention's offences, and provides that such services should compile information calculated to facilitate prevention and punishment and should stay in close contact with corresponding services in other states. This is the ancestor of the modern national anti trafficking coordination body. Article 15 provides that these services shall, to the extent permitted by domestic law and where they judge it desirable, furnish to their counterparts abroad particulars of offences and attempts, and particulars of searches, prosecutions, arrests, convictions, refusals of admission and expulsions, together with the movements of the persons concerned. The conditionality here is heavy: the obligation applies only so far as domestic law permits and only where the national service considers it desirable. 5.5 Prevention, protection and repatriation: Articles 16 to 20 Article 16 obliges parties to take or encourage, through public and private educational, health, social, economic and other related services, measures for the #prevention of prostitution and for the #rehabilitation and social adjustment of the victims of prostitution and of the offences covered by the Convention. This is a remarkable provision for 1949. It recognises, at least implicitly, that prostitution has social and economic causes and that the response cannot be purely penal. It is also frustratingly vague. There is no standard, no benchmark and no obligation of result. Article 17 requires parties, in connection with immigration and emigration, to adopt or maintain measures to check the traffic in persons for prostitution: to make regulations for the protection of immigrants and emigrants, in particular women and children, at places of arrival and departure and in transit; to arrange appropriate publicity warning the public of the dangers of the traffic; to keep watch at railway stations, airports, seaports, and in transit generally, for persons who may be carrying out the traffic; and to notify the authorities of the state of destination of the arrival of persons who appear to be principals, accomplices or victims. Article 18 requires parties, under the conditions laid down by domestic law, to take statements from alien persons in prostitution in order to establish their identity and civil status and to discover who induced them to leave their state, with the information communicated to the authorities of their state of origin with a view to eventual repatriation. Article 19 deals with #repatriation. Parties undertake, pending arrangements for repatriation of destitute persons who are victims of the Convention's offences, to make suitable provision for their temporary care and maintenance, and to repatriate such persons if they desire it, or if repatriation is demanded by a person exercising authority over them, or if there is an expulsion order. Repatriation is to take place only after agreement with the state of destination as to identity and nationality, and as to place and date of arrival. Where the person cannot repay the cost, it is borne by the state of residence up to the frontier and by the state of origin thereafter. Article 19 is worth pausing on, because it is where the Convention's protective ambitions are most visibly compromised by the logic of border control. The provision does contemplate temporary care and maintenance, and it does say that repatriation should follow the person's wishes. But it also contemplates repatriation demanded by "a person exercising authority" over the individual, and it contemplates repatriation pursuant to an expulsion order. The line between #victim_protection and removal of an unwanted foreigner is left uncomfortably blurred. This ambiguity was not resolved by international law until the residence permit and reflection period provisions of the 2005 Council of Europe Convention. Article 20 requires parties, if they have not already done so, to take the necessary measures for the supervision of #employment_agencies, in order to prevent persons seeking employment, in particular women and children, from being exposed to the danger of prostitution. This provision, largely ignored for half a century, has become newly relevant in an era of exploitative recruitment intermediaries in global labour markets. 5.6 Implementation and final clauses: Articles 21 to 28 Article 21 requires parties to communicate to the Secretary General their laws and regulations relating to the subject matter of the Convention, and an #annual_reports obligation concerning measures taken and results achieved. The Secretary General publishes the information periodically and communicates it to the members of the United Nations and to non member states to which the Convention has been officially communicated. This is the entire monitoring architecture. There is no committee of independent experts. There is no country visit. There is no individual complaints mechanism. There is no state to state complaint mechanism other than the general dispute settlement clause. The #monitoring_gap here is fundamental, and it is the single clearest structural difference between the 1949 Convention and every major human rights treaty adopted afterwards. Article 22 provides that a dispute between parties relating to the interpretation or application of the Convention which is not settled by negotiation shall be referred, at the request of any one of the parties to the dispute, to the #International_Court_of_Justice unless the parties agree to another mode of settlement. A number of ratifying states entered reservations to this clause, and objections to it explain some non ratifications. Articles 23 to 26 deal with signature, ratification, accession, entry into force and territorial application. Article 27 requires parties to take the legislative or other measures necessary to ensure the application of the Convention. Article 28 provides that the Convention terminates and replaces, as between the parties, the 1904 Agreement, the 1910 Convention, the 1921 Convention and the 1933 Convention, together with their amending protocols. The Final Protocol provides that nothing in the Convention shall be deemed to prejudice any legislation which lays down stricter conditions than those provided by the Convention. Parties may, in other words, go further. They may not, formally, go less far, although in practice the pervasive deference to #domestic_law throughout the text gave them considerable room to do exactly that. 6. The Abolitionist Paradigm and Its Internal Tensions 6.1 Three competing goals Careful reading of the Convention reveals that it is trying to do three things which do not fit comfortably together. The first goal is repression. The state must criminalise procurement, exploitation, brothel keeping and the financing of brothels, and must cooperate internationally in prosecuting these offences. This is a #criminal_justice_approach. The second goal is deregulation. The state must dismantle registration, licensing, special identity documents and special surveillance. This looks, at first glance, like a move towards #decriminalisation of the person selling sex, and it was defended in exactly those terms. The third goal is protection of the rights of people in prostitution against the state itself. Article 6 exists because registration systems produced a caste of women subject to arbitrary police power, and the drafters wanted to end that. The difficulty is that these goals pull in different directions. If you close the brothels but do not create alternative livelihoods, the trade moves outdoors and becomes more dangerous. If you criminalise everyone around the seller of sex, including landlords, you deprive her of housing and of the ability to work in company. If you remove her registration but keep her subject to vagrancy, immigration and public morality policing, you have changed the legal form of her subordination without changing its substance. This is the analytical point at the centre of recent historical work on the negotiation of the Convention: that regulation abolitionism was caught between law enforcement centred repression, the deregulation of prostitution, and the guarantee of the human rights of people in prostitution against the state, and that the balance struck was unstable. The scholarship concludes, in blunt terms, that the Convention restructured commercial sex to the detriment of those selling it, and that it did not eliminate trafficking. 6.2 The problem of consent The exclusion of #consent is the treaty's signature doctrine and its most contested one. The case for it is straightforward and serious. Consent given under conditions of poverty, coercion, addiction, prior abuse or immigration precarity is not meaningful consent. Traffickers routinely obtain formal agreement from their victims through deception or through the manipulation of desperate circumstances. If consent were a defence, exploiters would manufacture it. Making consent irrelevant closes that escape route and it also spares the victim the ordeal of having her past, her character and her choices put on trial. The case against it is equally serious. A rule that treats the consent of every adult as legally void treats adults, in practice mostly women, as incapable of choosing. It licenses paternalistic intervention against people who do not want to be rescued. It merges the coerced and the uncoerced into a single category, which makes it harder to identify and to prioritise the people who are actually being coerced. And it hands to the state a justification for policing a population in the name of protecting it. The Palermo Protocol later attempted a compromise. Its definition makes consent irrelevant only where one of the listed means, such as force, fraud, deception or abuse of a position of vulnerability, has been used. Where no such means is present, an adult's consent remains legally significant. The Protocol also declares itself without prejudice to how states address prostitution in domestic law. The 1949 Convention makes no such concession. It is, on this point, absolutist. 6.3 The purchaser is absent A striking feature of the 1949 text, often overlooked by students, is that it says nothing about the buyer of sex. It criminalises the procurer, the exploiter, the brothel keeper, the financier and the landlord. It does not criminalise the client. This omission was deliberate. Mid twentieth century abolitionism located the wrong in exploitation by a third party, not in the transaction itself. The later development that is sometimes called the #Nordic_model, which criminalises purchase while decriminalising sale, is a distinct and later idea. It shares the moral premise of the 1949 preamble but draws a different operative conclusion from it. Students should therefore resist the common shorthand that treats the 1949 Convention as the origin of the Nordic model. It is an ancestor in spirit, not in text. 7. What the Convention Achieved It is easy, and increasingly fashionable, to dismiss the 1949 Convention. That would be a mistake. Its achievements were real. It consolidated a fragmented regime. Four separate treaties, with different scopes, ages of protection and mechanisms, were replaced by one instrument. For states, this was a significant simplification. It universalised the subject matter. The earlier treaties were, in origin, European projects. The 1949 Convention was adopted by a body with global membership and was open to accession by all states. It abandoned racialised language. The disappearance of "white slave traffic" and the adoption of #gender_neutral_language mattered. Formally, the Convention protected persons, not only women and girls. That the enforcement practice remained overwhelmingly focused on women does not diminish the textual innovation. It removed the international element. The Convention applies to procurement and exploitation whether or not a frontier is crossed. This is more expansive than the enforcement culture that grew up around the later Protocol, which in practice has often been preoccupied with cross border movement. Domestic trafficking is trafficking, and the 1949 Convention said so first. It attacked state licensed prostitution. Whatever one thinks of the wisdom of Article 6, it was a genuine intrusion into an area that states had regarded as their exclusive domain, and it contributed to the closure of licensed brothel systems in a number of countries. It refused to punish the person in prostitution. The Convention does not require, and its logic does not support, the criminalisation of the seller of sex. This is the ancestor of what is now called the #non_punishment_principle, a principle that has since been elaborated far more explicitly in the 2005 Council of Europe Convention and in European Union law. It recognised social causation. Article 16's requirement of educational, health, social and economic measures reflects an understanding that criminal law alone will not solve the problem. This insight was subsequently lost for several decades and has only recently been rediscovered under the heading of a #labour_approach or a development approach to trafficking. It created institutional habits. The national coordination service required by Article 14, the information exchange of Article 15 and the reporting duty of Article 21 established practices that reappear, in far more developed form, in every subsequent instrument. 8. Weaknesses, Gaps and Critiques 8.1 No definition The Convention never defines "traffic in persons." It defines offences, and it assumes that everyone knows what trafficking is. In consequence, national implementations diverged enormously, and no common legal concept emerged. The absence of a #definition_of_trafficking is the most obvious deficiency of the pre 2000 regime, and it is precisely the gap that the Palermo Protocol was designed to fill. 8.2 Sexual exploitation only The Convention deals exclusively with prostitution. It has nothing to say about #forced_labour in agriculture, construction, fishing, domestic work or manufacturing. It has nothing to say about forced marriage, debt bondage, child soldiering, forced criminality or trafficking for the removal of organs. This narrowness has consequences that run deeper than mere incompleteness. By defining trafficking as a sexual harm, the 1949 regime helped to entrench what has been called sex work exceptionalism: the assumption that trafficking is essentially a sex problem, that its victims are essentially female, and that other forms of extreme labour exploitation belong in a different and less urgent legal category. That assumption survived the Protocol's formally broad definition and has taken decades to unsettle. 8.3 Deference to domestic law The recurring formula "to the extent permitted by domestic law," together with the sweeping saving clause in Article 12, weakens nearly every obligation in the treaty. A state that ratified but did not wish to change its practice had ample textual cover. The Convention respected #sovereignty so thoroughly that it undermined its own operative force. 8.4 No monitoring The reporting duty in Article 21 was not backed by any expert body, any review procedure or any publication of findings that would have created reputational pressure. There was no equivalent of the treaty bodies later created under the human rights conventions, no equivalent of the Council of Europe's expert group and no equivalent of the peer review processes that now exist in this field. Where there is no monitor, there is no compliance data, and where there is no compliance data, there is no accountability. The absence of a #treaty_body was not an oversight; it reflected the era, since the 1949 Convention predates the entire architecture of international human rights supervision. 8.5 Victim rights are thin Modern anti trafficking instruments contain detailed provisions on identification of victims, reflection and recovery periods, residence permits, non punishment for offences committed as a consequence of being trafficked, legal aid, #compensation, protection of privacy and special measures for child victims. The 1949 Convention contains almost none of this. It offers repatriation, temporary maintenance and a right for aliens to be parties to proceedings where domestic law allows. That is all. There is no right of #victim_identification, no right of non refoulement, no right to remain, no right to redress. 8.6 The dispute clause deterred ratification Article 22's referral to the International Court of Justice, without an opt out on the face of the treaty, produced reservations from ratifying states and outright refusal from others. This is a lesson in treaty design. A compulsory jurisdiction clause in an instrument with no other enforcement mechanism buys very little and costs a great deal in participation. 8.7 The critique from sex workers' rights The most sustained modern criticism comes from scholars and organisations working within a #sex_workers_rights framework. Their argument runs as follows. The 1949 Convention treats all commercial sex as a harm to dignity. It criminalises everyone around the sex worker while claiming to protect her. In practice this deprives her of the ability to work indoors, to work with others for safety, to rent premises, to employ a driver or a manager and to report violence to police who regard her workplace as a crime scene. Article 6 removed a stigmatising registration system but left the person selling sex without any recognised legal status at all. The result is not emancipation but invisibility. This literature also argues that the abolitionist framework, especially in its later carceral forms, has diverted resources from labour rights, migrant rights and social protection towards policing and prosecution, with limited results and with documented harms to the very groups the law claims to serve. Whether one accepts this critique or not, it must be engaged with, because it is now the dominant position in a large part of the socio legal academy. 8.8 The critique from the abolitionist side It is important, for balance, to note that the abolitionist tradition has not disappeared and has its own answer. Its defenders argue that the sex workers' rights literature generalises from a relatively privileged and visible minority of independent sellers to a population that is in fact dominated by poverty, coercion, minority status, migration precarity and childhood entry into the trade. They argue that legalised and regulated markets expand demand, and that expanded demand is met disproportionately by trafficked and migrant women. They argue that dignity is a proper legal standard and that the law routinely refuses to enforce consent in other contexts, from labour standards to organ sales. The empirical evidence in this debate is genuinely contested, and students should be wary of any account, on either side, that presents it as settled. 9. Ratification, Reservations and Reception The Convention has attracted roughly eighty two states parties, along with a smaller number of signatories that never ratified. This is a modest figure by any standard, and it is dwarfed by the roughly one hundred and eighty states parties to the 2000 Trafficking Protocol. Several patterns are worth noting. Many of the states that ratified were those that had already abolished, or were in the process of abolishing, licensed brothel systems, or that had strong domestic abolitionist movements. Ratification frequently followed rather than caused domestic reform. Many states with regulated systems did not ratify, because Article 6 would have required them to dismantle regulation. Several common law states, including some of the largest, never became parties. Their reasons were mixed: objections to the ICJ clause, objections to the mutual assistance provisions, and a general reluctance to accept treaty obligations that would require alteration of criminal procedure. A number of states in Europe were repeatedly urged, including by parliamentary bodies of the Council of Europe in the decades after 1949, to ratify, and declined. The consequence is that the Convention never achieved the near universality that would have allowed it to function as a genuine global standard. Its influence has run instead through imitation and through the diffusion of its ideas into domestic law, including the law of states that never joined it. 10. From 1949 to the Present Framework 10.1 Slavery and forced labour: a parallel track The 1926 Slavery Convention and the 1956 Supplementary Convention on the Abolition of Slavery, the Slave Trade, and Institutions and Practices Similar to Slavery addressed servile status, debt bondage, serfdom and certain practices involving women and children. The International Labour Organization's Forced Labour Convention of 1930, its Abolition of Forced Labour Convention of 1957 and its Worst Forms of Child Labour Convention of 1999 addressed compelled work. The 2014 #Protocol_of_2014 to the Forced Labour Convention modernised that regime, adding obligations of prevention, protection and access to remedies including compensation. For fifty years these tracks ran in parallel with the 1949 Convention and rarely met. The sexual track and the labour track had different treaties, different institutions, different constituencies and different vocabularies. Their eventual convergence, under the umbrella concepts of trafficking and #modern_slavery, is one of the defining developments of the last quarter century, and it remains incomplete and contested. 10.2 The human rights treaties Article 6 of the 1979 Convention on the Elimination of All Forms of Discrimination against Women obliges states to take all appropriate measures, including legislation, to suppress all forms of traffic in women and exploitation of prostitution of women. It is a compact provision, and notably it does not resolve the abolition versus regulation dispute; it simply requires suppression of exploitation and of trafficking. Its supervisory committee, however, gave the subject a monitoring home that the 1949 Convention never had. Article 35 of the 1989 Convention on the Rights of the Child requires states to take all appropriate measures to prevent the abduction, sale and traffic in children for any purpose and in any form, and the Optional Protocol on the sale of children, child prostitution and child pornography, adopted in 2000, elaborates this in detail. #child_rights instruments thus achieved, for minors, a comprehensiveness that the adult framework lacked. 10.3 The Palermo Protocol The Protocol to Prevent, Suppress and Punish Trafficking in Persons, Especially Women and Children, supplementing the United Nations Convention against Transnational Organized Crime, was adopted in 2000 and entered into force in 2003. It changed the field. Its Article 3 definition has three elements: an act, such as recruitment, transportation, transfer, harbouring or receipt of persons; a means, such as threat or use of force, coercion, abduction, fraud, deception, abuse of power or of a position of vulnerability, or the giving or receiving of payments or benefits to achieve the consent of a person having control over another; and a purpose of exploitation, which includes at a minimum the exploitation of the prostitution of others or other forms of sexual exploitation, forced labour or services, slavery or practices similar to slavery, servitude or the removal of organs. Where the victim is a child, the means element is not required. Four differences from 1949 are decisive. The Protocol supplies a definition. The 1949 Convention did not. The Protocol covers a wide range of exploitative purposes, not only sexual ones. The Protocol restores a role for consent in adult cases, by making irrelevance of consent conditional on the use of one of the listed means, and it deliberately leaves the regulation of prostitution to domestic law. The travaux preparatoires record this compromise explicitly. Abolitionist organisations regarded this as a retreat from 1949; sex workers' rights organisations regarded it as an overdue correction. The Protocol sits inside a criminal justice instrument on #transnational_organised_crime, which shaped its priorities. Its prevention and protection provisions are largely hortatory, expressed as things states shall consider or shall endeavour to do, whereas its criminalisation provisions are mandatory. This asymmetry is the most common criticism of the Protocol in the literature, and the parallel with the criticism of the 1949 text is instructive: both instruments are firmer about punishing offenders than about assisting victims. 10.4 Regional developments The 2005 Council of Europe Convention on Action against Trafficking in Human Beings, which entered into force in 2008, is the instrument that most fully addressed the protection deficit. It introduced a recovery and reflection period, provided for renewable residence permits, articulated a non punishment provision for victims compelled to commit offences, addressed compensation, and created an independent monitoring body, the Group of Experts on Action against Trafficking in Human Beings, together with a Committee of the Parties. In doing so it supplied precisely what the 1949 Convention lacked: an institution. In the European Union, Directive 2011/36/EU established a common definition, minimum penalties, victim assistance obligations and provisions on prevention. It was substantially amended by #Directive_2024_1712, which expanded the list of exploitative purposes to include forced marriage, illegal adoption and the exploitation of surrogacy, addressed offences committed through information and communication technologies, strengthened provisions on the liability of legal persons and addressed the knowing use of services exacted from victims. Recent legal analysis of the 2024 revision finds that it improves protection and expands the criminal law, while leaving significant gaps, particularly in relation to victim identification and access to assistance for undocumented persons. Other regional instruments, including those adopted within the Association of Southeast Asian Nations, the Organization of American States, the African Union and the South Asian Association for Regional Cooperation, have produced their own conventions and action plans, with varying degrees of institutionalisation. 10.5 The development turn Target 5.2 of the Sustainable Development Goals addresses the elimination of all forms of violence against women and girls, including trafficking and sexual exploitation. Target #SDG_8_7 calls for immediate and effective measures to eradicate forced labour, end modern slavery and human trafficking, and secure the prohibition and elimination of the worst forms of child labour. Target 16.2 addresses trafficking of children. This turn has reframed trafficking as a development and labour market problem, not merely a criminal one, and it has mobilised significant institutional resources. It has also attracted its own critique. Scholars have questioned whether anti trafficking funding actually reaches trafficked people or at risk groups, and have documented cases where it has done harm, for example by financing raid and rescue operations that displace or detain the people they claim to save. Careful #evidence_based_policy in this area requires taking that literature seriously. 11. Is the 1949 Convention Still Relevant? 11.1 It remains in force The first point is often misunderstood by students. The 1949 Convention has not been terminated, denounced en masse, or superseded in a formal legal sense by the Palermo Protocol. The two instruments coexist. A state that is party to both is bound by both. The Protocol contains no clause abrogating the 1949 Convention, and the 1949 Convention terminates only the four earlier instruments listed in its Article 28. For the roughly eighty two states parties, therefore, the obligations of Articles 1, 2 and 6 continue to bind. That includes the obligation to abolish special registration of persons in prostitution, which is directly relevant to any party that might contemplate introducing a licensing scheme. 11.2 It is nevertheless largely dormant The second point is equally important. In practice the Convention is rarely litigated, rarely invoked and rarely monitored. Its reporting obligation is not systematically enforced. No case has been brought under its dispute settlement clause. National courts occasionally cite it, usually as supporting authority rather than as operative law. Where a state is party to both instruments, its practice and its national legislation are almost invariably framed by the Protocol, not by the 1949 text. The Convention therefore occupies an unusual position: formally binding, practically inert. 11.3 Its ideas are alive The third point is the one that matters most. The Convention's ideas circulate far beyond its membership. The proposition that a person cannot validly consent to being exploited animates contemporary debates about the regulation of commercial sex, about labour standards, about surrogacy and about the sale of bodily material. The proposition that the state should never license or register sellers of sex remains the core of abolitionist and Nordic model advocacy. The proposition that the person selling sex should not be punished has hardened into the #non_punishment_principle, now a firm feature of European law. The proposition that trafficking is not fundamentally about borders, which the 1949 text embraced and which later enforcement practice forgot, is being rediscovered by scholars documenting large volumes of purely domestic exploitation. And the proposition, embedded in Article 16 and Article 20, that trafficking has social and economic roots, and that #employment_agencies and recruitment intermediaries require supervision, has become one of the most active areas of contemporary reform. The 1949 Convention should therefore be read less as an operative legal code and more as a founding statement of a normative tradition, one that is still contested and still politically powerful. 12. Contemporary Challenges Through a 1949 Lens 12.1 Digital recruitment and online exploitation The drafters imagined a procurer at a railway station. Today, recruitment happens on social media, in gaming platforms, on job boards and in encrypted messaging applications. #digital_recruitment allows an exploiter to reach thousands of potential victims at negligible cost, to build trust over months and to move a person without ever meeting them. #online_exploitation of sexual material, including material produced under coercion and distributed for profit, is now a significant category of harm that the 1949 text simply cannot see. Article 2's concept of a physical brothel, and of a building let or rented for prostitution, maps poorly onto a platform, a server or a payment processor. The 2024 European Union directive's engagement with information and communication technologies shows one way forward, but the underlying conceptual problem, namely how to attribute liability across a distributed technological chain, is unresolved. 12.2 Labour exploitation and #supply_chains The most numerous victims of trafficking worldwide are not in commercial sex. They are in agriculture, construction, fishing, domestic work, manufacturing, mining and services. This is the single largest blind spot of the 1949 framework and the strongest argument that it cannot serve as a modern instrument. Contemporary reform has moved towards supply chain transparency legislation, import bans on goods made with forced labour, and due diligence duties for corporations. Critical scholarship warns, however, that such measures can become symbolic, allowing corporations and states to perform virtue while leaving the underlying economic structures untouched. 12.3 Migration governance Restrictive #migration_governance is now widely recognised as a driver of trafficking rather than a remedy for it. When lawful routes close, people use unlawful ones, and unlawful routes require intermediaries, and intermediaries create debt, and debt creates leverage. The 1949 Convention's Article 17 and Article 19, with their focus on border checks, statements from aliens and repatriation, sit squarely within a border control paradigm. That paradigm has been extensively criticised as producing the very vulnerability it claims to address. 12.4 #armed_conflict and #climate_displacement Conflict destroys the institutions that protect people and creates enormous displaced populations with no legal status and no income. Climate related displacement is producing similar effects at growing scale. Both generate precisely the conditions in which trafficking flourishes, and neither is contemplated anywhere in the 1949 text. 12.5 #data_gaps Reliable measurement remains extraordinarily difficult. Detected cases are a function of enforcement effort as much as of underlying prevalence. Global estimates rely on modelling and extrapolation and carry wide uncertainty. Research methodology in this field has advanced considerably, including through respondent driven sampling and multiple systems estimation, but scholars reviewing the field two decades after the Protocol have concluded that the evidence base remains far weaker than the confidence of political rhetoric would suggest. This is a caution that applies with equal force to any claim about what the 1949 Convention did or did not accomplish. 13. Reform Pathways and Research Directions 13.1 Should the 1949 Convention be revised, replaced or retired? Three options present themselves. Formal retirement. States parties could agree to terminate the Convention on the ground that it has been overtaken. The advantage is tidiness. The disadvantage is that termination would send a political signal, namely that the international community has abandoned the abolitionist position, that many states are unwilling to send and that others would seize upon. Termination is therefore politically improbable. Revision. A protocol could modernise the text, adding a definition, extending the scope beyond sexual exploitation, adding victim rights and creating a monitoring body. This is theoretically attractive and practically unlikely. Reopening the text would reopen the underlying dispute about prostitution, which no state wants to litigate at the multilateral level, and any resulting instrument would duplicate the Protocol and the regional conventions. Benign coexistence. The most likely and probably the most sensible course is the current one: leave the Convention in force, treat it as a historical and interpretive resource, and concentrate reform effort on the instruments that actually operate. This is the position that international practice has already adopted, whether or not anyone has articulated it. 13.2 Lessons for treaty design The 1949 Convention teaches several lessons that generalise well beyond its subject matter. Define your terms. An instrument without a definition of its central concept cannot generate consistent practice. Build an institution. Substantive obligations without a monitoring body produce reports that nobody reads and obligations that nobody enforces. The contrast between the 1949 Convention and the Council of Europe Convention on this point is stark and instructive. Be careful with dispute clauses. Compulsory judicial settlement, in an instrument that is otherwise unenforceable, deters ratification without delivering compliance. Avoid excessive deference. The formula "to the extent permitted by domestic law," repeated often enough, can hollow out a treaty from the inside. Match your scope to your problem. A treaty that addresses one form of a phenomenon, and ignores the rest, will entrench a partial understanding of that phenomenon for generations. 13.3 Research directions Several avenues remain underexplored and would make excellent subjects for student research. The domestic implementation history of the Convention, especially in states of the global South, has attracted little attention. Most existing archival work focuses on European actors and on United Nations organs. The interaction between Article 6 and colonial administration deserves study. Licensed prostitution was a feature of colonial governance in many territories, and how, when and whether the abolition requirement reached those territories is a rich and largely unwritten story, closely linked to the #colonial_clause dispute of 1949. The relationship between the 1949 obligations and modern licensing schemes in states parties is a live doctrinal question. If a state party legalises and registers sex work, is it in breach of Article 6? The question has, remarkably, never been authoritatively answered. The reception of the Convention's consent doctrine in national jurisprudence, and its migration into fields far removed from commercial sex, would repay comparative study. 14. Limitations of This Study Three limitations should be acknowledged. This is a doctrinal and historical study. It does not present original empirical data on prevalence, prosecution rates or victim outcomes, and its claims about the practical effects of the Convention rest on the secondary literature. The archival scholarship on the drafting, while excellent, is still concentrated on a limited set of actors and archives, principally European and United Nations sources. Conclusions about the intentions of the drafters must therefore be read with appropriate caution. The normative dispute at the centre of the subject, whether commercial sex between consenting adults is inherently a violation of dignity, is not resolved here and cannot be resolved by legal analysis alone. The article has attempted to present both positions fairly, but readers should be aware that any account of this field is written from somewhere. 15. Conclusion The Convention for the Suppression of the Traffic in Persons and of the Exploitation of the Prostitution of Others was the first global agreement to confront trafficking and forced prostitution under the authority of a universal organisation. It gathered up half a century of fragmented treaty making, discarded a racialised vocabulary, extended its reach to purely domestic exploitation, and made a bold and controversial choice: to treat the consent of the exploited person as legally irrelevant and to demand the dismantling of the systems through which states had licensed and policed commercial sex. Its failures were, in retrospect, largely structural. It defined no crime called trafficking. It saw exploitation only in sexual form. It deferred so heavily to national law that its obligations became porous. It created no institution capable of asking a state what it had actually done. It attracted a modest membership and then fell into a long dormancy from which it has never fully emerged. And yet it is still there. It still binds its parties. And its ideas, above all the idea that dignity sets a limit on what a person may consent to and on what a state may license, remain at the centre of one of the most bitter and most consequential arguments in contemporary law and policy. The 1949 Convention did not end trafficking. It did something more modest and, in the long run, more durable. It established a way of talking about the problem, and every subsequent instrument, including the ones that reject its premises, has had to answer it. For students entering this field, that is the lesson worth carrying forward. #international_law does not only impose rules. It supplies the categories through which a problem becomes visible, and those categories, once established, are far harder to dislodge than any treaty text. Understanding the 1949 Convention is therefore not an exercise in legal archaeology. It is a necessary condition for understanding the law that governs #human_trafficking today, and for understanding why that law still cannot agree with itself. Hashtags #human_trafficking_law #suppression_of_traffic_in_persons #forced_prostitution_1949 #UN_Convention_1949 #international_criminal_law #exploitation_of_prostitution #trafficking_treaty_history #abolitionist_tradition #palermo_protocol_comparison #victim_rights_in_treaty_law #global_anti_trafficking_framework #legal_research_for_students #treaty_design_lessons #dignity_and_consent References Allain, J. (2021). Slavery in International Law: Of Human Exploitation and Trafficking. Brill Nijhoff, Leiden. Council of Europe (1950 recommendation series and subsequent parliamentary documentation on the ratification of the 1949 Convention). Parliamentary Assembly documentation, Strasbourg. Council of Europe (2005). Convention on Action against Trafficking in Human Beings, CETS No. 197, Warsaw, 16 May 2005, entered into force 1 February 2008. Dolinsek, S. (2022). Tensions of abolitionism during the negotiation of the 1949 Convention for the Suppression of the Traffic in Persons and of the Exploitation of the Prostitution of Others. European Review of History / Revue europeenne d'histoire, 29(2), 223 to 248. https://doi.org/10.1080/13507486.2022.2026893 European Parliament and Council of the European Union (2011). Directive 2011/36/EU of 5 April 2011 on preventing and combating trafficking in human beings and protecting its victims. Official Journal of the European Union, L 101. European Parliament and Council of the European Union (2024). Directive (EU) 2024/1712 of 13 June 2024 amending Directive 2011/36/EU on preventing and combating trafficking in human beings and protecting its victims. Official Journal of the European Union, L series. Fudge, J. (2025). Constructing Modern Slavery: Law, Capitalism, and Unfree Labour. Cambridge University Press, Cambridge. https://doi.org/10.1017/9781108562058 Gallagher, A. T. (2010). The International Law of Human Trafficking. Cambridge University Press, Cambridge. (Foundational doctrinal reference retained for its treatment of the pre 2000 instruments.) International Labour Organization (1930). Forced Labour Convention, No. 29, Geneva. International Labour Organization (2014). Protocol of 2014 to the Forced Labour Convention, 1930, Geneva. International Labour Organization, Walk Free and International Organization for Migration (2022). Global Estimates of Modern Slavery: Forced Labour and Forced Marriage. International Labour Office, Geneva. Kotiswaran, P. (ed.) (2017). Revisiting the Law and Governance of Trafficking, Forced Labor and Modern Slavery. Cambridge University Press, Cambridge. (Foundational edited collection retained for its analysis of definitional incoherence.) Kotiswaran, P. (2021). The sexual politics of anti trafficking discourse. Feminist Legal Studies, 29(1), 3 to 24. https://doi.org/10.1007/s10691-020-09447-x Lannier, S., Wozniak-Cole, V., and El Dekmak, N. (2025). The 2024 revision of the Anti Trafficking Directive 2011/36/EU: A critical assessment. Journal of Human Trafficking, Enslavement and Conflict-Related Sexual Violence, 6(1), 27 to 52. League of Nations (1921). International Convention for the Suppression of the Traffic in Women and Children, Geneva. League of Nations (1933). International Convention for the Suppression of the Traffic in Women of Full Age, Geneva. Sharapov, K., Hoff, S., Mendel, J., and Gerasimov, B. (2024). Editorial: Rethinking anti trafficking funding, following the money, again. Anti-Trafficking Review, 23, 1 to 9. United Nations (1949). Convention for the Suppression of the Traffic in Persons and of the Exploitation of the Prostitution of Others, approved by General Assembly resolution 317 (IV) of 2 December 1949, entered into force 25 July 1951, with Final Protocol. United Nations (1956). Supplementary Convention on the Abolition of Slavery, the Slave Trade, and Institutions and Practices Similar to Slavery, Geneva. United Nations (1979). Convention on the Elimination of All Forms of Discrimination against Women, New York. United Nations (1989). Convention on the Rights of the Child, New York. United Nations (2000). Protocol to Prevent, Suppress and Punish Trafficking in Persons, Especially Women and Children, supplementing the United Nations Convention against Transnational Organized Crime, New York. United Nations Office on Drugs and Crime (2022). Global Report on Trafficking in Persons 2022. United Nations, Vienna. United Nations Office on Drugs and Crime (2024). Global Report on Trafficking in Persons 2024. United Nations, Vienna. Zhang, S. X. (2022). Progress and challenges in human trafficking research: Two decades after the Palermo Protocol. Journal of Human Trafficking, 8(1), 4 to 12.

The 1961 Convention on the Reduction of Statelessness is the principal treaty that tells states when they must give #nationality and when they must not take it away. It was written to give practical content to the promise in Article 15 of the Universal Declaration of Human Rights that everyone has a right to a nationality. This article offers a full doctrinal and policy study of the treaty for students of international law, migration studies and human rights. It sets out the historical path that led from the interwar collapse of empires to the drafting conferences of 1959 and 1961; it explains, article by article, the rules on acquisition of nationality at birth, the narrow grounds on which nationality may be lost or withdrawn, the duty to avoid statelessness in cases of state succession, and the supervisory role assigned to the United Nations. It then examines why a treaty of such moral clarity has attracted only around eighty parties in more than sixty years, and why millions of people remain without any nationality. The analysis identifies five structural weaknesses: permissive exceptions that hollow out the core duties; the grandfathered right to deprive people of nationality preserved by Article 8(3); the absence of a definition of what it means to be "otherwise stateless"; the treaty's silence on gender discrimination and on evidentiary burdens; and the lack of any dedicated treaty body. The article argues that the 1961 Convention should now be read together with later human rights instruments, especially the Convention on the Rights of the Child and the Convention on the Elimination of All Forms of Discrimination against Women, so that its weakest provisions are interpreted in the most protective way available. It closes with a reform agenda addressed to states, courts, universities and civil society. Keywords: statelessness; nationality; the 1961 Convention; right to a nationality; deprivation of citizenship; state succession; birth registration; international human rights law 1. Introduction A person without a #citizenship is, in law, invisible. They may be born in a hospital and never receive a birth certificate. They may live in the same town for fifty years and never vote, never own registered property, never open a bank account, never obtain a passport, and never be lawfully employed. When they fall ill, they may be turned away from a public clinic. When they die, their death may go unrecorded. Statelessness is not merely a technical defect in a person's paperwork. It is the removal of the legal status through which most other rights are claimed. The #Convention_on_the_Reduction_of_Statelessness, adopted in New York on 30 August 1961 and in force since 13 December 1975, is the single most important treaty designed to stop this from happening. Its purpose can be stated in one sentence: it sets rules for the conferral and the non-withdrawal of nationality so that statelessness does not arise. Where the 1954 Convention relating to the Status of Stateless Persons asks what rights a stateless person should enjoy once they are already stateless, the 1961 Convention asks a prior and more ambitious question. It asks how the condition can be prevented from occurring at all. That difference of focus matters. Protection instruments manage a harm. Prevention instruments seek to make the harm unnecessary. The 1961 Convention belongs to the second family. It obliges contracting states to grant nationality to a #child_born_stateless on their territory who would otherwise have none. It obliges them to grant nationality to a child born abroad to their own nationals where no other nationality is available. It restricts the circumstances in which nationality may be lost through residence abroad, through marriage, through renunciation, or through naturalisation elsewhere. It forbids, subject to narrow exceptions, any #deprivation_of_nationality that would leave a person stateless. It prohibits deprivation on racial, ethnic, religious or political grounds without exception, an #absolute_prohibition. And it requires that transfers of territory between states be arranged so that nobody living in the transferred territory is left without a nationality. Underlying the whole scheme is a delicate compromise. States retain the #sovereign_competence to decide who their nationals are. That competence was affirmed by the Permanent Court of International Justice in the Tunis and Morocco Nationality Decrees advisory opinion of 1923 and codified in the 1930 Hague Convention on Certain Questions relating to the Conflict of Nationality Laws. The 1961 Convention does not take that competence away. What it does is fence it in. States may still write their own #nationality_law, but they must write it in a way that does not manufacture stateless people. Sovereignty over membership is preserved; sovereignty to produce #rightlessness is not. Sixty-five years after adoption, the record is mixed. The treaty has fewer parties than almost any comparable #human_rights_instrument. Roughly eighty states have ratified or acceded to it, which is well under half the membership of the United Nations. Meanwhile, at the end of 2024, at least 4.4 million stateless persons or persons of undetermined nationality were reported across around one hundred countries, and children made up close to half of that reported population. Most experts believe the true #global_figure is considerably higher, because stateless people are, by definition, the population that official statistics are least likely to capture. The United Nations High Commissioner for Refugees ran a ten-year #IBelong campaign from 2014 to 2024 with the stated aim of ending statelessness within the decade. The decade closed with real progress in some countries and with the problem plainly unsolved. A broader Global Alliance to End Statelessness was launched in late 2024 to carry the work forward with a wider set of partners, including organisations led by #stateless_people themselves. This article has four aims. First, to explain what the 1961 Convention actually says, in plain language, with enough legal detail to be useful to a student writing an essay, a practitioner advising a client, or an activist lobbying a parliament. Second, to place the treaty inside the wider architecture of #international_human_rights_law, which has developed substantially since 1961 and which now supplies interpretive resources the drafters could not have imagined. Third, to identify honestly where the treaty is weak, where states have exploited its flexibility, and where the scholarly literature has become sceptical of the whole prevention paradigm. Fourth, to propose a realistic agenda for reform that does not depend on the unlikely event of a new global treaty. The article proceeds as follows. Section 2 reviews the recent scholarly literature. Section 3 explains the method. Section 4 sets out the historical background. Section 5 provides a provision-by-provision analysis. Section 6 situates the treaty within human rights law and regional systems. Section 7 examines implementation and state practice. Section 8 analyses the principal weaknesses. Section 9 presents regional and thematic case studies. Section 10 engages the critical literature. Section 11 offers a reform agenda. Section 12 concludes. 2. Literature Review Statelessness_studies has grown from a marginal specialism into a recognised interdisciplinary field. The founding of a dedicated journal, the widening of the topic beyond legal doctrine into sociology, anthropology and political theory, and the emergence of scholarship written by people with lived experience of #statelessness have all changed the shape of the debate within a single decade. Four strands of recent work are especially relevant to any study of the 1961 Convention. The doctrinal strand. A body of careful legal writing continues to map the content of the #right_to_a_nationality and the obligations it generates. Von Rütte's monograph on the human right to citizenship argues that the right to a nationality has matured from an aspirational declaration into an enforceable individual entitlement, situating it within both universal and regional #human_rights instruments and drawing on a wide body of case law. Vlieks provides a close treatment of European law on preventing and solving statelessness, showing how European Union law, the European Convention on Human Rights and the 1997 European Convention on Nationality interlock with, and sometimes exceed, the 1961 standards. Anniversary commentaries published in 2022 for the treaty's sixtieth year, including Foster's overview of its history and continuing relevance, Mezmur's analysis of its relationship with the Convention on the Rights of the Child, and Robertson's practical proposals for implementing Article 1(1) in Australia, remain among the most useful short introductions available. The critical strand. A second literature is deeply sceptical of the way statelessness has been framed. Kochenov argues that the dominant paradigm, which he calls citizenism, treats the possession of any nationality as automatically good and the absence of nationality as automatically bad, and that this framing conceals the far larger injustice of a world in which citizenships are radically unequal and racially stratified. On this view, pushing a stateless person into a low-value nationality may do very little for their rights while allowing the international community to record a success. The critique builds on earlier work by Bloom and Kingston, whose edited volume relocates the "problem" from statelessness to citizenship itself, and who describe the #weaponisation_of_citizenship as a tool of punishment, erasure and social control. This literature does not argue that the 1961 Convention is worthless. It argues that measuring progress by counting reductions in the stateless population, without asking what quality of membership people are being given, mistakes a statistic for a solution. The state-behaviour strand. Jain's study of the #manufacture_of_statelessness is a landmark. It shows that states increasingly produce statelessness not through openly discriminatory statutes, which are now easy to challenge, but through apparently neutral criteria: rules about time, rules about territory, and above all administrative practice. A #documentary_requirement that a minority cannot in practice satisfy, a registration deadline set too short, an archive conveniently lost, a decision that is never formally refused but never granted either. These techniques evade the classic legal tests because they are facially neutral. Manby's work on what she calls Schrödinger's citizenship, and Szigeti's parallel treatment, examine a related phenomenon: the person whose nationality is neither confirmed nor denied, who is treated as a national of another state for the purpose of removal and as a foreigner for the purpose of rights, and who therefore falls outside both the 1954 definition and the 1961 safeguards. Cheong's field research in Malaysia demonstrates how the notion of "potential #citizenship elsewhere" can be used to deny recognition to people who are stateless in every practical sense. The thematic strand. Specific drivers of statelessness have attracted focused study. Gender discrimination in #nationality_laws remains a leading cause of childhood statelessness; roughly two dozen states still deny women the right to confer nationality on their children on equal terms with men, and around fifty retain some gender-discriminatory provision. United Nations human rights mechanisms have addressed this repeatedly, and the Special Rapporteur on violence against women and girls reported to the General Assembly in 2023 that sex-based discrimination in nationality laws and statelessness reinforce one another. The use of #citizenship_stripping as a #counter_terrorism measure has generated a large literature, much of it critical, examining the deprivation of nationality from returning foreign fighters and the practice's dubious security value; Paulussen sets out the human rights and humanitarian law constraints, and Prener's monograph on denationalisation traces the legal, political and moral implications of the trend. Balaton-Chrimes documents how ethnic registration practices in Kenya produce statelessness at the administrative counter rather than in the statute book. Van Thillo analyses the protection gaps exposed by displacement from Ukraine. Sardelić examines the civic marginalisation of Romani minorities in Europe, a population historically at high risk of statelessness after the dissolution of Yugoslavia. Taken together, this literature suggests a field in transition. The doctrinal question of what the 1961 Convention requires is comparatively settled. The harder questions now are empirical and political: why states do not accede, how they evade the obligations they have accepted, and whether the prevention paradigm is asking the right question in the first place. This article attempts to hold all three questions in view. 3. Methodology and Scope This is a #doctrinal_analysis and analytical study, supported by secondary literature and by publicly reported policy developments. It uses three methods. The first is #textual_analysis of the Convention itself. Each substantive article is read closely, with attention to the difference between mandatory language ("shall grant") and permissive language ("may provide"), and with attention to the conditions and exceptions attached to each duty. The drafting history, including the two sessions of the United Nations Conference on the Elimination or Reduction of Future Statelessness held in 1959 and 1961, is used to explain why particular compromises were reached. The second is #systemic_interpretation. Under the general rule of treaty interpretation reflected in the Vienna Convention on the Law of Treaties, a treaty must be read in its context and in the light of relevant rules of international law applicable between the parties. This means that the 1961 Convention cannot responsibly be read in isolation from the International Covenant on Civil and Political Rights, the #Convention_on_the_Rights_of_the_Child, the Convention on the Elimination of All Forms of Discrimination against Women, the International Convention on the Elimination of All Forms of Racial Discrimination, and the relevant regional treaties. Where the 1961 text is ambiguous or permissive, later and more protective standards should inform its interpretation. The third is #comparative_assessment and critical evaluation. State practice, national litigation and the recent scholarly critique are used to test whether the treaty is achieving what it set out to achieve. Two limits should be stated openly. First, the article does not attempt an exhaustive survey of every state party's implementing legislation; that would require a book. It uses illustrative examples. Second, statistics on statelessness are unreliable by their nature. Reported figures capture the people whom governments and international agencies have managed to count, and the whole condition of statelessness tends to place people beyond the reach of counting. Numbers in this article should therefore be treated as indicative rather than exact. 4. Historical Background: How the Convention Came About 4.1 The interwar rupture #Mass_statelessness became a European problem before it became an international legal category. The collapse of the Ottoman, Austro-Hungarian, Russian and German imperial systems after the First World War left millions of people outside the membership of any of the successor states. Revolutionary governments denationalised political opponents in large numbers. The interwar period produced the #Nansen_passport, an improvised travel document issued to stateless refugees, and produced the first serious attempt to codify the problem in the 1930 Hague Conference, which yielded the Convention on Certain Questions relating to the Conflict of Nationality Laws and a Protocol dealing with a specific case of statelessness at birth. Those instruments were modest. They accepted that each state decides who its nationals are, and they tried, at the margins, to reduce the conflicts of law that leave people in the gaps between nationality regimes. They did not create any right to a nationality. 4.2 The postwar reconstruction of the idea The #Second_World_War transformed the moral register of the debate. The mass denationalisations that preceded deportation and extermination demonstrated that the removal of legal membership was not a technicality but a precondition of atrocity. This is the context in which Hannah Arendt developed her account of the stateless as those who have lost the right to have rights, and in which Article 15 of the #Universal_Declaration_of_Human_Rights was drafted: everyone has the right to a nationality, and no one shall be arbitrarily deprived of their nationality or denied the right to change it. Article 15 was a declaration, not a treaty. The task of turning it into binding obligations fell to the United Nations. In 1949 the General Assembly placed nationality, including statelessness, on the International Law Commission's list of topics for codification. The Economic and Social Council gave the item priority. The 1951 Refugee Convention had been intended to cover both refugees and stateless persons but ultimately covered only refugees, leaving a visible hole. 4.3 Two treaties, two functions The hole was filled in two stages. The 1954 Convention relating to the Status of Stateless Persons supplied the definition that still governs the field: a stateless person is someone who is not considered as a national by any state under the operation of its law. It also set out a #protection_regime for such people, modelled closely on the Refugee Convention. The International Law Commission, meanwhile, produced two alternative drafts in 1953: a Draft Convention on the Elimination of Future Statelessness, which was ambitious, and a Draft Convention on the Reduction of Future Statelessness, which was cautious. The General Assembly convened a conference of plenipotentiaries in 1959. It could not reach agreement and was reconvened in 1961. The word that survived in the final title tells the whole story. States chose #reduction over elimination. They were prepared to narrow the pathways into statelessness. They were not prepared to close them. That choice explains many of the treaty's compromises. Nearly every core duty is followed by a list of conditions a state may impose or exceptions it may invoke. Article 8, which prohibits deprivation of nationality that would cause statelessness, contains a clause allowing states, at the moment of signature or accession, to preserve grounds for deprivation that already existed in their domestic law. The Convention entered into force only in 1975, fourteen years after adoption, once the required six ratifications had accumulated. UNHCR was subsequently designated by the General Assembly as the body to which a person claiming the benefit of the Convention may turn, giving the agency a supervisory mandate under Article 11. 5. The Architecture of the Convention The treaty has twenty-one articles. Substantively it divides into four blocks: acquisition of nationality (Articles 1 to 4); loss and deprivation of nationality (Articles 5 to 9); state succession (Article 10); and institutional and final provisions (Articles 11 to 21). 5.1 Acquisition of nationality: Articles 1 to 4 #Article_1: nationality by birth on the territory. This is the heart of the treaty. A contracting state shall grant its nationality to a person born on its territory who would otherwise be stateless. The obligation is unconditional in its object, but the state may choose how to fulfil it. It may grant nationality automatically at birth by operation of law, or it may grant it on application. If it chooses the application route, it may impose conditions, but only those listed exhaustively in Article 1(2): a period within which the application must be lodged, beginning not later than at age eighteen and ending not earlier than at age twenty-one; habitual residence for a period the state specifies, not exceeding ten years in total and not exceeding five years immediately preceding the application; the absence of certain criminal convictions of a specified gravity; and the condition that the person has always been stateless. Crucially, an application that meets the prescribed conditions may not be rejected. The state has no residual discretion to refuse. The design deserves attention. It creates a conditional #jus_soli, a right to the nationality of the place of birth that activates only where no other nationality is available. It does not require unconditional birthright citizenship of the kind found in much of the Americas. It is a #safety_net, not a general rule of attribution. Article 1 also addresses the child of a national who cannot obtain nationality because of the parent's status: where a person would be stateless but for the operation of a state's rules on foreign-born descent, the state may extend nationality by descent subject to conditions similar to those above. #Article_2: foundlings. A #foundling found on the territory of a contracting state shall, in the absence of proof to the contrary, be considered to have been born within that territory to parents possessing the nationality of that state. This is a presumption, not an irrebuttable rule, and its scope has been contested. Some states apply it only to newborn infants; the more protective reading, endorsed by UNHCR guidance, extends it to any child found in the territory while still of an age at which they cannot establish their own parentage or place of birth. #Article_3: births on ships and aircraft. A birth on a ship is deemed to have taken place in the territory of the state whose flag the ship flies; a birth on an aircraft, in the territory of the state of registration. This is a small provision that closes a real gap. #Article_4: nationality by descent for those born abroad. Where a person would otherwise be stateless and is not born in the territory of a contracting state, that state shall grant its nationality if one of the parents held its nationality at the time of the person's birth. Again the state may choose between automatic conferral and conferral on application, subject to closed conditions: an application period beginning no later than the person's twenty-third birthday, habitual residence not exceeding three years, continuous statelessness, and the absence of another nationality at birth. Read together, Articles 1 to 4 construct an interlocking net. If the state of birth applies #jus_sanguinis only and the parents are stateless, Article 1 catches the child. If the child is born abroad to a national of a contracting state and cannot inherit that nationality under normal rules, Article 4 catches the child. If the child's parentage and birthplace are unknown, Article 2 catches the child. If the child is born in transit, Article 3 catches the child. The net has holes, as we shall see, but the intention behind its design is coherent. 5.2 Loss and deprivation of nationality: Articles 5 to 9 The second block moves from conferral to non-withdrawal. The organising principle is that a person should not fall out of #nationality into nothing. #Article_5: change of personal status. If the law of a contracting state provides for loss of nationality as a consequence of marriage, divorce, legitimation, recognition or adoption, that loss must be conditional upon the possession or acquisition of another nationality. This was a significant provision in 1961, when many legal systems automatically stripped a woman of her nationality upon marriage to a foreigner. The article also protects children legitimated by a foreign father, requiring that they retain nationality unless they acquire another. #Article_6: derivative loss. If loss of nationality by a spouse or a parent entails loss for the other spouse or for children, that consequential loss must be conditional on possession or acquisition of another nationality. In other words, statelessness may not be inherited as a side effect. #Article_7: renunciation, departure and residence abroad. Renunciation of nationality shall not result in loss unless the person possesses or acquires another nationality. Naturalisation abroad may trigger loss of the original nationality only where the new nationality has actually been obtained or assured. A national of a contracting state may not lose nationality on grounds of departure, residence abroad or failure to register, unless he or she thereby becomes stateless is expressly avoided. There are exceptions: a naturalised person who resides abroad for at least seven consecutive years may lose nationality if he or she fails to declare an intention to retain it; and a person born outside the territory may be required, after majority, to reside in the territory or register with the authorities in order to retain the nationality. Article 7(6) then supplies a residual guarantee: except in the circumstances the article itself specifies, a person shall not lose the nationality of a contracting state if such loss would render him stateless, even where the loss is not expressly prohibited elsewhere in the treaty. This is an important sweep-up clause. #Article_8: deprivation. A contracting state shall not deprive a person of nationality if the deprivation would render that person stateless. This is the flagship prohibition. It is then qualified. Deprivation is permitted where nationality was obtained by #misrepresentation or fraud, and in the residence-abroad circumstances allowed under Article 7. And then comes Article 8(3), the most criticised clause in the treaty. At the time of signature, ratification or accession, a state may declare that it retains the right to deprive a person of nationality, even where statelessness results, on grounds already existing in its national law at that moment. Those grounds are limited to two families: conduct seriously prejudicial to the vital interests of the state, including service to or receipt of emoluments from a foreign state contrary to an express prohibition and behaviour inconsistent with the duty of loyalty; and taking an oath or making a formal declaration of allegiance to another state, or giving definite evidence of a determination to repudiate allegiance. Article 8(4) provides a procedural safeguard: any exercise of such a power must be in accordance with law, and the person must have the right to a fair hearing before a court or other independent body. #Article_9: the absolute prohibition. A contracting state may not deprive any person or group of persons of their nationality on racial, ethnic, religious or political grounds. This is the one rule in the treaty with no exception, no reservation and no escape clause. It is the crystallised memory of the twentieth century. It is also the provision most directly relevant to some of the largest #protracted_statelessness situations in the world today, in which entire communities have been excluded from nationality on ethnic or religious grounds. 5.3 State succession: Article 10 Where a treaty provides for the #transfer_of_territory, it must include provisions to ensure that no person becomes stateless as a result. Where no such provisions are made, the state acquiring the territory shall confer its nationality on persons who would otherwise become stateless. The provision was drafted with decolonisation and boundary adjustment in mind. It was tested severely by the dissolution of the Soviet Union, Yugoslavia and Czechoslovakia in the 1990s, events that produced hundreds of thousands of stateless people, and it remains directly relevant wherever borders move or new states emerge. 5.4 Institutional provisions: Articles 11 to 21 #Article_11 required the establishment, within the United Nations framework, of a body to which a person claiming the benefit of the Convention may apply for examination of their claim and for assistance in presenting it to the appropriate authority. The General Assembly assigned this function to UNHCR in the 1970s. The mandate has since expanded through successive resolutions into a general responsibility for the identification, prevention and reduction of statelessness and for the protection of stateless persons. Article 12 deals with retroactivity, providing that where a state grants nationality at birth by operation of law under Article 1(1) or Article 4, those provisions apply to persons born before as well as after entry into force. #Article_13 is a savings clause: nothing in the Convention displaces more generous provisions in national law or in other treaties. This matters, because it means the 1961 standards are a floor and not a ceiling. Article 14 provides for referral of interstate disputes to the International Court of Justice. #Article_17 restricts reservations to Articles 11, 14 and 15. In other words, the substantive obligations in Articles 1 to 10 cannot be the subject of reservations. This is a striking feature. A state that accedes accepts the core rules in full, with the single, structurally different exception of the Article 8(3) declaration, which is not technically a reservation but a permitted retention of pre-existing powers, exercisable only at the moment of joining. 6. The Convention in the Wider Human Rights System The 1961 Convention was drafted before most of the modern #human_rights_treaties existed. Reading it today without reference to what came after would be a serious error of method. 6.1 The universal instruments The International Covenant on Civil and Political Rights (1966), in Article 24(3), provides that every child has the right to acquire a nationality. The Human Rights Committee has consistently read this as imposing a duty on the state of birth to grant nationality where the child would otherwise be stateless, and has found violations in individual communications where states failed to do so. The Convention on the Rights of the Child (1989), in Articles 7 and 8, is the most powerful ally the 1961 Convention has. Article 7 gives every child the right to be registered immediately after birth, the right to a name, and the right to acquire a nationality, and requires states to ensure the implementation of these rights in particular where the child would otherwise be stateless. Article 8 protects the child's identity, including nationality. The CRC has near-universal ratification. This is decisive. A state that has not joined the 1961 Convention but has ratified the CRC, which is almost every state on earth, has already accepted an obligation to prevent childhood statelessness. Article 3's principle of the best interests of the child supplies an interpretive lens under which any ambiguity in nationality law should be resolved in favour of the child acquiring a nationality. As Mezmur argues, the CRC effectively universalises the core commitment of the 1961 Convention, and reading the two instruments together gives states far less room to claim that they are unbound. The Convention on the Elimination of All Forms of Discrimination against Women (1979), in Article 9, requires equal rights for women with respect to the nationality of their children and prohibits automatic changes to a woman's nationality on marriage. The 1961 Convention says almost nothing about #gender_discrimination, which is one of its most glaring silences. CEDAW fills the gap directly, and the CEDAW Committee's general recommendations on the gender-related dimensions of nationality and statelessness have developed the point considerably. The International Convention on the Elimination of All Forms of Racial Discrimination (1965) prohibits racial discrimination in the enjoyment of the right to nationality and reinforces the absolute rule in Article 9 of the 1961 Convention. The Convention on the Rights of Persons with Disabilities (2006), in Article 18, restates the right to a nationality for persons with disabilities and prohibits deprivation of nationality on the basis of disability, a form of exclusion that had long gone unremarked. 6.2 The regional instruments Europe. The 1997 #European_Convention_on_Nationality is in several respects more advanced than the 1961 Convention. It states a general principle that statelessness shall be avoided, requires facilitated naturalisation for stateless persons, and prohibits deprivation resulting in statelessness except in cases of fraudulent acquisition. The Council of Europe's 2006 Convention on the Avoidance of Statelessness in relation to State Succession develops Article 10 of the 1961 Convention into a detailed regime. Within the European Union, the Court of Justice has held in a line of cases beginning with Rottmann that where the withdrawal of national citizenship also removes European Union citizenship, the decision falls within the scope of EU law and must satisfy a proportionality test. Meanwhile the European Court of Human Rights, although the European Convention on Human Rights contains no right to nationality, has developed a protective jurisprudence under Article 8, treating arbitrary denial or revocation of nationality as capable of interfering with private life. Africa. The #African_Charter on the Rights and Welfare of the Child, in Article 6, guarantees every child the right to acquire a nationality and requires states to grant nationality to a child born on their territory who would otherwise be stateless. The African Committee of Experts has applied this provision robustly, most notably in the Nubian children case concerning Kenya. The African Court on Human and Peoples' Rights, in Anudo v Tanzania, held that arbitrary deprivation of nationality violated the African Charter and international law, and it is widely read as treating the right to a nationality as customary. A draft protocol on the right to a nationality and the eradication of statelessness in Africa has been under development for some years and would, if adopted, be the most comprehensive regional instrument in the world. The Americas. Article 20 of the #American_Convention_on_Human_Rights is unusually strong: everyone has the right to a nationality; everyone has the right to the nationality of the state in whose territory they were born if they do not have the right to any other nationality; and no one shall be arbitrarily deprived of nationality or of the right to change it. The Inter-American Court has applied this vigorously, notably in Yean and Bosico v Dominican Republic, which concerned the denial of birth certificates to children of Haitian descent, and in Expelled Dominicans and Haitians v Dominican Republic. The region also has a strong tradition of unconditional jus soli, which by itself prevents most statelessness at birth. Asia and the Middle East. These are the regions with the weakest regional frameworks and, not coincidentally, some of the largest stateless populations. There is no binding regional human rights treaty on nationality in Asia. The Arab Charter on Human Rights includes a right to nationality but its enforcement machinery is weak. Immanuel has argued that, notwithstanding low treaty ratification, a customary obligation to avoid, reduce or prevent statelessness can be identified in South Asian practice, which if accepted would bind states regardless of accession. 6.3 Customary international law The most consequential legal question in this area is whether the prohibition of arbitrary deprivation of nationality, and the obligation to avoid statelessness at birth, have hardened into #customary_international_law. If they have, then the low ratification rate of the 1961 Convention matters far less than it appears to, because the substance of the obligation would bind non-parties too. The evidence is suggestive but not conclusive. Article 15 of the Universal Declaration is widely regarded as reflecting custom. The near-universal ratification of the CRC establishes very broad acceptance of the child's right to acquire a nationality. Judicial statements, including from the African Court, have described the right to a nationality in customary terms. Von Rütte's monograph makes a sustained case for a human right to citizenship grounded in this convergence. Against that, state practice remains inconsistent, and states continue to assert wide discretion. The prudent conclusion is that the prohibition of arbitrary deprivation of nationality has the strongest claim to customary status, while the positive duty to confer nationality on otherwise stateless children born on the territory is an emerging norm whose consolidation is well advanced but not complete. 7. Implementation and State Practice 7.1 The accession problem The most obvious failure of the 1961 Convention is arithmetic. Around eighty states are parties. Of these, a meaningful minority acceded only after 2014, in response to the #IBelong campaign, which shows that sustained diplomatic pressure works. But the treaty remains far less ratified than the Refugee Convention, and entire regions are barely represented. Most states in Asia and the Middle East, which host some of the world's largest stateless populations, are not parties. Why do states stay out? Several reasons recur. The first is a fear of loss of control. Nationality is understood as the last redoubt of sovereignty, and governments are reluctant to accept external rules about who belongs. The second is misunderstanding. Officials frequently believe, wrongly, that acceding would require them to adopt unconditional jus soli or to naturalise large numbers of migrants. It would not. The obligation bites only where the person would otherwise be stateless. The third is #strategic_exclusion. In some cases the exclusion of a particular group from nationality is not an accident to be remedied but a policy to be preserved. A state that has deliberately excluded a minority from citizenship has no interest in accepting a treaty that would require it to stop. The fourth is #institutional_inertia. Statelessness has no domestic constituency. Stateless people cannot vote. There is rarely a ministry whose performance is measured by reducing statelessness, and the issue seldom reaches the top of a cabinet agenda. 7.2 What implementation looks like when it works Where states have implemented the Convention seriously, several features recur. A #statutory_safeguard in the nationality act. The most effective single reform is a clear provision granting nationality to any child born in the territory who would otherwise be stateless, operating automatically or on a simple application, with no discretionary escape hatch. Countries that have added such clauses have seen measurable declines in childhood statelessness. Universal and free #birth_registration. A safeguard is worthless if the child cannot prove where they were born. Birth registration is the evidentiary foundation of nationality. States that have removed fees, decriminalised late registration, permitted registration regardless of parents' immigration status, and taken mobile registration units to remote or marginalised communities have made far more progress than states that have merely amended the statute book. This is where the 1961 Convention meets Sustainable Development Goal target 16.9 on legal identity for all. A #statelessness_determination_procedure. Although this is formally an obligation flowing from the 1954 Convention rather than the 1961, in practice the two are inseparable. A state cannot fulfil its Article 1 duty unless it can determine whether a person would "otherwise be stateless". A dedicated, accessible, evidentially realistic procedure with a shared burden of proof and a right of appeal is essential infrastructure. Removal of #gender_discrimination. States that have equalised the right of mothers and fathers to confer nationality on their children have eliminated one of the single largest generators of childhood statelessness at a stroke. Several countries have made this reform in recent years, often after long civil society campaigns. Confirmation campaigns and #mass_naturalisation. Where a stateless population already exists, targeted programmes to confirm or grant nationality have produced the field's biggest wins. Kyrgyzstan's resolution of its known stateless caseload is the most frequently cited example. 7.3 What implementation looks like when it fails The characteristic failures are subtler than outright non-compliance. The first is the #discretionary_override. A statute contains the correct safeguard, but the grant is expressed as something the minister "may" do rather than "shall" do. Discretion, in practice, means refusal. The second is the #evidentiary_trap. The applicant must prove a negative: that no other state considers them a national. Proving that requires responses from foreign consulates, which may simply never reply. Silence is then treated as an unresolved question rather than as evidence of non-nationality, and the applicant is left in permanent limbo. This is the Schrödinger's citizenship problem: the person is neither a national nor a non-national, and the indeterminacy is itself the harm. The third is the #procedural_cliff. Application windows that open at eighteen and close at twenty-one are compatible with the letter of Article 1(2), but a young person who does not know the law exists, cannot afford a lawyer, and has no documents will miss the window. The safeguard formally exists and is practically unreachable. The fourth is the administrative production of statelessness identified by Jain. Neutral-looking requirements, applied to a group that cannot meet them, produce exclusion without ever writing exclusion into law. 8. Structural Weaknesses of the Convention 8.1 The permissive architecture The Convention's core duties are hedged with #permitted_conditions. A state may require ten years of habitual residence. A state may exclude applicants with certain criminal convictions. A state may set application windows. Each condition was a concession extracted at the drafting conference. Cumulatively, they mean that a state can be fully compliant with Article 1 while a stateless child born on its territory waits until adulthood, remains stateless throughout childhood, and then must navigate a bureaucratic process to obtain what the treaty describes as an entitlement. This is why the interpretive move described in Section 6 matters so much. Under Article 7 of the CRC, read with the best interests principle, the delay of nationality until adulthood is very difficult to justify. UNHCR guidance recommends that states grant nationality automatically at birth, and that where an application procedure is used, it should be available from birth and should not be conditioned on lengthy residence. The 1961 Convention permits a slow route; the child rights framework strongly discourages it. 8.2 Article 8(3) and the return of denationalisation The retained right to deprive people of nationality on loyalty grounds, even at the cost of statelessness, was already an anomaly in 1961. It has become a live problem. Since 2000, a significant number of states have introduced or expanded powers to strip citizenship, with the sharpest expansion in Europe, driven by the return of people who travelled to conflict zones in Syria and Iraq. The #legal_debate is fierce. Defenders argue that deprivation is a legitimate exercise of the state's right to define its membership and to protect national security, and that where a person holds another nationality, no statelessness arises. Critics respond on several grounds. The measure is discriminatory in effect, because it can only lawfully be used against dual nationals, who in Europe are disproportionately citizens of migrant background; it therefore creates two classes of citizen, one whose membership is secure and one whose membership is conditional. It is of doubtful security value, since it does not remove the threat but relocates it, and it shifts responsibility onto other states, often states less able to manage it. It sits uneasily with the principle that a state should deal with its own nationals rather than exporting them. And where the "other" nationality is theoretical rather than actual, deprivation produces statelessness in fact while denying it in law. The high-profile British litigation over the withdrawal of citizenship from a woman who travelled to Syria as a teenager, in which the domestic courts upheld the deprivation on the basis that she was entitled to another nationality that she had never held or claimed, illustrates precisely the problem Manby and Szigeti describe. The rule against statelessness was satisfied by an attribution of nationality that the other state did not accept. Prener's monograph on denationalisation argues that this practice hollows out the meaning of citizenship for everyone, not only for those who lose it. Article 9 remains the outer limit. Deprivation on racial, ethnic, religious or political grounds is prohibited absolutely. Whether a policy that in practice affects almost exclusively citizens of one ethnic or religious background is caught by Article 9, or evades it because it is facially neutral, is one of the sharpest unresolved questions in the field. 8.3 The undefined threshold: what does "otherwise stateless" mean? The Convention repeatedly turns on the #otherwise_stateless test, that is, whether a person "would otherwise be stateless". It never says how that is to be established. The 1954 definition speaks of a person not considered a national by any state "under the operation of its law", and UNHCR guidance is clear that this requires an assessment of law as applied in practice, not merely law on the books. If a state's nationality statute would, on its face, grant a person nationality, but the competent authority in fact refuses to recognise them, that person is stateless. Many states resist this reading. They prefer to ask whether the person is theoretically eligible for another nationality somewhere. Cheong's research shows how this "potential citizenship elsewhere" logic operates to deny recognition. It allows a state to say, truthfully in a narrow legal sense, that the person is not stateless, while the person remains without any nationality that any state will actually acknowledge. The gap between formal eligibility and practical recognition is where a large proportion of the world's stateless population lives. 8.4 Silence on gender The Convention's near-total #silence_on_gender is a product of its era. Article 5 addresses loss of nationality on marriage, which was the pressing gender issue of 1961, but the treaty says nothing about the far more consequential rule that in some states only fathers can transmit nationality to children. Where a mother cannot pass on her nationality and the father is stateless, unknown, absent, or unwilling or unable to acknowledge the child, the child is left with nothing. This remains one of the largest single drivers of childhood statelessness. Around two dozen states still restrict mothers' ability to confer nationality on their children on equal terms. The remedy is not in the 1961 Convention but in CEDAW Article 9, in CRC Article 7 read with Article 2, and in domestic constitutional equality guarantees, which have been used successfully in litigation in several countries. 8.5 No treaty body, no reporting, no jurisprudence Unlike the CRC, CEDAW or the ICCPR, the 1961 Convention has no committee, no periodic reporting cycle, no individual complaints mechanism and no general comments. Article 11 gives UNHCR a supervisory and assistance role, but UNHCR is an operational agency dependent on the cooperation of the very states it monitors, not a quasi-judicial body. Article 14 permits interstate disputes to go to the International Court of Justice, and no state has ever used it. The consequence is that the Convention generates very little authoritative interpretation of its own. Its meaning is developed instead through UNHCR guidelines, through the work of other treaty bodies applying overlapping norms, and through national and regional courts. This is a serious #institutional_gap. It also explains why the practical strategy of statelessness lawyers is so often to litigate under the CRC, CEDAW or a regional human rights convention rather than under the 1961 Convention itself. 8.6 Climate change and the coming test of Article 10 Article 10 was drafted for territorial transfer between existing states, not for #climate_displacement. It was not drafted for the disappearance of territory. Yet the prospect that low-lying island states may become uninhabitable raises a question the treaty cannot answer: what happens to the nationality of the population of a state whose territory ceases to sustain human life? Most scholars now argue that the presumption of continuity of statehood should apply, that a state does not cease to exist merely because its territory becomes uninhabitable, and that its nationality law therefore continues to operate. Recent bilateral arrangements offering residence and citizenship pathways to affected populations point toward a practical solution. But this remains an area where existing law is being asked to answer a question it was never designed to address, and where the 1961 Convention offers little more than an analogy. 9. Case Studies 9.1 Protracted statelessness on ethnic and religious grounds The largest and most widely reported stateless populations are those excluded from nationality on ethnic or religious grounds. Where a national law or its administration excludes an entire community from citizenship on such grounds, Article 9 of the 1961 Convention is engaged in its clearest form. The prohibition is absolute and is not subject to any exception, reservation or national security override. That the states concerned are often not parties does not dispose of the matter, since the prohibition of racial discrimination is a peremptory norm and the prohibition of discriminatory deprivation of nationality is widely regarded as customary. The 1961 Convention's contribution here is normative rather than mechanical: it names the wrong precisely and forecloses the usual excuses. 9.2 Statelessness produced by administration: the Kenyan example Balaton-Chrimes documents how, in Kenya, the process of vetting applicants for identity documents on the basis of ethnic origin creates an administrative apparatus that can render members of certain communities effectively stateless without any change in the law. The statute grants nationality; the counter refuses to record it. This is Jain's manufactured statelessness in concrete form, and it demonstrates why any serious implementation strategy must look at registration practice and not only at legislation. The African Committee of Experts' decision in the Nubian children case established that this kind of discriminatory administrative practice violates the African Charter on the Rights and Welfare of the Child. 9.3 State succession and displacement in Europe The dissolution of Yugoslavia and of the Soviet Union produced large populations who were not automatically absorbed into any successor state's citizenry. Romani communities were disproportionately affected, as Sardelić's work on civic marginalisation shows, because they were less likely to hold the residence registrations and documents that successor states used to determine who belonged. Three decades later, residual caseloads remain. The war in Ukraine has generated a new set of problems. Van Thillo's analysis identifies persistent gaps in the protection of people who are both stateless and forcibly displaced: they may struggle to establish eligibility for temporary protection, they may lack the documents needed to cross borders, and children born in displacement may face obstacles to birth registration. The overlap between statelessness and displacement is one of the least well handled areas of the international protection system. 9.4 The Gulf and the question of undetermined nationality Long-term residents in several Gulf states have lived for generations without recognised nationality, often described by the authorities as having undetermined status rather than as being stateless. The distinction is not innocent. To call a population "undetermined" is to defer the question indefinitely and to avoid the legal consequences that recognition of statelessness would trigger. The Marambio case note on the treatment of such persons holding travel documents of convenience from third states illustrates how the attribution of a nominal foreign nationality can be used to justify removal or to withhold protection, even where that nationality confers no rights at all. 9.5 Statelessness and immigration detention Because a stateless person has no state obliged to receive them, they cannot lawfully be removed. In systems where immigration detention is justified by the prospect of removal, this can produce indefinite detention of people who have committed no crime and cannot be deported anywhere. The Australian High Court's decision in NZYQ, discussed by Gordon, held that detention cannot lawfully continue where there is no real prospect of removal in the reasonably foreseeable future. The case is a reminder that statelessness is not only a problem of nationality law but a problem of liberty. 10. Critical Perspectives: Is Prevention the Right Frame? A responsible article must engage the strongest objections to its own subject matter. 10.1 The citizenism critique Kochenov's #citizenism critique runs roughly as follows. The international community treats statelessness as a pathology and citizenship as the cure. But citizenships are not equal. The bundle of rights attached to a passport varies enormously, and that variation tracks global hierarchies of wealth and race. To insist that everyone must have a nationality, without asking what that nationality is worth, is to naturalise a system in which most of humanity is assigned, at birth, to a status that severely restricts their freedom of movement and their access to opportunity. On this account, statelessness is a symptom of an unjust global membership regime, not a deviation from a just one. Solving statelessness by conferring a nationality of little practical value may register as a success in the statistics while changing nothing in the life of the person concerned. This critique deserves to be taken seriously, and it has an obvious force. It is also, in my assessment, incomplete as a guide to action. The reason is that the two harms are not commensurable. Holding a low-value nationality is bad. Holding no nationality at all is categorically different. The stateless person is not merely at the bottom of a hierarchy of citizenships; they are outside the hierarchy altogether. They cannot be lawfully employed, cannot register a marriage, cannot inherit, cannot be issued a death certificate, and cannot invoke the diplomatic protection of any state. The citizenism critique is a powerful diagnosis of the global order. It is not an argument for leaving people without documents while we wait for that order to change. The most defensible synthesis is that the critique should discipline our metrics rather than replace our objectives. Ending statelessness should be pursued, and pursued urgently, while resisting the temptation to treat a reduction in the headline number as evidence that justice has been done. 10.2 The participation critique A second criticism concerns who produces knowledge about statelessness. Hyma and others argue that the field has been dominated by researchers and institutions in the global North writing about populations in the global South, and that stateless people themselves have been objects of study rather than participants in it. The launch of the Global Alliance to End Statelessness with formal roles for stateless-led organisations is a response to this. The methodological implication for students is direct: an essay about a stateless community that cites no work by members of that community has a gap in it. 10.3 The sovereignty rejoinder The most common objection from governments is that nationality is the core of #sovereignty. This deserves a straight answer. The 1961 Convention does not tell any state whom to admit as a national in general. It tells states not to leave people with nothing. A state remains free to set naturalisation requirements, to require language tests, to demand periods of residence, and to refuse citizenship to those who do not meet its criteria, provided those people have a nationality somewhere. Article 13, which preserves more generous national provisions, confirms that the treaty is a floor. The obligation is residual. It engages only in the gap where no other state will act. Characterising it as an assault on sovereignty misdescribes it. 11. A Reform Agenda The prospects for a new global treaty on statelessness are negligible. Any realistic agenda must work with the instruments that exist. For states that are not parties. Accede. The most common objections rest on misunderstanding. Accession does not require unconditional jus soli, does not require mass naturalisation of migrants, and does not surrender control over immigration. It requires a safety net. States that fear the treaty should read Article 13 and Article 17 together: the substantive obligations cannot be reserved, but they are narrower than most officials assume. For states that are parties. Withdraw #Article_8_3 declarations. Replace discretionary language in nationality statutes with mandatory language. Grant nationality to otherwise stateless children automatically at birth rather than on application at eighteen. Establish a statelessness determination procedure with a shared burden of proof, free legal aid and a right of appeal. Treat consular silence as evidence of non-nationality rather than as an unresolved question. For all states, party or not. Equalise the rights of women and men to confer nationality on their children. Make birth registration universal, free, immediate, and available regardless of the parents' immigration status or documentation. This single measure would do more to prevent statelessness worldwide than any other reform, and it is comparatively cheap. For courts. Read the 1961 Convention systemically. Where its text is permissive, interpret it in the light of CRC Articles 3, 7 and 8 and CEDAW Article 9. Subject deprivation decisions to genuine proportionality review rather than deferential security review. Refuse to accept a theoretical foreign nationality as an answer where the foreign state does not in fact recognise the person. For universities and students. Statelessness is under-taught. It belongs in every course on international human rights law, refugee law, migration and citizenship. Student legal clinics have proved unusually effective in this area, because the work is document-intensive, individually decisive and chronically under-resourced. For civil society and international organisations. Continue to press for accessions; the post-2014 record shows they can be obtained. Invest in the participation of stateless-led organisations. And be candid about metrics: report not only how many people acquired a nationality, but what that nationality allows them to do. 12. Conclusion The 1961 Convention on the Reduction of Statelessness is a treaty of modest ambition and enormous importance. It did not try to abolish statelessness; it tried to reduce it, and the language of the title records the limit of what states were willing to accept in 1961. Its provisions are hedged with conditions, its central prohibition on deprivation carries a grandfathered exception that has proved dangerously useful to modern governments, and it has neither a treaty body nor a reporting mechanism nor a body of case law of its own. And yet its central claim has held up remarkably well. Nationality is not a gift. It is not a reward for good behaviour, and it is not something a state may withhold simply because it finds a person inconvenient. The rules on conferral, and the rules on non-withdrawal, together express a simple principle: no state may act in a way that leaves a human being outside the membership of every state on earth. Article 9, which prohibits deprivation of nationality on racial, ethnic, religious or political grounds without any exception whatsoever, is the clearest statement of that principle in international law, and it was written by people who had watched what happens when it is ignored. The task now is not to rewrite the treaty. It is to give its weakest provisions the strongest available reading, to connect it to the human rights law that grew up around it, to press the remaining states to join, and to attend to the administrative machinery, the birth registers and the identity counters, where statelessness is actually made and unmade. The person waiting at that counter without a document is not an abstraction. They are the entire subject of this treaty, and the true measure of whether it has worked. Hashtags #statelessness #nationality_law #right_to_a_nationality #1961_Convention #stateless_persons #citizenship_rights #birth_registration #childhood_statelessness #deprivation_of_nationality #international_law #human_rights #legal_identity #jus_soli_safeguard #state_succession #UNHCR References Balaton-Chrimes, S. (2024). Jus tribalis: statelessness, ethnicity and registration in Kenya. Statelessness and Citizenship Review, 6(1), 63-87. https://doi.org/10.35715/SCR6001.114 Bloom, T., and Kingston, L. N. (Eds.). (2021). Statelessness, governance, and the problem of citizenship. Manchester: Manchester University Press. https://doi.org/10.7765/9781526156426 Cheong, A. R. (2025). Who counts as a stateless person? Nation-statist logics and the liabilities of potential citizenship elsewhere. Ethnic and Racial Studies, 48(4), 883-905. https://doi.org/10.1080/01419870.2024.2404482 Foster, M. (2022). The 1961 Convention on the Reduction of Statelessness: history, evolution and relevance. Statelessness and Citizenship Review, 4(1), 188-193. Gordon, H. (2024). NZYQ v Minister for Immigration, Citizenship and Multicultural Affairs: the end of indefinite immigration detention in Australia. Statelessness and Citizenship Review, 6(1), 144-151. Hyma, R. (2024). Participatising statelessness research: towards more inclusive scholarship. Statelessness and Citizenship Review, 6(1), 130-136. Immanuel, A. M. P. (2023). The customary obligation to avoid, reduce, or prevent statelessness in South Asia. Asian Journal of International Law, 13(2), 244-268. Jain, N. (2022). Manufacturing statelessness. American Journal of International Law, 116(2), 237-288. https://doi.org/10.1017/ajil.2022.2 Kingston, L. N. (2021). The weaponisation of citizenship: punishment, erasure, and social control. In T. Bloom and L. N. Kingston (Eds.), Statelessness, governance, and the problem of citizenship (pp. 99-111). Manchester: Manchester University Press. https://doi.org/10.7765/9781526156426.00016 Kochenov, D. V. (2024). Statelessness: a radical rethinking of the dominant citizenism paradigm. Annual Review of Law and Social Science, 20, 117-139. https://doi.org/10.1146/annurev-lawsocsci-041822-045326 Manby, B. (2024). Schrodinger's citizenship: framing perspectives for the resolution of statelessness. Statelessness and Citizenship Review, 6(1), 5-37. Marambio, H.-U. (2024). Germany: prohibition of deportation on Bidoons with Comoros passports. Statelessness and Citizenship Review, 6(1), 152-158. Mezmur, B. D. (2022). Making their days count: the 1961 Convention on the Reduction of Statelessness and the Convention on the Rights of the Child. Statelessness and Citizenship Review, 4(1), 198-204. Mulmi, S., and Biedscheid, J. (2025). World Conference on Statelessness 2024: key takeaways. Statelessness and Citizenship Review, 6(2), 214-222. https://doi.org/10.35715/SCR6002.114 Paulussen, C. (2021). Stripping foreign fighters of their citizenship: international human rights and humanitarian law considerations. International Review of the Red Cross, 103(916-917), 605-618. Prener, C. B. (2022). Denationalisation and its discontents: citizenship in the 21st century. Leiden: Brill Nijhoff. Ramnath, K. (2023). Boats in a storm: law, migration, and decolonization in South and Southeast Asia, 1942-1962. Stanford: Stanford University Press. Robertson, K. (2022). Practical measures to meaningfully implement Article 1(1) of the 1961 Convention on the Reduction of Statelessness in Australian law and practice. Statelessness and Citizenship Review, 4(1), 194-197. Sardelic, J. (2021). The fringes of citizenship: Romani minorities in Europe and civic marginalisation. Manchester: Manchester University Press. Sperfeldt, C. (2021). Statelessness and governance at the periphery: nomadic populations and the modern state in Thailand, Cote d'Ivoire, and Lebanon. In T. Bloom and L. N. Kingston (Eds.), Statelessness, governance, and the problem of citizenship (pp. 196-209). Manchester: Manchester University Press. https://doi.org/10.7765/9781526156426.00026 Stewart, M. (2023). Review of Statelessness: a modern history, by Mira L. Siegelberg. American Journal of International Law, 117(2), 371-378. https://doi.org/10.1017/ajil.2022.74 Szigeti, P. D. (2025). The dilemmas of Schrodinger's citizenship. Harvard International Law Journal, 66(1), 231-280. Van Thillo, C. (2024). Persistent gaps in protection: stateless and forcibly displaced in Ukraine and abroad. Statelessness and Citizenship Review, 6(1), 88-108. Vlieks, C. (2022). Nationality and statelessness in Europe: European law on preventing and solving statelessness. Cambridge: Intersentia. Von Rutte, B. (2022). The human right to citizenship: situating the right to citizenship within international and regional human rights law. Leiden: Brill Nijhoff. Yadav, A. (2023). The human right to citizenship [book review]. Statelessness and Citizenship Review, 5(2), 245-251. https://doi.org/10.35715/SCR5002.1111

The International Convention on the Protection of the Rights of All Migrant Workers and Members of Their Families, adopted by the United Nations General Assembly on 18 December 1990 and in force since 1 July 2003, remains the most complete treaty ever written on the rights of people who cross borders to work. It covers ninety three articles arranged in nine parts, and it is the only core United Nations human rights instrument that speaks directly to workers in an irregular situation as well as to those with papers. Yet more than three decades after adoption, the treaty has attracted only about sixty States parties, almost all of them countries that send workers abroad rather than countries that receive them. This article asks a simple question with a complicated answer: why does a treaty that so many governments praise in principle remain so widely ignored in practice, and what does that gap mean for the roughly 169 million people who work outside their country of birth? The study uses doctrinal legal analysis combined with a review of secondary literature published mainly between 2021 and 2026, together with the reporting record of the Committee on Migrant Workers. It maps the structure and content of the Convention, examines its monitoring machinery, explains the political economy of non ratification, and compares the treaty with International Labour Organization standards and with the 2018 Global Compact for Safe, Orderly and Regular Migration. The article argues that the Convention is not a failed instrument but an under used one. Its low ratification count understates its real influence, which travels through national legislation, regional agreements, soft law commitments, and the advocacy work of trade unions and civil society. At the same time, the article does not treat the ratification gap as harmless. Without destination State participation, the treaty cannot deliver the reciprocity that its drafters imagined, and the burden of protection falls on the States least able to carry it. The article ends with a set of practical recommendations for governments, treaty bodies, universities, and student researchers. Keywords: migrant workers, international human rights law, labour migration, United Nations treaty bodies, ratification, irregular migration, migration governance Introduction Human movement for work is one of the oldest patterns in economic life, but the legal protection of the people who move is a modern and still unfinished project. The International Labour Organization estimates that about 169 million people worldwide are #migrant_workers, meaning they are employed, or looking for employment, in a country other than the one where they hold citizenship. These workers build stadiums, harvest crops, care for elderly people, staff hospitals, drive delivery vehicles, clean offices, crew fishing boats, and write software. They send home remittances that exceed, in many low income countries, the combined value of foreign aid and foreign direct investment. They are also, as a group, more exposed to abuse than almost any other category of worker. The United Nations answered this reality with a dedicated treaty. The International Convention on the Protection of the Rights of All Migrant Workers and Members of Their Families, usually shortened to the ICMW or the ICRMW, was adopted by General Assembly resolution 45 slash 158 on 18 December 1990 without a vote. It took almost thirteen years to gather the twenty ratifications needed for entry into force, which finally came on 1 July 2003. The treaty is one of the nine core United Nations #human_rights instruments, and it is monitored by a body of fourteen independent experts known as the Committee on the Protection of the Rights of All Migrant Workers and Members of Their Families, or simply the Committee on Migrant Workers. On paper, the Convention is impressive. It recognises that a person does not lose their humanity by crossing a border without permission. It extends a substantial floor of rights to every migrant worker, regardless of status, and then adds a further layer of rights for those who are documented. It names specific groups whose situations are easily forgotten, including frontier workers, seasonal workers, seafarers, workers on offshore installations, itinerant workers, project tied workers, specified employment workers, and self employed migrants. It obliges States to cooperate with one another in managing #labour_migration in a way that is sound, equitable, humane and lawful. In practice, the Convention occupies an awkward position. Its record of #ratification is the weakest of any core United Nations human rights treaty. Around sixty States are parties. Not a single major destination country in Western Europe, North America, the Gulf, or the industrialised Asia Pacific has joined. The States that have joined are overwhelmingly countries of origin, such as Mexico, the Philippines, Morocco, Sri Lanka, Bangladesh, Indonesia, Egypt, and a large number of West African and Latin American States. This produces a strange result. The treaty binds most tightly the governments whose nationals are abroad and whose leverage over employers in foreign countries is smallest, while leaving untouched the governments whose labour markets actually absorb migrant labour. This article examines that paradox in detail. It is written for students and early career researchers, and it therefore tries to explain legal concepts in ordinary language without simplifying them into inaccuracy. The article has four aims. First, to describe accurately what the Convention says and how it is structured. Second, to explain how the monitoring system works and what it has and has not achieved. Third, to analyse the reasons for low ratification, drawing on recent scholarship rather than on the assumptions that circulate in policy debate. Fourth, to assess the Convention in relation to neighbouring instruments, especially the labour standards of the International Labour Organization and the 2018 Global Compact for Safe, Orderly and Regular Migration, and to ask whether the Convention still matters in a governance environment increasingly shaped by non binding cooperation. The central argument is that the Convention should be read as an unfinished constitutional settlement rather than as a dead letter. Its normative content has spread far beyond its formal membership. Courts, national legislatures, regional bodies, and even non ratifying States have absorbed parts of its logic, sometimes without acknowledging the source. But the argument is not celebratory. The absence of destination States means that the treaty cannot function as the reciprocal bargain it was designed to be, and the gap between the standards on paper and the conditions in fields, factories, construction sites and private homes remains wide. 1.1 Research questions The article is organised around five questions. First, what protections does the Convention actually create, and how do they differ from protections already available under other human rights treaties? Second, how effective is the monitoring machinery of the Committee on Migrant Workers, given that the complaint procedures in Articles 76 and 77 have never entered into force? Third, what explains the persistent refusal of destination States to ratify, and are the standard explanations supported by evidence? Fourth, how does the Convention interact with the International Labour Organization framework and with the Global Compact for Migration, and does the growth of #soft_law weaken or strengthen the treaty? Fifth, what practical reforms could raise the real world impact of the Convention without waiting for a wave of ratifications that may never come? 1.2 Structure of the article Section 2 reviews the relevant literature. Section 3 sets out the methodology and its limits. Section 4 traces the historical background and the drafting process. Section 5 describes the structure and scope of the treaty. Section 6 examines the substantive rights, distinguishing the universal floor in Part III from the additional entitlements in Part IV. Section 7 analyses State obligations on migration management under Part VI. Section 8 studies the Committee on Migrant Workers. Section 9 addresses the ratification gap. Section 10 offers regional case studies. Section 11 places the Convention alongside the International Labour Organization system and the Global Compact. Section 12 discusses contemporary pressures, including domestic work, climate related mobility, platform employment, and the aftermath of the COVID nineteen pandemic. Section 13 provides a critical assessment. Section 14 discusses implications. Section 15 sets out recommendations. Section 16 notes limitations and future research directions, and Section 17 concludes. Literature Review 2.1 The founding debate Scholarship on the Convention began almost as soon as the text was adopted, and it was pessimistic from the start. Early commentators noticed that the treaty had been drafted mainly by sending States and that receiving States had participated in the working group without any serious intention of joining the result. This early literature established a framing that still dominates: the Convention as a well written instrument stranded by political indifference. The most influential collective statement of that view remains the volume edited by Cholewinski, de Guchteneire and Pecoud, which asked directly why so few States had ratified and offered historical, political, legal and institutional answers. That volume is now old enough that its data is dated, but its analytical categories continue to shape the field. Later work, especially the collection edited by Desmond, tried to shift the conversation away from ratification counting and towards the question of added value. If the Convention says things that no other treaty says, then its content matters even where it is not formally binding. 2.2 The added value debate A second strand of literature asks whether the Convention actually adds anything. Sceptics point out that the International Covenant on Civil and Political Rights and the International Covenant on Economic, Social and Cultural Rights already apply to everyone within a State's jurisdiction, migrants included. On this reading, the Convention restates existing rights and adds only administrative detail. The counter argument, developed in detail by Georgopoulou, Schrempf and Venturi, is that the Convention does several things the general treaties do not. It defines categories of migrant worker with precision, which matters because vague categories are the entry point for exclusion. It addresses situations that only arise in migration, such as the confiscation of identity documents, the destruction of passports, the denial of #consular_assistance, the treatment of workers whose employer also controls their visa, and the transfer of earnings across borders. It expressly refuses to allow irregular status to be used as a reason to deny basic protection. And it imposes duties of interstate cooperation that no other human rights treaty imposes. Recent scholarship broadly accepts the added value argument while remaining sober about enforcement. The debate has moved from whether the Convention is worth having to how its content can be delivered through channels other than ratification. 2.3 The turn to soft law The adoption of the #Global_Compact_for_Migration in December 2018 reorganised the literature again. The Compact is not legally binding, contains twenty three objectives, and was endorsed by a large majority of United Nations members, including many destination States that have never gone near the Convention. Some scholars welcomed this as pragmatic progress. Others, including Desmond, warned that a non binding instrument endorsed by everyone could quietly displace a binding instrument ratified by few, producing the appearance of progress without the substance. Panizzon, Vitiello and Molnar examined this question directly, asking whether soft normativity relativises both the risks and the gains of global compacts. Bast, Endres de Oliveira and Wessels showed how the Compact can be used to strengthen rights claims in a regional legal order that has otherwise resisted the Convention. Atak and colleagues, reviewing the first cycle of Compact reviews, found that the reporting exercise produced a great deal of self description and rather little accountability. Jouzier, writing on Canada, argued that the Compact retains untapped potential to advance migrant rights if it is read through mobility facilitation rather than through migration control. The relevant point for this article is that the Committee on Migrant Workers itself has now entered this debate. Its General Comment No. 6, adopted in December 2024, deals explicitly with convergent protection through the Convention and the Global Compact, insisting that the Compact must be interpreted consistently with binding obligations rather than as an alternative to them. 2.4 Regional and sectoral literature A large empirical literature examines conditions in specific corridors, and it supplies the factual foundation for legal analysis. Work on the Gulf Cooperation Council States and on the #kafala sponsorship system shows how legal residence tied to a single employer produces dependency that no amount of formal labour law can cure. Mustaqeem, writing on the South Asia to Gulf corridor, documents how the pandemic exposed the fragility of temporary migration regimes and argues for a regional alliance among sending States. Burmeister Rudolph shows how categories of skill are constructed in regional governance processes such as the Colombo Process, with direct consequences for who is protected. Sectoral literature on #domestic_workers matters because domestic work sits at the intersection of gender, migration and the informal economy. Liem and colleagues argue that ratification of International Labour Organization Convention No. 189 is a precondition for meaningful protection, noting that private households are rarely treated as workplaces and are therefore rarely inspected. Health research, such as the study by Sharma and colleagues on Nepali workers in the Gulf and Malaysia, records the psychological consequences of precarious status in ways legal scholarship cannot capture alone. 2.5 The gap in the literature Three gaps stand out. First, most legal writing on the Convention is produced in and for the Global North, even though the treaty's membership is overwhelmingly in the Global South, which means that the perspectives of States parties are under represented. Second, there is very little systematic study of what actually changes inside a country after it ratifies, partly because the Committee's reporting cycle produces uneven data. Third, student oriented and teaching material on the Convention is thin, which matters because the next generation of migration lawyers and policy officers is being trained largely on refugee law, where the doctrinal literature is far richer. This article tries to speak to the third gap directly. Methodology 3.1 Approach This is a doctrinal and analytical study. It does not generate new empirical data. It works from three bodies of material. The first is the primary legal text: the Convention itself, its ninety three articles, and the general comments and concluding observations issued by the Committee on Migrant Workers. Particular attention is given to General Comment No. 1 on migrant domestic workers, General Comment No. 2 on the rights of migrant workers in an irregular situation, the joint general comments issued with the Committee on the Rights of the Child on children in the context of international migration, General Comment No. 5 on liberty and freedom from arbitrary detention, and General Comment No. 6 on convergent protection with the Global Compact. The second is the comparative treaty framework, principally International Labour Organization Conventions No. 97, No. 143 and No. 189, together with the Global Compact for Safe, Orderly and Regular Migration. The third is peer reviewed secondary literature, selected with a preference for material published from 2021 onwards, supplemented by a small number of older works whose analytical influence is unavoidable. 3.2 Selection criteria Sources were selected on three criteria: relevance to the Convention or to the conditions the Convention regulates, recency, and reliability. Journal articles in indexed publications, monographs from academic publishers, and official documents of United Nations bodies were preferred. Advocacy material and news reporting were consulted for context but are not relied upon for legal propositions. 3.3 Limitations Doctrinal analysis has well known limits. It tells us what the law says, not what happens in a labour camp at five in the morning. This study therefore treats empirical scholarship as a check on doctrinal optimism. A second limitation is language. The literature consulted is overwhelmingly in English, which under represents scholarship in Spanish, Arabic, French, Bahasa Indonesia, and the South Asian languages, all of which are important given the geography of the treaty's membership. A third limitation is that the Committee's own record is incomplete: many States parties report late or not at all, and the absence of an operational #individual_complaints procedure means there is no case law to analyse. These limitations are discussed again in Section 16. Historical Background and Drafting 4.1 The road to 1990 Concern about the treatment of foreign workers reached the United Nations in the early 1970s. In 1972 the Economic and Social Council expressed alarm at the illegal transportation of workers to Europe and at the exploitation of African workers in conditions close to slavery and forced labour. In the same year, the General Assembly condemned discrimination against foreign workers and called on governments to end such practices. These early resolutions matter because they set the tone. The Convention was born out of a scandal, not out of an abstract wish for legal tidiness. The International Labour Organization had already been in this field for decades. Convention No. 97 of 1949 dealt with migration for employment. Convention No. 143 of 1975 dealt with migrations in abusive conditions and with equality of opportunity and treatment. Both had modest ratification records. The United Nations decision to draft a separate, broader instrument was therefore partly a judgment that the labour standards approach was not enough, and partly a reflection of pressure from sending States who wanted a treaty with a human rights frame rather than a purely industrial relations frame. 4.2 The working group A General Assembly working group, open to all Member States, began drafting in 1980 and worked for a full decade. The composition of the group tells much of the story. Countries of origin, especially Mexico and Morocco, together with a group of Mediterranean and Nordic States, drove the process. Major destination States attended, argued for restrictions, and secured important concessions, but never committed to joining. Two of those concessions shaped the final text. The first was the preservation of #state_sovereignty over admission. Article 79 makes clear that nothing in the Convention affects the right of each State to establish the criteria governing admission of migrant workers and members of their families. The treaty regulates treatment, not entry. This was the price of adoption without a vote. The second was the two tier structure. Destination States were unwilling to grant the full range of rights to workers in an irregular situation. The compromise was to create a universal floor in Part III, applicable to all migrant workers whatever their status, and a wider set of entitlements in Part IV reserved for those who are documented or in a regular situation. This structure is the most distinctive feature of the treaty and, as Section 6 argues, also its most contested. 4.3 Adoption and the long wait The General Assembly adopted the text on 18 December 1990 without a vote, which is often presented as consensus. It is more accurate to describe it as an absence of opposition. States that had no intention of ratifying saw no reason to force a division. The treaty then sat for years. Before 1998 only nine States had ratified. The pace picked up between 1998 and 2003, largely through the efforts of a transnational advocacy campaign involving trade unions, migrant associations, church networks, and United Nations agencies. The twentieth ratification arrived in March 2003 and the Convention entered into force on 1 July 2003. The Committee held its first session in March 2004. That timeline is itself an argument. A treaty that takes thirteen years to reach a threshold of twenty ratifications, in a period when the Convention on the Rights of the Child reached near universal membership in a few years, is telling us something about the political value that States assign to migrants. Structure and Scope of the Convention 5.1 The nine parts The Convention has ninety three articles in nine parts. Part I, Articles 1 to 6, covers scope and definitions. Part II, Article 7, sets out the principle of #non_discrimination. Part III, Articles 8 to 35, contains the human rights of all migrant workers and members of their families, irrespective of migration status. Part IV, Articles 36 to 56, contains other rights of migrant workers and members of their families who are documented or in a regular situation. Part V, Articles 57 to 63, contains provisions applicable to particular categories of migrant workers. Part VI, Articles 64 to 71, addresses the promotion of sound, equitable, humane and lawful conditions in connection with international migration. Part VII, Articles 72 to 78, deals with application, and establishes the Committee. Part VIII, Articles 79 to 84, contains general provisions. Part IX, Articles 85 to 93, contains final provisions. The architecture matters. Rights are not scattered randomly; they are graded by status and by category, which makes the treaty easier to negotiate and harder to apply. 5.2 Who counts as a migrant worker Article 2 defines a migrant worker as a person who is to be engaged, is engaged, or has been engaged in a remunerated activity in a State of which he or she is not a national. Three tenses are used deliberately. The definition covers the person preparing to depart, the person currently employed, and the person whose employment has ended. This is significant, because many abuses happen at the recruitment stage and at the termination stage rather than during employment. Article 2 then names specific categories. A frontier worker retains habitual residence in a neighbouring State and returns there daily or at least once a week. #frontier_workers are common along land borders in Europe, Southern Africa, and North America, and their situation is legally awkward because they belong to two labour markets at once. A #seasonal_workers arrangement covers work that by its character depends on seasonal conditions and is performed only during part of the year. Agriculture and tourism dominate this category. A seafarer is employed on a vessel registered in a State of which he or she is not a national. #seafarers are among the least visible workers in the world economy, and jurisdictional confusion is chronic. A worker on an offshore installation is under the jurisdiction of a State of which he or she is not a national. An itinerant worker has habitual residence in one State but travels to another for short periods due to the nature of the occupation. A project tied worker is admitted for a defined period to work solely on a specific project. A specified employment worker is sent for a limited period to perform a specific duty or assignment, or engages in work requiring professional, commercial, technical or other highly specialised skill. A self employed worker engages in remunerated activity otherwise than under a contract of employment. The inclusion of #self_employed_migrants is unusual and important, because much migrant labour is disguised as independent contracting precisely to escape labour law. 5.3 Exclusions Article 3 excludes several groups. These include persons sent or employed by international organisations, persons sent or employed by a State outside its territory to perform official functions, persons taking up residence in a State different from their State of origin as investors, refugees and stateless persons unless national legislation or an international instrument in force for the State party provides otherwise, students and trainees, and seafarers or workers on offshore installations who have not been admitted to take up residence and engage in remunerated activity in the State of employment. The exclusion of students and trainees is a live problem. Student visa routes are widely used as labour migration channels in Australia, Canada, the United Kingdom, Japan and the Gulf, and the workers who enter through them are frequently among the most exploited. A treaty that excludes them by definition leaves a hole in its own coverage. 5.4 Family members Article 4 defines members of the family as persons married to migrant workers or having with them a relationship that produces effects equivalent to marriage according to applicable law, together with dependent children and other dependent persons recognised as family members by applicable legislation or by bilateral or multilateral agreements. The reference to applicable law is a weakness. It allows a State of employment to define family narrowly and then claim compliance. 5.5 Documented and undocumented Article 5 draws the central distinction. Migrant workers are documented or in a regular situation if they are authorised to enter, stay and engage in a remunerated activity in the State of employment pursuant to the law of that State and to international agreements to which that State is a party. They are non documented or in an irregular situation if they do not comply with those conditions. Article 35 then adds a crucial safeguard against a predictable objection. Nothing in Part III shall be interpreted as implying the regularisation of the situation of migrant workers who are non documented, or any right to such regularisation. In other words, giving rights to a worker without papers does not give that worker papers. This provision was inserted to reassure destination States. It has not persuaded them, but it does clarify the logic of the treaty: protection and status are separate questions. The Rights Framework 6.1 The universal floor: Part III Part III is the moral core of the Convention. It applies to every migrant worker and every member of their family, whether or not they hold papers. Its provisions can be grouped into five clusters. 6.1.1 Physical integrity and freedom from exploitation Article 9 protects the right to life. Article 10 prohibits torture and cruel, inhuman or degrading treatment. Article 11 prohibits slavery, servitude and #forced_labour. These provisions restate existing law, but the restatement is not empty. The International Labour Organization has found that migrant workers are more than three times as likely to be in forced labour as non migrant adult workers. Restating the prohibition in a migration specific treaty forces States to think about the mechanisms that produce forced labour in migration: debt from recruitment, confiscated passports, unpaid wages used as a retention device, and legal residence tied to a single employer. Article 21 addresses one of these mechanisms directly. It prohibits anyone other than a public official duly authorised by law from confiscating, destroying or attempting to destroy identity documents, documents authorising entry or stay, or work permits. It also prohibits the destruction of a passport in any circumstances. Nobody who has read the empirical literature on the Gulf, on Malaysian plantations, or on European agriculture can regard this as a minor provision. Passport confiscation is the single most effective tool of control available to an abusive employer. 6.1.2 Liberty, due process, and expulsion Articles 16 to 20 create a detailed code on liberty and #due_process. Migrant workers are entitled to liberty and security of person and to effective protection by the State against violence, physical injury, threats and intimidation, whether by public officials or by private individuals, groups or institutions. This express extension to private actors is important: much of the violence migrant workers face comes from employers, recruiters, landlords and gangs, not from the police. Article 16 also requires that a migrant worker deprived of liberty be told the reasons for arrest, as far as possible in a language they understand. Article 17 requires humane treatment in detention and respect for cultural identity. Article 18 guarantees equality with nationals before courts and tribunals, including free assistance of an interpreter. Article 22 is one of the treaty's strongest clauses. It prohibits measures of #collective_expulsion. Each case of expulsion must be examined and decided individually. Expulsion may only be carried out in pursuance of a decision taken by a competent authority in accordance with law, and the decision must be communicated to the person in a language they understand. The person must, except where a final decision is pronounced by a judicial authority, have the right to submit reasons against expulsion and to have the case reviewed. If an expulsion decision already executed is subsequently annulled, the person is entitled to seek compensation, and the earlier decision may not be used to prevent re entry. Article 22 also provides that expulsion shall not prejudice rights acquired under the law of the State of employment, including the right to receive wages and other entitlements. That last sentence closes a common trick: expelling a worker in order to avoid paying them. Article 23 gives migrant workers the right to have recourse to the protection and assistance of the consular or diplomatic authorities of their State of origin. #detention without consular notification is a routine violation and a routine cause of disappearance. 6.1.3 Labour rights Article 25 requires that migrant workers enjoy treatment not less favourable than that which applies to nationals of the State of employment in respect of remuneration and other conditions of work, including overtime, hours of work, weekly rest, holidays with pay, safety, health, and termination of the employment relationship. It also applies to other terms of employment, including minimum age of employment and restriction on home work. Crucially, Article 25 states that it is not lawful for employers to derogate from this principle in private contracts, and that States must take all appropriate measures to ensure that migrant workers are not deprived of rights derived from this principle by reason of any irregularity in their stay or employment. In plain terms, an employer cannot pay a worker less because the worker's visa has expired. The #equality_of_treatment guarantee survives the loss of status. Article 26 gives all migrant workers the right to take part in meetings and activities of trade unions, to join them freely, and to seek their aid and assistance. Note the limit. Part III grants the right to join and participate. The right to form associations and trade unions appears in Part IV, Article 40, and is therefore reserved for documented workers. This is one of the clearest examples of the two tier structure operating to the disadvantage of the workers who need collective power most. Article 27 addresses #social_protection. Migrant workers and their families enjoy the same treatment as nationals in respect of social security, in so far as they fulfil the requirements provided for by the applicable legislation of the State of employment and the applicable bilateral and multilateral treaties. Where the applicable legislation does not allow a benefit, States must examine the possibility of reimbursing contributions made. This is a weaker guarantee than it appears, because it is conditioned on national law, but the reimbursement clause is a genuine advance: it treats contributions as the worker's property rather than as a windfall for the host State's treasury. Article 28 guarantees the right to receive any medical care that is urgently required for the preservation of life or the avoidance of irreparable harm to health, on the basis of equality of treatment with nationals. Such #emergency_medical_care may not be refused by reason of any irregularity with regard to stay or employment. This is a floor, not a ceiling, and it is one of the provisions that destination States most frequently cite as a reason for non ratification, on the theory that it undermines immigration control. 6.1.4 Children and family Article 29 provides that each child of a migrant worker has the right to a name, to registration of birth, and to a nationality. Birth registration is the gateway document for every other right, and unregistered children of migrant parents are among the world's most invisible populations. Article 30 provides that each child of a migrant worker has the basic right of access to education on the basis of equality of treatment with nationals of the State concerned. The article states plainly that access to public preschool educational institutions or schools shall not be refused or limited by reason of the irregular situation with respect to stay or employment of either parent or by reason of the irregularity of the child's stay in the State of employment. The #right_to_education for the children of undocumented workers is one of the treaty's most concrete achievements, and it has migrated into national law in several countries that have never ratified. Article 31 requires States to ensure respect for the cultural identity of migrant workers and their families and not to prevent them from maintaining cultural links with their State of origin. 6.1.5 Economic autonomy and information Article 32 provides that upon termination of stay, migrant workers and their families have the right to transfer their earnings and savings and their personal effects and belongings. #remittances are the economic heart of labour migration, and any obstacle to transferring earnings is a direct tax on the migrant. Article 33 requires that migrant workers and their families have the right to be informed by the State of origin, the State of employment or the State of transit, as the case may be, concerning their rights under the Convention, the conditions of their admission, and their rights and obligations under the law of the States concerned. Information must be provided free of charge and, as far as possible, in a language they are able to understand. This is a small clause with large consequences. A right that a worker does not know about is not a right. 6.2 The additional layer: Part IV Part IV applies only to documented migrant workers and their families. It contains the rights that States were unwilling to extend to everyone. Article 37 requires that workers be fully informed, before departure or at the latest at the time of admission, of all conditions applicable to their admission, stay and remunerated activity, and of the requirements they must satisfy in the State of employment. Article 38 requires States of employment to make every effort to authorise temporary absences without effect on the authorisation to stay or work, and to take account of the special needs and obligations of migrant workers with respect to their States of origin. Article 39 guarantees liberty of movement within the territory of the State of employment and freedom to choose residence, subject to restrictions provided by law and necessary to protect national security, public order, public health or morals, or the rights and freedoms of others. Article 40 grants the right to form associations and trade unions in the State of employment. As noted above, this creates a serious anomaly. #freedom_of_association is the mechanism through which workers defend every other right. Denying it to undocumented workers effectively denies them the means of enforcement. Article 41 grants the right to participate in public affairs of the State of origin and to vote and be elected in that State, in accordance with its legislation. Article 42 encourages the establishment of procedures or institutions through which the special needs of migrant workers can be taken into account in the State of employment, and provides that States of employment may, in accordance with national legislation, grant political rights. The word may is doing a great deal of work in that sentence. Article 43 establishes equality of treatment with nationals in relation to access to educational institutions and services, vocational guidance and placement services, vocational training and retraining facilities, housing including social housing schemes, social and health services, cooperatives and self managed enterprises, and access to and participation in cultural life. It also provides for protection against exploitation in respect of rents and prices. Article 44 is the family provision. States must recognise that the family is the natural and fundamental group unit of society and is entitled to protection, must take appropriate measures to ensure the protection of the unity of the families of migrant workers, and must take measures they deem appropriate and within their competence to facilitate the reunification of migrant workers with their spouses and dependent unmarried minor children. The formula is deliberately weak: measures they deem appropriate. #family_reunification is thus encouraged rather than guaranteed, and this is one of the treaty's real disappointments given that its title promises protection for families. Article 45 extends equality of treatment to family members in relation to education, vocational guidance and training, social and health services, and cultural life, and requires States of employment, in collaboration with States of origin where appropriate, to pursue a policy aimed at facilitating the integration of children of migrant workers in the local school system, particularly in respect of teaching them the local language. Articles 46 to 48 deal with economic matters: exemption from import and export duties on personal and household effects and on work equipment, protection against double taxation, and an obligation on States of employment to permit and facilitate the transfer of earnings and savings. Articles 49 and 51 protect status. Where separate residence and work authorisations are required, the residence permit must run for at least as long as the work authorisation, and a worker does not fall into irregularity merely because the job ends before the permit expires. Workers who are not free to choose their employment keep their residence authorisation and may seek alternative work for the remainder of the permit. Article 52 allows States to restrict free choice of employment, including for a qualifying period that may not exceed two years. Article 54 grants equality of treatment in respect of protection against dismissal, unemployment benefits, public employment schemes, and access to alternative work. Article 56 provides that documented migrant workers and their families may not be expelled except for reasons defined in the national legislation of the State of employment, and subject to the safeguards established in Part III. It adds that expulsion shall not be resorted to for the purpose of depriving a migrant worker of the rights arising out of the authorisation of residence and the work permit, and that in considering expulsion, account should be taken of humanitarian considerations and of the length of time the person has already resided in the State of employment. 6.3 Particular categories: Part V Part V, Articles 57 to 63, adapts the framework to specific categories. Frontier workers and seasonal workers are entitled to those Part IV rights that can be applied to them by reason of their presence and work in the State of employment, and States must consider allowing seasonal workers employed for a significant period to take up other work. Itinerant workers, project tied workers, specified employment workers and self employed workers each receive tailored treatment. The provisions are technical but the underlying principle is consistent: the shorter and more conditional the stay, the fewer the rights. That principle is understandable as politics and troubling as ethics, because precisely those workers with the shortest and most conditional stays are the ones with the least ability to complain. State Obligations on Migration Management: Part VI Part VI is often overlooked, but it is what makes the Convention different from a pure human rights treaty. It imposes duties of cooperation. Article 64 requires States parties, without prejudice to Article 79, to consult and cooperate with a view to promoting sound, equitable and humane conditions in connection with international migration of workers and their families. Such consultation shall have due regard not only to labour needs and resources but also to the social, economic, cultural and other needs of migrant workers and their families, as well as to the consequences of migration for the communities concerned. Article 65 requires States to maintain appropriate services to deal with questions concerning international migration, including the formulation and implementation of migration policies, the exchange of information with other States, the provision of information to employers, workers and their organisations, and the provision of information and appropriate assistance to migrant workers regarding authorisations, formalities, arrangements for departure, travel, arrival, stay, remunerated activities, exit and return, as well as conditions of work and life and customs, currency and tax laws. Article 66 restricts recruitment. Operations for the recruitment of workers for employment in another State shall be limited to public services or bodies of the State in which such operations take place, public services or bodies of the State of employment on the basis of agreement between the States concerned, and a body established by virtue of a bilateral or multilateral agreement. Subject to authorisation, approval and supervision by public authorities, an employer or a person acting on their behalf may also be permitted to carry out such operations. This provision is the treaty's answer to the exploitative recruitment industry, and it points directly at the problem of #recruitment_fees, which remain the largest single driver of debt bondage in labour migration. Article 67 requires cooperation on orderly return, including on measures regarding the orderly return of migrant workers to the State of origin when they decide to return or their authorisation of stay expires, and, in the case of documented workers, on measures to promote adequate economic conditions for their resettlement and to facilitate their durable social and cultural reintegration. Article 68 requires States to collaborate to prevent and eliminate illegal or clandestine movements and employment of migrant workers in an irregular situation. Measures include appropriate action against the dissemination of misleading information relating to emigration and immigration, measures to detect and eradicate illegal or clandestine movements and to impose effective sanctions on persons, groups or entities that organise, operate or assist in organising such movements, and measures to impose effective sanctions on persons, groups or entities that use violence, threats or intimidation against migrant workers in an irregular situation. States of employment must also take all adequate and effective measures to eliminate employment of irregular migrant workers, including sanctions on employers. Importantly, the article adds that these measures shall not impair the rights of migrant workers vis a vis their employer arising from employment. Article 69 requires States parties in whose territory migrant workers are in an irregular situation to take appropriate measures to ensure that this situation does not persist. Whenever they consider the possibility of #regularization, they must take account of the circumstances of entry, the duration of stay, and other relevant considerations, in particular those relating to family situation. Article 70 requires that States take measures not less favourable than those applied to nationals to ensure that working and living conditions of migrant workers and their families in a regular situation are in keeping with standards of fitness, safety, health and principles of human dignity. Article 71 requires States to facilitate the repatriation to the State of origin of the bodies of deceased migrant workers or members of their families, and to provide assistance in the settlement of compensation matters. This clause is easy to skim past and impossible to forget once one has read reporting on the number of construction workers whose bodies are returned to South Asia each year, frequently with the cause of death recorded as natural causes or cardiac arrest, without autopsy. Part VI, read as a whole, shows that the Convention is not simply a list of entitlements. It is a framework for cooperative #migration_governance in which States of origin, transit and employment each carry duties. That framework only functions if all three types of State are inside it. The fact that destination States remain outside is therefore not merely an ideological problem. It is a structural defect that stops the machine from running. The Committee on Migrant Workers 8.1 Composition and mandate Part VII establishes the Committee on the Protection of the Rights of All Migrant Workers and Members of Their Families. It consists of fourteen independent experts of high moral standing, impartiality and recognised competence, elected by States parties and serving in their personal capacity. The Committee held its first session in March 2004 and is the smallest of the United Nations #treaty_body institutions, a fact that has practical consequences for its workload and visibility. 8.2 State reporting Articles 73 and 74 create the reporting procedure. States parties undertake to submit a report on the legislative, judicial, administrative and other measures they have taken to give effect to the Convention within one year of entry into force for the State concerned, and thereafter every five years and whenever the Committee so requests. The Committee examines the reports, engages in a constructive dialogue with the State delegation, and issues concluding observations containing recommendations. The system has three well documented weaknesses. First, many States report late. #periodic_reports arriving several years overdue are common, and some States parties have never submitted an initial report. Second, the reports are self assessments, and governments have an obvious interest in describing law rather than practice. Third, the Committee's recommendations are not binding, and there is no systematic follow up mechanism with teeth. Still, the reporting process is not theatre. The dialogue forces ministries to gather data they would otherwise not collect, creates an entry point for #civil_society and for #national_human_rights_institutions through shadow reports, and produces a documentary record that domestic litigants, journalists and opposition politicians can use. 8.3 General comments The Committee's most influential output has been its general comments, which interpret the Convention's provisions authoritatively. General Comment No. 1 (2011) addressed migrant domestic workers, recognising that domestic work is performed in private homes, is overwhelmingly done by women, is frequently excluded from labour law, and is therefore uniquely difficult to regulate and inspect. General Comment No. 2 (2013) addressed the rights of migrant workers in an irregular situation, and stands as the clearest statement anywhere in international law that irregular status does not deprive a person of human rights. Joint General Comments Nos. 3 and 4 (2017), issued together with the Committee on the Rights of the Child, addressed the human rights of children in the context of international migration, and stated that immigration detention of children is never in the best interests of the child. General Comment No. 5 (2021) addressed migrants' rights to liberty and freedom from arbitrary detention, insisting that administrative detention for migration purposes must be exceptional, time limited, subject to judicial review, and never applied to children. General Comment No. 6 (2024) addressed convergent protection through the Convention and the Global Compact for Safe, Orderly and Regular Migration. It reaffirmed that a human rights based approach is central to migration governance and that all migrants, whatever their status, are rights holders rather than objects of management. Its practical purpose was to prevent the non binding Compact from being used to dilute binding obligations. 8.4 The complaint procedures that do not exist Article 76 allows a State party to declare that it recognises the competence of the Committee to receive communications from another State party alleging non fulfilment of obligations. Article 77 allows a State party to declare that it recognises the competence of the Committee to receive communications from individuals subject to its jurisdiction who claim their rights have been violated. Each mechanism requires ten declarations to enter into force. Neither threshold has been reached. As of 2026, only a handful of States have made the Article 76 declaration, and only a handful the Article 77 declaration. In both cases the number remains below ten. The consequence is severe. The Committee cannot decide individual cases. There is no jurisprudence. There is no body of authoritative rulings that lawyers can cite, that judges can be shown, or that students can study. Compare this with the Human Rights Committee, whose views under the First Optional Protocol have shaped domestic law across the world, and the gap is obvious. The absence of an operational #individual_complaints procedure is arguably the single greatest weakness in the Convention's institutional design, and it is a weakness that States parties could fix tomorrow without any need for destination State cooperation. Ten declarations from among the existing sixty States parties would activate the mechanism. That so few have made the declaration suggests that even ratifying States are wary of #accountability, particularly because many of them are also destination States for workers from poorer neighbours, a point too often left out of the polite version of this story. Explaining the Ratification Gap 9.1 The numbers Roughly sixty States are parties. The list is dominated by countries of origin, including Mexico, Guatemala, Honduras, Ecuador, Peru, Bolivia, Colombia, Chile, Argentina, Uruguay, the Philippines, Indonesia, Sri Lanka, Bangladesh, Kyrgyzstan, Tajikistan, Azerbaijan, Turkey, Egypt, Morocco, Algeria, Mali, Senegal, Ghana, Nigeria, Uganda, Rwanda, Mozambique and Seychelles. Not one member of the European Union has ratified. Neither has the United States, Canada, Australia, New Zealand, Japan, the Republic of Korea, the Russian Federation, China, India, South Africa, or any Gulf Cooperation Council State. 9.2 The standard explanations Five explanations recur in the literature. The first is #political_will, or its absence. Migration is electorally toxic in many #destination_states, and no government wins votes by extending rights to foreigners, particularly foreigners without papers. Ratification carries a political cost with no corresponding political benefit. The second is the objection to Part III. Destination States object to the extension of rights to workers in an irregular situation, arguing that this weakens deterrence and rewards unlawful entry. Article 35 explicitly denies that Part III implies any right to regularisation, but the objection is not really a legal one. It is a symbolic one: governments do not wish to be seen to be granting anything at all to people they describe as illegal. The third is the claim of redundancy. Destination States argue that their existing obligations under the two Covenants, the Convention against Torture, the Convention on the Elimination of All Forms of Racial Discrimination, and domestic labour law already cover migrants, so a further treaty adds administrative burden without adding protection. The literature generally rejects this claim, since the Convention contains provisions with no analogue elsewhere, but the argument is politically useful because it allows a government to sound committed to rights while refusing the instrument. The fourth is complexity. At ninety three articles, the Convention is long, technical, and difficult to translate into domestic administrative practice. Ministries of interior and ministries of labour must coordinate, and in many States they do not. The fifth is the reciprocity problem, and it is the most interesting. Sending States have ratified in order to signal concern for their nationals abroad and to gain moral standing in bilateral negotiations. But because destination States have not ratified, the treaty produces no reciprocal duties. Sri Lanka, for example, has publicly reasoned that accepting the Committee's competence over individual complaints might expose it to claims regarding its own obligations while leaving the obligations of receiving countries untouched, placing it on an unequal footing. This is a rational calculation, and it shows how the absence of #reciprocity corrodes the treaty from within, discouraging even committed States parties from strengthening the system. 9.3 What the standard explanations miss Two further factors deserve more attention than they usually receive. The first is the two tier structure itself. By separating rights into a status blind floor and a status dependent tier, the drafters produced a text that satisfies nobody. Destination States object to the floor. Migrant rights advocates object to the tier. A treaty designed to be acceptable to everybody ended up looking like a concession to the wrong side from both directions. The second is the rise of alternative governance channels. Since 2018, States that wish to demonstrate engagement with migration governance can point to the Global Compact instead. The Compact costs nothing legally, generates conferences, reviews, and reports, and allows a government to appear cooperative. Chetail's observation that a management based approach to migration increasingly competes with a rights based approach captures this precisely. The existence of an attractive alternative reduces the pressure to ratify. 9.4 Does non ratification mean no effect? No. This is the central corrective that recent scholarship offers. The Convention influences law in at least four ways that do not require ratification. It shapes the reporting of other treaty bodies, which increasingly ask States about migrants' rights using concepts drawn from the Convention. It informs the Universal Periodic Review, in which States are routinely recommended to ratify, and in which the substance of the Convention is raised even by States that have not ratified it themselves. It provides the normative vocabulary used by trade unions, migrant associations and non governmental organisations, whose campaigns have produced concrete national reforms in countries that remain outside the treaty. And it supplies interpretive material to domestic and regional courts, which cite the Convention as evidence of international standards even where it does not bind them. The measure of a treaty's success is therefore not only how many States have signed it, but how far its ideas have travelled. On that measure the Convention has done better than its ratification list suggests. It has not, however, done well enough to change the daily reality of most migrant workers, and the #implementation_gap remains the honest headline. Regional Perspectives 10.1 Latin America Latin America is the region where the Convention has the deepest roots. Mexico, which co sponsored the drafting effort, and most Central and South American States are parties. The region has also produced the most ambitious jurisprudence on migrants' rights, notably the advisory opinion of the Inter American Court of Human Rights on the juridical condition and rights of undocumented migrants, which held that the fundamental principle of equality and non discrimination has entered the domain of peremptory norms and that a State may not subordinate labour rights to the immigration status of the worker. The region also demonstrates the limits of ratification. Several Latin American States parties are themselves significant destination countries, and the treatment of Haitian, Venezuelan, Nicaraguan and Bolivian workers within the region has attracted repeated criticism from the Committee. Ratification is a beginning, not an achievement. 10.2 The South Asia to Gulf corridor This corridor is where the gap between the treaty and reality is widest. Millions of workers from India, Bangladesh, Nepal, Pakistan and Sri Lanka work in the Gulf Cooperation Council States. Several origin States have ratified. No Gulf State has. The structural problem is the sponsorship system. Under #kafala, a migrant worker's legal residence is tied to a named sponsor, usually the employer. Leaving that employer without permission can convert a lawful worker into an absconder, subject to arrest, detention and deportation. The system does not merely permit abuse. It manufactures dependency, and dependency is the precondition of abuse. Reforms have been introduced. Qatar, under intense international scrutiny before the 2022 World Cup, removed exit permit requirements for most workers, introduced a minimum wage, and formally allowed workers to change employer without a no objection certificate. Bahrain, Kuwait, Oman, Saudi Arabia and the United Arab Emirates have adopted partial measures. The scholarly assessment is cautious. Several studies conclude that reforms have altered the vocabulary more than the structure, that enforcement is weak, that domestic workers are frequently excluded, and that employers retain practical control through wage withholding, accommodation, and the threat of absconding charges. Kalush and Saraswathi describe this pattern as reform without rights, noting that the sponsorship system is not a single law but an interlocking set of laws, regulations and entrenched practices, so that changing one element leaves the others intact. Mustaqeem's analysis of the pandemic period is instructive. When COVID nineteen struck, workers in this corridor lost wages, lost access to healthcare, were blamed as vectors of infection, and were stranded between host States that no longer wanted them and origin States that could not absorb them. The experience revealed that temporary migration regimes are not resilient, and that origin States acting alone have almost no leverage. Her proposal for a regional alliance among South Asian sending States is essentially an attempt to reconstruct, through collective bargaining, the reciprocity that the Convention was supposed to supply. 10.3 Africa Many African States are parties, and regional bodies such as the Economic Community of West African States operate long standing free movement protocols, while the African Union migration policy framework draws on the Convention's language. The obstacles here are capacity rather than principle: labour inspectorates are under resourced, informal employment predominates, and the data needed to monitor compliance is scarce. South Africa, the continent's largest destination country, has not ratified, and the treatment of workers from Zimbabwe, Mozambique, Lesotho and Malawi remains a persistent concern. 10.4 Southeast Asia Southeast Asia contains both major origin States, notably the Philippines and Indonesia, and major destination States, notably Malaysia, Thailand and Singapore. The Philippines and Indonesia have ratified; Malaysia, Thailand and Singapore have not. Research on Thailand illustrates the dynamic. Laikram and Pathak argue that Thailand's non ratification of the Convention, alongside its non ratification of the core International Labour Organization freedom of association conventions, leaves migrant workers without effective protection, deprives them of equal treatment with nationals, and prevents them from organising, and they contend that the pandemic sharpened rather than created this vulnerability. The regional body, the Association of Southeast Asian Nations, has adopted a consensus instrument on migrant workers, but it is non binding and its coverage of undocumented workers is contested. 10.5 Europe No European Union member State has ratified, and the European Union itself has consistently declined to encourage ratification. The standard explanation is that Union law already regulates labour migration through a series of directives, and that the Convention's provisions on irregular migrants conflict with the Union's returns policy. The scholarly response is that this misunderstands the relationship. European directives regulate admission and treat labour migration as a market management issue, leaving Member States free to decide how many workers to admit. The Convention is a human rights treaty that regulates treatment, not numbers, and Article 79 expressly preserves State control over admission. The two instruments are not in conflict; they operate on different planes. The refusal to ratify is a political choice, not a legal necessity. Desmond has described the relationship between the Convention and the Union as vexed rather than incompatible, and has argued that the Union's human rights rhetoric sits uneasily with its unwillingness to engage with the one treaty designed for the people its labour markets depend on. The Convention in a Crowded Field 11.1 International Labour Organization standards The Convention did not arrive on empty ground. Convention No. 97 of 1949 addressed migration for employment, requiring equality of treatment in respect of remuneration, membership of trade unions, social security, employment taxes and legal proceedings. Convention No. 143 of 1975 addressed migration in abusive conditions and required States to respect the basic human rights of all migrant workers, including those in an irregular situation, and to promote equality of opportunity and treatment for those lawfully within their territory. Convention No. 189 of 2011 on decent work for domestic workers is the most important recent addition. It requires that domestic workers enjoy fair terms of employment, decent working conditions, written particulars of employment, limits on hours, weekly rest, and protection against abuse, harassment and violence. Its premise, as its supporters emphasise, is that domestic workers are neither servants nor members of the family nor second class workers. Ratification remains limited, and no Gulf State has ratified it. The relationship between the two systems is complementary rather than competitive. #ILO_standards supply technical detail, tripartite governance involving employers and unions, and a supervisory machinery with real experience. The Convention supplies a human rights frame, coverage of family members, and explicit protection irrespective of status. Cholewinski has long argued that the two should be used together, and the Committee on Migrant Workers has consistently drawn on International Labour Organization material in its general comments. 11.2 The Global Compact for Migration The Global Compact for Safe, Orderly and Regular Migration was adopted on 19 December 2018. It is expressly non binding, upholds State sovereignty, and sets out twenty three objectives ranging from data collection and information provision to ethical recruitment, decent work, access to basic services, and cooperation on return and reintegration. It is the first intergovernmental agreement covering all dimensions of international migration, and it was endorsed by a very large majority of States, including many that have never ratified the Convention. Three assessments of the Compact are worth distinguishing. The optimistic reading holds that a widely endorsed soft instrument can build political consensus that a narrowly ratified hard instrument never achieved, and that objectives such as Objective 6 on fair and ethical recruitment and Objective 22 on portability of social security translate the Convention's content into language governments will accept. The sceptical reading holds that the Compact allows States to substitute performance for obligation. Atak and colleagues, reviewing the first cycle of regional reviews, found that States used the process largely to describe existing policy rather than to submit to scrutiny. The mediating reading, which this article adopts, is that the Compact is neither replacement nor threat but a channel, whose value depends on whether it is interpreted consistently with binding law. That is what General Comment No. 6 attempts to secure. Bast, Endres de Oliveira and Wessels show how the Compact can strengthen rights based arguments even within the European Union, which has resisted the Convention, and Jouzier makes a parallel argument for Canada, suggesting that its potential is unrealised because it is read through the lens of control rather than mobility. The risk that should worry students of this field is not that the Compact will be used against the Convention. It is that the Compact will be used instead of it, quietly, and that in twenty years the Convention will be a historical curiosity taught in seminars while the actual governance of #labour_migration proceeds through non binding declarations, bilateral memoranda and private recruitment contracts. 11.3 The Sustainable Development Goals The 2030 Agenda gave migration a formal place in development policy. Target 8.8 calls for the protection of labour rights and the promotion of safe and secure working environments for all workers, including migrant workers, in particular women migrants, and those in precarious employment. Target 10.7 calls for facilitating orderly, safe, regular and responsible migration and mobility of people, including through the implementation of planned and well managed migration policies. The #SDGs give ministries of finance and planning a reason to care about migrant workers, which is a different and sometimes more effective lever than the human rights argument alone. Contemporary Pressures 12.1 Domestic work Domestic work is the hardest case for the Convention. The workplace is a private home. Labour inspectors rarely enter it. The worker often lives there, so the employer controls housing, food, movement and communication. The workforce is overwhelmingly female and disproportionately migrant, which means that migration status, gender and class disadvantage compound one another. General Comment No. 1 recognised this, and Article 25 of the Convention requires equality of treatment in conditions of work regardless of status. But equality of treatment with nationals is worth little where national law itself excludes domestic workers from the labour code, which remains the position in many States. Liem and colleagues argue that ratification of International Labour Organization Convention No. 189 is therefore a necessary complement, because it obliges States to bring domestic work inside labour law rather than merely to treat migrant domestic workers the same as excluded national ones. #women_migrant_workers in this sector remain the clearest example of a group for whom formal equality delivers substantive nothing. 12.2 Recruitment and debt The Convention's Article 66 restricts recruitment operations, and Objective 6 of the Global Compact commits States to fair and #ethical_recruitment. Neither has ended the practice of charging workers for their own jobs. Fees are often disguised as training costs, medical examinations, visa processing, or agent commissions, and they are frequently financed by loans secured against family land. A worker who arrives in debt cannot leave a bad employer, cannot refuse unsafe work, and cannot complain. The employer pays principle, endorsed by the International Labour Organization and by a growing number of corporate buyers, is the correct policy response, but it requires enforcement in both the origin and the destination State, which returns us to the reciprocity problem. 12.3 Wage theft #wage_theft is the most common single violation reported by migrant workers, and it is also the least remedied. A worker whose visa is tied to an employer, who is deported when the employment ends, and who has no access to legal aid abroad has, in practice, no way of recovering unpaid wages once they have left the country. Article 22 provides that expulsion shall not prejudice rights already acquired, including wages, and Article 25 forbids derogation by contract. These provisions are correct in principle and almost never enforced. Cross border wage claim mechanisms, portable legal representation, and bilateral agreements that permit claims to continue after return are the practical answers, and they are underdeveloped. 12.4 Detention and criminalisation General Comment No. 5 addressed the growth of immigration #detention. The Convention permits expulsion but demands individual assessment, reasons, review, and compensation where an expulsion is annulled. It does not authorise mass administrative detention, and it does not authorise the detention of children. The criminalisation of irregular entry and stay, and increasingly of those who assist migrants, sits uneasily with the architecture of the treaty, which treats irregular status as an administrative condition rather than a moral one. 12.5 Health and the pandemic legacy The COVID nineteen period was a natural experiment in what happens when migrant workers are excluded from social protection. Workers were housed in dense accommodation, were excluded from national relief schemes, were unable to return home when borders closed, and were blamed for outbreaks. The evidence base on the health consequences is now substantial. Sharma and colleagues document high rates of psychological morbidity among Nepali migrant workers in the Gulf and Malaysia, associated with long hours, hazardous conditions, discrimination and separation from family. Article 28 guarantees emergency care; it does not guarantee inclusion in public health systems, and the pandemic showed why that distinction is dangerous for everyone, not only for migrants. 12.6 Platform work and new forms of employment The Convention's definition of a migrant worker includes those engaged in remunerated activity otherwise than under a contract of employment, which means self employed migrants are covered. This is fortunate, because a growing share of migrant labour in delivery, transport, care and cleaning is now mediated by digital platforms that classify workers as independent contractors. #platform_work raises questions the drafters did not anticipate: who is the employer for the purposes of Article 25, how does an algorithm comply with Article 22, and what does equality of treatment mean when the comparator group of nationals is itself misclassified. This is the most obvious frontier for future research on the treaty. 12.7 Climate related mobility People displaced by drought, flooding, salinity and heat increasingly move across borders as workers rather than as refugees, because there is no other lawful route. #climate_mobility therefore arrives at the door of labour migration law. The Convention does not mention environmental displacement, but its status blind floor is the only widely applicable instrument that protects such people once they are working abroad without authorisation. This under used capacity deserves attention. Critical Assessment 13.1 Strengths The Convention's first strength is its refusal to make protection conditional on status. Part III is a statement that a human being does not become less human by overstaying a visa, and no other treaty says it as clearly. Its second strength is specificity. General human rights treaties speak in abstractions. The Convention speaks about passports, wage slips, work permits, recruitment agencies, and the repatriation of bodies. Specificity is what makes a treaty usable by a labour inspector rather than only by a professor. Its third strength is coverage of family life. The title promises protection to families, and Articles 29, 30, 44 and 45 attempt to deliver it, particularly through the guarantee of education for children irrespective of the parents' status. Its fourth strength is the framework for cooperation in Part VI, which recognises that no single State can regulate a phenomenon that is by definition transnational. 13.2 Weaknesses The first weakness is membership. A treaty without destination States cannot govern destination States. The second is the two tier structure, which reserves the right to form trade unions and the substance of family reunification for documented workers, and therefore withholds the tools of self protection from the workers who most need them. The third is enforcement. Without Articles 76 and 77 in force, there is no case law, no individual remedy, and no incentive for compliance beyond reputational discomfort during a five yearly dialogue. The fourth is the deference to national law. Repeated formulas such as in accordance with applicable legislation, and as they deem appropriate, allow States to define compliance downwards. The fifth is the exclusion of students and trainees, categories that have become major labour migration routes since 1990 and that now function, in several countries, as the principal channel for the recruitment of low paid workers. 13.3 Is the Convention a failure? The honest answer is that it is a partial success measured against an impossible standard. It was designed to regulate a global labour market on the assumption that States would accept binding limits on how they treat non citizens. That assumption was optimistic in 1990 and is more so now. Judged as a binding regime, it has underperformed. Judged as a normative reference point, it has done considerable work, and the vocabulary of migrants as rights holders rather than as units of labour is largely its achievement. Discussion The findings support three propositions. First, the ratification gap is structural, not accidental. It follows from the fact that the benefits of the treaty accrue to non voters, while the costs, real or perceived, fall on governments accountable to voters. Any strategy for increasing ratification that ignores this is wishful. Second, the treaty's influence is broader than its membership, and the appropriate research agenda is therefore about diffusion rather than about counting States. The interesting question is not how many governments have ratified, but which provisions have travelled into national law, court decisions, corporate due diligence standards and bilateral agreements, and why those provisions and not others. Third, the growth of soft law is genuinely double edged. The Global Compact has expanded the number of governments willing to discuss migrant rights. It has also given governments a way to appear engaged without accepting obligation. The Committee's General Comment No. 6 is the correct institutional response: insist on convergence, and refuse the substitution. For students, the practical lesson is that the study of this Convention cannot stop at the text. The text is clear, well drafted and largely uncontroversial as a statement of principle. The difficulty lies entirely in the politics of implementation, and any analysis that does not engage with that politics will be legally accurate and analytically useless. Recommendations 15.1 For States parties Make the declarations under Articles 76 and 77. Ten declarations would activate the individual complaints mechanism and give the Committee the case law it needs. This does not depend on destination States and could be done within a single reporting cycle. Report on time, and report on practice rather than only on law. Include disaggregated data by sex, sector, nationality and status. Ratify International Labour Organization Conventions No. 97, No. 143 and No. 189 alongside the Convention, so that the human rights frame and the labour standards frame reinforce each other. 15.2 For destination States that have not ratified Consider partial engagement. A State may align domestic law with specific provisions, such as Article 21 on identity documents, Article 22 on collective expulsion, Article 25 on equality in conditions of work, and Article 30 on access to education, without ratifying the treaty as a whole. This is intellectually untidy, but it protects people, and protecting people is the point. Recognise that Article 79 preserves control over admission. The most common objection to ratification, that it would surrender border control, is simply not supported by the text. 15.3 For origin States Coordinate. The single greatest structural weakness of origin States is that they compete with one another to supply cheap labour, and competition drives down the standards each is willing to demand. Regional alliances among sending States, of the kind Mustaqeem proposes for South Asia, would convert a race to the bottom into collective bargaining. Negotiate #bilateral_labour_agreements that incorporate Convention standards by reference, that provide for #portable_benefits and for the continuation of wage claims after return, and that are published rather than kept confidential. 15.4 For the Committee and the United Nations system Continue the strategy of convergence with the Global Compact, and use the review cycles of the Compact as an entry point into non ratifying States. Invest in follow up procedures, and publish accessible summaries of concluding observations in the main languages of migrant communities, not only in the working languages of the United Nations. 15.5 For universities and students Teach the Convention. Its near absence from mainstream international law curricula is one reason its provisions are so rarely argued in court. Encourage empirical dissertations that test whether ratification changes anything, and that follow specific provisions into national systems. Limitations and Future Research This article is doctrinal and secondary. It does not test its claims against primary empirical data, and its conclusions about diffusion are therefore hypotheses rather than findings. The literature consulted is predominantly in English, which biases the account towards the perspectives of scholars in the Global North writing about States in the Global South. #data_gaps are severe throughout this field: reliable, comparable statistics on wage theft, recruitment costs, workplace deaths, and access to justice for migrant workers do not exist for most corridors, and this absence is not accidental. Four research agendas follow. First, comparative studies of what changes in domestic law and administrative practice after ratification, using matched pairs of ratifying and non ratifying States. Second, provision level diffusion studies, tracing individual articles into national statutes, court decisions and corporate standards. Third, empirical work on the effect of the Global Compact review cycles on the behaviour of non ratifying destination States. Fourth, doctrinal work on the application of the Convention to platform mediated and algorithmically managed labour, an area where the 1990 text is silent and the 2020s reality is loud. Conclusion The International Convention on the Protection of the Rights of All Migrant Workers and Members of Their Families is the most complete legal statement the international community has produced on the rights of people who work outside their own country. It says clearly that migrant workers are workers and human beings before they are immigration cases, that the loss of legal status does not extinguish the right to be paid, to be safe, to see a doctor in an emergency, and to send one's children to school, and that the governance of labour migration is a shared duty of origin, transit and destination States. It also remains the least ratified of the core human rights treaties, unsupported by a single major destination State, with complaint procedures that have never entered into force after more than two decades, and with an enforcement record that consists largely of recommendations addressed to governments that are already the most sympathetic. Both statements are true at once, and the tension between them is the subject of this article. The Convention has not failed; it has been declined. The distinction matters, because a failed instrument should be replaced, while a declined one should be pressed. The most useful contribution students and researchers can make is not to mourn the ratification gap but to work out, provision by provision and corridor by corridor, how the standards in this treaty can be made to bite in the places where they are needed. That work is legal, empirical, political and slow. It is also, given that around 169 million people are currently working outside the country where they were born, among the more urgent tasks in contemporary international law. Hashtags #ICMW_1990 #Migrant_Workers_Convention #ICRMW #UN_Treaty_Bodies #International_Human_Rights_Law #Labour_Migration_Governance #Rights_of_Migrant_Families #Irregular_Migrants_Rights #Committee_on_Migrant_Workers_CMW #Global_Compact_Migration #Decent_Work_for_Migrants #Ratification_Gap #Migration_and_Development #Protecting_Migrant_Families #Migration_Law_Research References Atak, I., Nakache, D., Crepeau, F., and colleagues (2023). Reviewing the reviews: the global compacts' added value in access to asylum procedures and immigration detention. Frontiers in Human Dynamics, 5. doi:10.3389/fhumd.2023.1264942 Bast, J., Endres de Oliveira, P., and Wessels, J. (2023). Enhancing the rights of protection seeking migrants through the Global Compact for Migration: the case of EU asylum policy. International Journal of Refugee Law, 35(4), 461 to 474. doi:10.1093/ijrl/eead007 Burmeister Rudolph, M. (2024). Reproducing and reformulating categories of skill in regional and global migration governance: evidence from India and Bangladesh. Ethnic and Racial Studies. doi:10.1080/01419870.2024.2404485 Committee on the Protection of the Rights of All Migrant Workers and Members of Their Families (2021). General comment No. 5 (2021) on migrants' rights to liberty, freedom from arbitrary detention and their connection with other human rights. United Nations, Geneva. Committee on the Protection of the Rights of All Migrant Workers and Members of Their Families (2024). General comment No. 6 (2024) on the convergent protection of the rights of migrant workers and members of their families through the Convention and the Global Compact for Safe, Orderly and Regular Migration. United Nations, Geneva. Desmond, A. (2020). A new dawn for the human rights of international migrants? Protection of migrants' rights in light of the UN's SDGs and Global Compact for Migration. International Journal of Law in Context, 16(3), 222 to 238. doi:10.1017/S1744552320000191 Farahat, A., and Bast, J. (2022). A global view on the Global Compact for Migration: introduction. Verfassung und Recht in Ubersee (World Comparative Law), 55(1), 3 to 12. International Labour Organization (2021). ILO global estimates on international migrant workers: results and methodology, third edition. International Labour Office, Geneva. International Labour Organization (2022). Global estimates of modern slavery: forced labour and forced marriage. International Labour Office, Geneva. International Organization for Migration (2024). World migration report 2024. International Organization for Migration, Geneva. Jouzier, B. (2026). Analysing the impact of the GCM in Canada through the lens of mobility facilitation: an invitation to rethink the GCM's potential to advance the human rights of migrants. Laws, 15(2), 18. Kalush, B., and Saraswathi, V. (2024). Reforms without rights: the GCC states' blinkered view of labour reforms. Gulf Labour Markets, Migration and Population Programme, Explanatory Note No. 2, Gulf Research Center, Jeddah. Laikram, S., and Pathak, S. (2022). Ratification of ICRMW toward the ILO conventions amid COVID nineteen in Thailand. Emerging Science Journal, 6 (Special Issue on COVID nineteen: Emerging Research), 133 to 148. doi:10.28991/esj-2022-SPER-010 Liem, A., and colleagues (2024). Securing the rights and health of domestic workers: the importance of ratifying the ILO's C189. Globalization and Health, 20. doi:10.1186/s12992-024-01065-5 Mustaqeem, M. (2025). South Asian migrant workers and the challenges of Gulf migration regimes: advocating for a regional alliance. Journal of World Affairs, 1(2), 246 to 268. doi:10.1177/29769442251350945 Panizzon, M., Vitiello, D., and Molnar, T. (2022). The rule of law and human mobility in the age of the global compacts: relativizing the risks and gains of soft normativity? Laws, 11(6), 89. doi:10.3390/laws11060089 Sharma, A., Adhikari, R., Parajuli, E., Buda, M., Raut, J., Gautam, E., and Adhikari, B. (2023). Psychological morbidities among Nepalese migrant workers to Gulf and Malaysia. PLOS ONE, 18(11). doi:10.1371/journal.pone.0267784 United Nations Economic and Social Commission for Asia and the Pacific (2024). Asia Pacific migration report 2024: assessing implementation of the Global Compact for Migration. United Nations, Bangkok. United Nations Network on Migration (2022). Implementing the Global Compact for Safe, Orderly and Regular Migration: guidance for governments and all relevant stakeholders. United Nations, Geneva. United Nations Office of the High Commissioner for Human Rights (2021). The international convention on migrant workers and its committee, fact sheet No. 24, revision 1. United Nations, Geneva.