Israeli control of the West Bank and East Jerusalem has run continuously since June 1967, 59 years to 2026, per the US Department of State’s Office of the Historian. That duration is the spine of the apartheid allegations: the argument is that what began as military administration settled into two legal and territorial regimes over one piece of ground, one for Israeli settlers and one for Palestinians. The law of occupation itself contemplates separate legal machinery, and Article 64 of the Fourth Geneva Convention (1949) says so in terms; the 1973 Apartheid Convention adds an element the occupation texts do not carry, the purpose of maintaining domination by one racial group over another. That gap, between conduct that is documented and purpose that is inferred, is where the published determinations and their named critics divide.
Key Findings
- Israeli control of the West Bank and East Jerusalem has run continuously since June 1967, per the US Department of State’s Office of the Historian, 59 years to 2026.
- Article 64 of the Fourth Geneva Convention (1949) permits an occupying power to keep local penal law in force while subjecting the population to its own provisions and trying offenders in its own military courts; Article II of the 1973 Apartheid Convention adds what the occupation texts do not contain, a purpose of establishing and maintaining domination by one racial group over another.
- Israel’s Supreme Court, sitting as the High Court of Justice with an eleven-justice panel, dismissed the petitions against Basic Law: Israel as the Nation-State of the Jewish People on 8 July 2021 in HCJ 5555/18 Hasson v. Knesset, while holding that the Knesset may not deny the state’s core character as Jewish and democratic.
- The Israel Central Bureau of Statistics counted Jerusalem past one million residents at the end of 2023, 60.5% Jews and Others and 39.5% Arabs, in a single municipal figure that includes the East Jerusalem neighbourhoods annexed after 1967.
- 5.2% of the UK antisemitic comments coded by Becker, Ascone and Troschke (2022) used the apartheid analogy, against 39.8% invoking an “evil” stereotype, in 1,504 comments on British media Facebook pages during the May 2021 escalation.
Which occupation features do the apartheid findings rest on?
Four features recur across the published determinations: land taken, settlements built, movement controlled by permit, and separate legal treatment. Human Rights Watch, in its report of 27 April 2021, put the land figure at more than two million dunams confiscated in the West Bank since 1967, roughly 2,000 square kilometres, its own count on its own definitions rather than a government cadastral statistic.
The settler figure most often repeated has a different provenance again. The Office of the UN High Commissioner for Human Rights published a statement in March 2022 by Michael Lynk, then Special Rapporteur on the situation of human rights in the Palestinian territory occupied since 1967, giving 700,000 Israeli settlers in some 300 settlements across the West Bank and East Jerusalem. That is a mandate-holder’s statement, not an enumeration by a statistical office, and it is dated four years back from the present.
Amnesty International, in its report of 1 February 2022, describes a single system of oppression and domination extending across all areas under Israeli control, which is a wider geographic claim than the occupation itself. The International Human Rights Clinic at Harvard Law School, in a March 2022 submission written with Addameer, frames the same material as a legal argument: that the occupation begun in June 1967 set in motion land confiscation, settlement and physical and legal segregation, producing what the submission calls a bifurcated system of citizenship and a dual regime of legal rights.
Counting conventions matter to every one of these totals. The Israel Central Bureau of Statistics counts East Jerusalem residents and West Bank settlers inside Israel’s resident population, as its Rosh Hashana release of September 2025 records in reaching about 10.148 million residents. Monitoring organisations that publish settlement statistics keep the West Bank and the annexed East Jerusalem neighbourhoods as separate series. The two conventions describe the same people and produce non-comparable totals, so a settler count without its convention attached cannot be checked against another. What the treaty texts require before any of this becomes apartheid is a separate question, and it is answered by different documents.
Is the occupation still a temporary military arrangement?
Continuously since June 1967. The US Department of State’s Office of the Historian dates Israeli control of the West Bank, East Jerusalem, the Gaza Strip, the Sinai Peninsula and the Golan Heights to the June 1967 war, and control of the West Bank and East Jerusalem has not lapsed since, a span of 59 years to 2026.
On the calendar, then, the premise holds. The arrangement has outlasted the working life of most of the officials who established it, and any description of it as a short-term military measure is a description the dates do not support. The reframe is what occupation law itself says about duration, which is nothing. Article 43 of the Hague Regulations of 1907 obliges an occupant who has taken authority over territory to take all measures in its power to restore and ensure public order and safety, while respecting, unless absolutely prevented, the laws in force in the country. The provision sets a standard of administration and no term at all; sovereignty is not transferred, and no clock runs.
The Special Rapporteur’s March 2022 statement published by OHCHR frames the same duration as deliberate, describing settlements and barriers as creating facts on the ground that confine Palestinians in disconnected enclaves. That is the mandate-holder’s characterisation of purpose, dated and attributable to him, and it is the step the counter-arguments contest.
The instruments themselves can be laid out and read without any characterisation attached.
Does occupation law itself permit two separate legal systems?
In part, yes, and the text is explicit. Article 64 of the Fourth Geneva Convention (1949) allows the occupying power to keep the penal laws of the occupied territory in force, and to subject the population to provisions it enacts itself, with offenders tried before its own properly constituted, non-political military courts sitting in the occupied country.
Article 43 of the Hague Regulations (1907) sits behind that, requiring the occupant to respect the laws in force unless absolutely prevented. Read together, the two texts contemplate an administration in which the occupied population answers to military jurisdiction and the occupier’s own nationals do not. Separate legal machinery is therefore not, by itself, a departure from the law of occupation. It is what the law of occupation describes.
The Apartheid Convention of 1973 asks something different of the same facts. Article II defines the crime as inhumane acts of the kinds it enumerates committed for the purpose of establishing and maintaining domination by one racial group of persons over any other racial group of persons and systematically oppressing them. The operative words are the purpose and the institutionalised character, not the separateness as such.
That is the seam the dispute runs along. The March 2022 Harvard Law School clinic submission argues that the two systems in the West Bank are not the incidental machinery of belligerent occupation but a bifurcated regime of citizenship and rights, sustained across decades and tied to land transfer. The article published in the Fordham International Law Journal under the title “Not an Apartheid State, a Defamed State” argues the opposite reading of the same arrangement, that the measures are the security-driven administration a belligerent occupation authorises, and that applying the apartheid label to them misdescribes the legal category. A working paper posted to SSRN, “Why Allegations that Israel Is An ‘Apartheid’ State Are False under International Law,” makes the same argument at greater length and is a working paper rather than a peer-reviewed article.
Neither position is settled by the treaty texts alone, because the texts fix the elements and not their application. What Article 64 permits and what Article II prohibits can both be true of a single administration, and the difference between them is a finding about purpose.
Has any court ruled on the occupation’s legality since 1967?
Two advisory opinions and no conviction. The International Court of Justice issued an advisory opinion on 9 July 2004 concerning the separation barrier, and a second on 19 July 2024 on Israel’s policies and practices in the occupied Palestinian territory. Both are advisory opinions rather than contentious judgments between parties, and no court has convicted any party of the crime of apartheid.
Article 7(1)(j) of the Rome Statute (1998) lists apartheid among the crimes against humanity within the jurisdiction of the International Criminal Court, and the Apartheid Convention (1973) has been in force since 1976. In the half-century since, the criminal record is empty of a conviction under either instrument in relation to this territory. That is a fact about the court record, and it settles nothing about the underlying conduct in either direction; an unprosecuted allegation and a rejected one leave the same trace in a docket.
The judicial record that is binding on anyone here is domestic. On 8 July 2021, the Supreme Court of Israel, sitting as the High Court of Justice with an eleven-justice panel, decided HCJ 5555/18 Hasson v. Knesset, dismissing fifteen petitions against Basic Law: Israel as the Nation-State of the Jewish People. The Basic Law was upheld. The Court held at the same time that the Knesset, acting as constituent authority, is in its own words not all powerful, and has no authority, even by Basic Law, to deny or facially contradict the core character of the State of Israel as a Jewish and democratic state. The test it set is narrow: whether an amendment denies the core democratic identity of the state or inflicts mortal harm on the minimal core characteristics of its identity as Jewish and democratic. The Court derived that limit from the Declaration of Independence, the Basic Laws, Knesset legislation and its own case law.
Both halves of that judgment belong on the page. The law stands, and the country’s highest court recorded a constitutional line the law could have crossed and, in the Court’s finding, did not. The 2018 statute the case concerned is examined clause by clause in the report on the Israeli laws called Jewish supremacist.
How does Israel’s own record treat the territories it holds?
Jerusalem’s residents passed one million at the end of 2023, and the Israel Central Bureau of Statistics counted them as 60.5% Jews and Others and 39.5% Arabs, a single municipal population that includes the East Jerusalem neighbourhoods annexed after 1967. That is the Bureau’s own preliminary count, and the composition of the denominator is the point.
The enacted texts run the same way. Basic Law: Jerusalem, Capital of Israel (1980) declares Jerusalem, complete and united, the capital of Israel. Clause 3 of Basic Law: Israel as the Nation-State of the Jewish People (2018) repeats it in the Knesset’s own English text: “The complete and united Jerusalem is the capital of Israel.” Clause 7 of the same law reads: “The State views the development of Jewish settlement as a national value, and shall act to encourage and promote its establishment and consolidation.” Clause 1(c) reads: “The realization of the right to national self-determination in the State of Israel is exclusive to the Jewish People.” The Knesset labels its English translation unofficial.
The statistical practice has a documented history of its own. Hirsch-Hoefler and Ben Shitrit (2020), in a POMEPS report that is a think-tank publication rather than a peer-reviewed study, trace the Central Bureau of Statistics yearbook from 1967 forward: East Jerusalem entered the general population count for Israel proper in the 1968 yearbook; a 1970 footnote folded settlers into Israel’s total as “Israeli residents in the Administrated Territories”; by 1983 settlement localities were counted within the total number of Israeli localities; and in 1997 the separate territories chapter was removed. The authors call this the normalization of settlements, which is their analytic framing and travels with their names attached. They also record that the Bureau excludes unrecognised outposts, and that a Green Line straddling locality can appear in Israel’s South District in the official series while monitoring organisations count it as a settlement.
None of this is a finding about how anyone is treated. It is what a state’s own enacted texts and its own statistical office record about the territory it holds, which is the evidence the argument is being made about.
Can a long occupation become apartheid, and who disputes it?
The named positions divide on that question, and they were published within months of each other. The Harvard Law School International Human Rights Clinic, in its March 2022 submission with Addameer, argues that the occupation begun in June 1967 set in motion a deliberate policy of land confiscation, dispossession, settlement and legal segregation, and that duration and structure together carried the arrangement across the threshold.
The counter-position holds that duration changes the political problem and not the legal category. The Fordham International Law Journal article “Not an Apartheid State, a Defamed State” argues that the West Bank arrangement is a belligerent occupation with a security rationale, and that the apartheid designation imports a racial-domination framework that the facts of a territorial and national conflict do not fit. The SSRN working paper “Why Allegations that Israel Is An ‘Apartheid’ State Are False under International Law” advances the same conclusion from the treaty elements and, as a working paper, has not been through peer review.
The Institute for National Security Studies, in its 2017 volume “Six Days, Fifty Years: The June 1967 War and its Aftermath,” treats the post-1967 arrangement as an unresolved territorial dispute fifty years on, framing the West Bank question as one of borders, security and political settlement rather than of civil status within a single polity. Published at the fiftieth anniversary, it is a research-institute volume from an Israeli institute, and its framing is the one most often set against the single-regime reading.
The two readings do not disagree much about the physical record. Both accept that settlements exist, that Israeli civilians in the West Bank and Palestinian residents of the West Bank are governed by different legal machinery, and that the arrangement has lasted since 1967. They disagree about what the machinery is for. One side reads permanence and pattern as evidence of a purpose; the other reads the same permanence as an unresolved conflict that has not been settled by negotiation, with the legal machinery following from occupation law rather than from a design to dominate.
Because the disagreement is about purpose rather than about the physical record, it cannot be closed by adding more description of the physical record. That is why the treaty elements, and specifically what would count as proof of purpose, do the deciding.
What would have to be shown to prove a domination purpose?
Three elements, per Article 7(2)(h) of the Rome Statute (1998): inhumane acts of a character similar to those listed in Article 7(1), committed in the context of an institutionalised regime of systematic oppression and domination by one racial group over any other racial group or groups, and committed with the intention of maintaining that regime.
The 1973 Apartheid Convention states the same structure in Article II: the enumerated inhumane acts must be committed for the purpose of establishing and maintaining domination by one racial group of persons over any other and systematically oppressing them. In both instruments the conduct and the purpose are separate elements, and the second is not established by proving the first.
The conduct element is where the documentary record is thickest, because acts leave records: land declarations, building permits, court files, statistical series, population counts. The purpose element leaves fewer traces of its own kind. Human Rights Watch, in April 2021, and Amnesty International, in February 2022, reach the purpose element by inference, reading a pattern of policy sustained over decades as evidence of the intention behind it. That is a reasoning method rather than a measurement, and each organisation states it as its own determination on its own application of the treaty definitions.
The named critics answer at the same element. The Fordham International Law Journal article and the SSRN working paper hold that the pattern is explained by a security rationale within belligerent occupation, and that a pattern consistent with two explanations does not establish either one. Article 64 of the Fourth Geneva Convention (1949) gives that argument its footing, because the separate military jurisdiction the critics point to is provided for by the law of occupation itself.
Israel’s own highest court has spoken to the adjacent constitutional question rather than to this one. In HCJ 5555/18 (8 July 2021) the Supreme Court upheld Basic Law: Israel as the Nation-State of the Jewish People while holding that the Knesset may not deny the state’s core identity as Jewish and democratic. That is a ruling on the constitutional limits of Israeli legislation, not a finding under the Apartheid Convention, and no ruling of the second kind exists in the record.
How does occupation vocabulary appear in measured antisemitic content?
5.2%. Of the antisemitic comments coded in the United Kingdom sample by Becker, Ascone and Troschke (2022, peer-reviewed), 5.2% used the apartheid analogy, in a study of 4,500 comments on the Facebook pages of leading British, French and German media outlets reacting to the May 2021 escalation of the Arab-Israeli conflict.
The study coded against a scheme built on the working definition of antisemitism adopted by the International Holocaust Remembrance Alliance. 26.9% of the 1,504 UK comments analysed were coded antisemitic, against 12.6% of 1,500 French comments and 13.6% of 1,520 German comments. Within the UK antisemitic comments the apartheid analogy was one of the smaller categories, and the authors record that thread context was needed to infer the antisemitic meaning in 38.7% of them, so the coding depends on material beyond the comment itself.
Survey research measures the adjacent association. Harber, Bulska, Malloy and Vila (2026, peer-reviewed, abstract-sourced) report across three preregistered studies (N = 373, N = 243, N = 493) that conspiracy beliefs implicating Israel and Zionists mediated the association between antisemitism and anti-Israel attitudes, with their model accounting for over 55% of the variance in anti-Israel attitudes, while conspiracy beliefs about Jews unrelated to Israel did not mediate it and a general conspiratorial mindset did not explain the result. Ozer, Obaidi and Bergh (2025, peer-reviewed, abstract-sourced) ran two experiments with British samples (N = 354 and N = 490) priming participants with war-in-Gaza material, and report two separate pathways to anti-Jewish hostility, one running through social dominance orientation and one through out-group identity fusion with the Palestinian people.
These are associations measured in samples, not statements about any individual who uses the term, and the boundary they sit on is examined separately in the report on anti-Zionism and antisemitism.
Methodology and limitations
This report assembles four kinds of evidence and labels each where it is used: treaty texts as adopted (the Hague Regulations of 1907, the Fourth Geneva Convention of 1949, the Apartheid Convention of 1973, the Rome Statute of 1998); official records (the US Department of State’s Office of the Historian on June 1967, the Knesset’s published Basic Laws, the Israel Central Bureau of Statistics releases, and the judgment in HCJ 5555/18); scholarly and institutional legal analysis (the Harvard Law School clinic submission of March 2022, the Fordham International Law Journal article, an SSRN working paper, the 2017 Institute for National Security Studies volume, and the 2020 POMEPS report by Hirsch-Hoefler and Ben Shitrit); and peer-reviewed measurement of discourse and attitudes (Becker, Ascone and Troschke, 2022; Harber and colleagues, 2026; Ozer and colleagues, 2025).
Where an organisation’s determination is itself the circulating claim, it is named, dated and attributed as that body’s own count on its own definitions: the more-than-two-million-dunam figure is Human Rights Watch’s own (April 2021), the 700,000 settlers and 300 settlements figure comes from the March 2022 OHCHR statement of the Special Rapporteur and is a mandate-holder’s statement rather than an enumeration by a statistical office, and the single-system characterisation is Amnesty International’s (February 2022). None of these is a government or census figure, and the settler figure is four years old.
Several limitations bear on the reading. The 2004 and 2024 ICJ opinions are advisory, not contentious judgments, and confer no criminal finding. Population and settler counts are governed by conventions that are not interchangeable: the Israel Central Bureau of Statistics counts East Jerusalem residents and West Bank settlers within Israel’s resident total, while monitoring organisations keep them as separate series, so totals from the two conventions must not be compared or combined. The Hirsch-Hoefler and Ben Shitrit report is not peer-reviewed and its normalization framing is the authors’ own. The Harber and Ozer findings are drawn from published abstracts and are correlational, from samples in the United States and Britain respectively. The Becker study covers one platform, three countries and one escalation window in May 2021, and its shares are of coded antisemitic comments within that sample, not of public discourse generally.
Conclusion
Does duration turn an occupation into something else? On the record set out here, 59 years of continuous control since June 1967 changes what has to be explained, and it does not by itself change the legal category. Article 64 of the Fourth Geneva Convention contemplates the separate military jurisdiction that the argument points to; Article II of the 1973 Apartheid Convention and Article 7(2)(h) of the Rome Statute add a purpose element that occupation law never contained, and the entire dispute lives in that added element.
What the record fixes is substantial and one-sided in neither direction. The physical arrangement is not in dispute between the named parties: settlements exist, two legal regimes operate over one territory, and Israel’s own statistical office has counted East Jerusalem inside Jerusalem’s municipal total since the 1968 yearbook, reporting 60.5% Jews and Others against 39.5% Arabs at the end of 2023. The enacted texts say what they say, including Clause 7 of the 2018 Basic Law on Jewish settlement as a national value and Clause 1(c) on exclusive national self-determination. Israel’s own Supreme Court upheld that law in July 2021 while setting a constitutional limit it could have breached. And no court has convicted any party of the crime of apartheid in the 50 years the 1973 Convention has been open. Human Rights Watch and Amnesty International reach the purpose element by inference from the pattern; the Fordham article and the SSRN paper read the same pattern as belligerent occupation with a security rationale. Two readings, one physical record, no adjudicated finding between them.
Then there is the vocabulary. The apartheid framing appeared in 5.2% of the UK antisemitic comments Becker, Ascone and Troschke coded in May 2021, alongside an “evil” stereotype at 39.8% and Israel’s sole guilt at 27.9%, and Harber and colleagues measured conspiracy beliefs implicating Israel and Zionists carrying the association between antisemitism and anti-Israel attitudes. A word can be a legal term of art in one venue and a carrier in another, and the same three syllables travel through both. Which of those two lives the framing is mostly leading, when it moves from a treaty citation into a comment thread, is a question this record poses and does not close.
Frequently Asked Questions
What is a dunam, and how large is two million of them?
A dunam is about 1,000 square metres, so the more-than-two-million-dunam figure Human Rights Watch published in April 2021 works out to roughly 2,000 square kilometres. That is the organisation’s own count on its own definitions, not a figure drawn from a government land registry.
Which treaty texts define apartheid as a crime?
Two. The International Convention on the Suppression and Punishment of the Crime of Apartheid (1973) defines it in Article II, and the Rome Statute of the International Criminal Court (1998) lists it among crimes against humanity in Article 7(1)(j) and defines it in Article 7(2)(h). Both require an institutionalised regime and a purpose of maintaining domination, not discriminatory treatment alone.
Is an International Court of Justice advisory opinion binding?
No. An advisory opinion is issued at the request of an authorised body and does not decide a case between parties or impose obligations the way a contentious judgment does. The Court issued advisory opinions on 9 July 2004 and 19 July 2024 touching on Israeli policies in the occupied territory.
Why do published settler counts differ so much between sources?
Because the counting conventions differ. The Israel Central Bureau of Statistics counts West Bank settlers and East Jerusalem residents inside Israel’s resident population, reaching about 10.148 million in its September 2025 release, while monitoring organisations report the West Bank and the annexed East Jerusalem neighbourhoods as separate series. Totals built on one convention cannot be compared with totals built on the other.
What does occupation law require of an occupying power?
Article 43 of the Hague Regulations (1907) requires the occupant to take all measures in its power to restore and ensure public order and safety, while respecting, unless absolutely prevented, the laws in force in the country. Sovereignty is not transferred by occupation, and the text sets no time limit.
Does research measure a link between attitudes to Israel and antisemitism?
Allington, Hirsh and Katz (2023, peer-reviewed) surveyed 1,790 UK adults in December 2021 and found coronavirus conspiracy suspicion correlated with all measured forms of antisemitism, with the correlation strongest for attitudes toward Jews as Jews and the anti-Zionist link accounted for by its overlap with that measure. These are correlations in a quota sample, not statements about any individual’s views.
Sources
- Allington, D., Hirsh, D., and Katz, L., 2023. Correlation Between Coronavirus Conspiracism and Antisemitism: A Cross-Sectional Study in the United Kingdom. Scientific Reports. DOI 10.1038/s41598-023-41794-y. Peer-reviewed.
- Becker, M., Ascone, L., and Troschke, H., 2022. Antisemitic Comments on Facebook Pages of Leading British, French, and German Media Outlets. Humanities and Social Sciences Communications. DOI 10.1057/s41599-022-01337-8. Peer-reviewed.
- Fordham International Law Journal, n.d. Not an Apartheid State, a Defamed State. ir.lawnet.fordham.edu. Law review article.
- Harber, K., Bulska, D., Malloy, T., and Vila, J., 2026. Antisemitism, Conspiracy Beliefs, and Anti-Israel Attitudes. American Psychologist. DOI 10.1037/amp0001635. Peer-reviewed (abstract-sourced).
- Harvard Law School International Human Rights Clinic and Addameer, 2022. Submission on apartheid in the West Bank. hrp.law.harvard.edu. University legal clinic analysis.
- Hirsch-Hoefler, S., and Ben Shitrit, L., 2020. So, how many settlements are there? Counting, tracking, and normalizing Jewish settlements in the Israeli Central Bureau of Statistics Yearbook, 1967 to the present. POMEPS. Think-tank report, not peer-reviewed.
- Institute for National Security Studies, 2017. Six Days, Fifty Years: The June 1967 War and its Aftermath. inss.org.il. Research institute volume.
- International Committee of the Red Cross, 1949. Geneva Convention (IV) relative to the Protection of Civilian Persons in Time of War, Article 64. Treaty text. Official record.
- International Peace Conference, 1907. Convention (IV) respecting the Laws and Customs of War on Land, Annex, Article 43 (Hague Regulations). Treaty text. Official record.
- Israel Central Bureau of Statistics, 2024. Selected Data on the Occasion of Jerusalem Day, 2024, Media Release 165/2024. cbs.gov.il. Official record.
- Israel Central Bureau of Statistics, 2025. Population of Israel on the Eve of Rosh Hashana 5786. cbs.gov.il. Official record.
- Knesset, 1980. Basic Law: Jerusalem, Capital of Israel. main.knesset.gov.il. Official record (enacted legislation).
- Knesset, 2018. Basic Law: Israel as the Nation-State of the Jewish People, unofficial English translation. main.knesset.gov.il. Official record (enacted legislation).
- Ozer, S., Obaidi, M., and Bergh, R., 2025. The Impact of Globalized Conflicts: Examining Attitudes Toward Jews Among Britons in the Political Context of the War in Gaza. International Journal of Intercultural Relations, vol. 107. DOI 10.1016/j.ijintrel.2025.102184. Peer-reviewed (abstract-sourced).
- SSRN working paper. Why Allegations that Israel Is An “Apartheid” State Are False under International Law. papers.ssrn.com, abstract 4343950. Working paper, not peer-reviewed.
- Supreme Court of Israel sitting as High Court of Justice, 2021. HCJ 5555/18 Hasson v. Knesset, judgment of 8 July 2021. Official record (court judgment).
- United Nations General Assembly, 1973. International Convention on the Suppression and Punishment of the Crime of Apartheid, Article II. Treaty text. Official record.
- United Nations Diplomatic Conference, 1998. Rome Statute of the International Criminal Court, Article 7. Treaty text. Official record.
- US Department of State, Office of the Historian, n.d. The 1967 Arab-Israeli War. history.state.gov. Official record.