Those who call the Gaza blockade collective punishment, Amnesty International among them since 2008, rest the label on restrictions that reach the Strip’s whole population. The restricting party’s own record documents one such measure: COGAT, an Israeli Ministry of Defense unit, states that “no aid entered the Gaza Strip between March 2 to May 19 2025” under a “directive from the political echelon.” The population carrying that burden is measured: the Site Management Cluster counted an estimated 1,708,822 people in 1,591 displacement sites in May 2026. The legal record runs on different tests. Israel’s Supreme Court tied the term “collective punishment” to the disproportionate use of a punitive demolition power in a 2014 judgment, and Israel’s Foreign Ministry states that compliance with the law of armed conflict is “conduct-oriented, not result-oriented.”
Key Findings
- COGAT, the Israeli Ministry of Defense unit that coordinates entry into Gaza, records on its Aid Data dashboard that “no aid entered the Gaza Strip between March 2 to May 19 2025” by “directive from the political echelon,” a party’s own published record (COGAT, 2026).
- The Site Management Cluster estimated 1,708,822 people living in 1,591 verified displacement sites on data extracted 16 May 2026, a figure it describes as a social count rather than a census (Site Management Cluster, 2026).
- COGAT’s own truck-level record counts 32,946 trucks between 21 October 2023 and 31 May 2024, ranging from 9 trucks on 8 May 2024 to 467 on 9 April 2024; the totals are the Institute’s arithmetic on Israel’s own published rows (COGAT, 2026).
- Israel’s Supreme Court held on 31 December 2014, in a case on punitive house demolition, that disproportionate use of the power “amounts to collective punishment that is prohibited under international law” (Israeli Supreme Court, 2014).
- Israel’s Ministry of Foreign Affairs states, in a legal paper updated to 2 November 2023, that compliance “is conduct-oriented, not result-oriented,” a party’s statement of its own legal position (Israel Ministry of Foreign Affairs, 2023).
Why is the Gaza blockade called collective punishment?
COGAT, the Israeli Ministry of Defense unit that coordinates entry into Gaza, records on its Aid Data dashboard that “no aid entered the Gaza Strip between March 2 to May 19 2025,” citing “the directive from the political echelon.” The collective-punishment label rests on measures of this kind: applied to everyone in the Strip at once, with no individual criterion attached. The line is a party’s own published record, read on 10 September 2026.
The full sentence on the dashboard reads: “According to the directive from the political echelon, no aid entered the Gaza Strip between March 2 to May 19 2025.” COGAT gives the period by those two dates and attaches to it no criterion distinguishing one resident of the Strip from another. The dashboard is a live display stamped “Data Last Updated: 9/10/2026 2:43:42 PM,” and it carries a date-range selector, so its numerical panels describe whichever range was selected when they were read; the quoted line is the dashboard’s own text.
The characterisation turns on two separate questions. The first is reach. Article 33 of the Fourth Geneva Convention of 1949 prohibits collective penalties against protected persons, and the dashboard line describes a measure applied to the Strip as a whole. That strand turns on who bears a measure: the whole population rather than the individuals alleged to have committed an offence.
The second strand is purpose: whether a measure reaching the whole population was imposed to pressure or penalise that population. That strand turns on an inference about intent, a different question from reach. COGAT’s dashboard attributes the 2025 nil-entry period to “the directive from the political echelon” and states no purpose for it, so the record on purpose is thinner than the record on reach, and Israel’s own courts have their own method for testing a stated purpose.
The treaty text and the determinations issued on aid restrictions as collective punishment are set out at length in an earlier Institute report. The two strands matter separately because the evidence for each is of a different kind: reach is a matter of counts of people and supplies, and purpose is a matter of what drove a decision.
Who calls the blockade collective punishment, and since when?
The Site Management Cluster, the humanitarian coordination body for Gaza’s displacement sites, counted an estimated 1,708,822 people living in displacement sites on data extracted 16 May 2026. That is the kind of population-wide burden that named holders of the collective-punishment label have described since at least 2008. The Cluster’s count is an estimate built from site enumeration, not a census.
The holders, in date order, are named bodies with dated statements. Amnesty International described the blockade as collective punishment in a 2008 document titled “Israel/Occupied Palestinian Territories: Gaza blockade.” In October 2023 Amnesty International called for Israel to lift an “illegal and inhumane blockade on Gaza.” Each is a position held by its author, on its author’s own definitions, and neither is a measurement of the burden it describes.
The factual premise beneath those positions, that the burden falls across the population, is measured. The Site Management Cluster (2026) verified 1,591 active sites across all five governorates, with fieldwork from 3 February to 19 April 2026 carried out by 165 enumerators from twelve partner organisations. It classed 83% of those sites as makeshift, meaning informal, open-air and tented. For scale against the pre-war population, Crawford (2025), a Brown University Costs of War paper that is not peer-reviewed, relays the Palestinian Central Bureau of Statistics estimate of approximately 2.2 million residents in July 2023. The two figures carry different dates and different definitions, so no share is computed from them.
Three limits travel with the count. The 1,708,822 counts people in sites at a stated date, not everyone displaced at any point since October 2023. The Cluster’s 354,550 households are derived by applying a standard household size of five where disaggregated data was missing. The Cluster describes the 1,591 as “a social count,” a measure of organised displaced communities reached. The count records where people live; it carries no finding about the purpose of any restriction, which is the second strand of the characterisation and the subject of the next two sections. The Institute’s reading of the blockade and human dignity sets these enumerations against the published legal standards.
What has Israel’s own Supreme Court said collective punishment is?
Israel’s Supreme Court, in HCJ 8091/14 delivered on 31 December 2014, a case on punitive house demolition under Regulation 119 rather than on the Gaza blockade, held that “the disproportionate use of said authority by the military commander, which amounts to collective punishment that is prohibited under international law, is precluded.” In that judgment the court ties the term to disproportion in how a power is used, not to the existence of the power.
The petition, brought by eight organisations, sought a declaration that forfeiting, demolishing and sealing the homes of persons suspected of hostile activity under Regulation 119 of the Defence (Emergency) Regulations, 1945, is unlawful. Justice E. Rubinstein wrote the judgment, with Justices E. Hayut and N. Sohlberg concurring, and the disposition reads: “Subject to Paragraph 27, we cannot grant the Petition.” The English text is the Cardozo Israeli Supreme Court Project’s translation; the Hebrew governs.
Paragraph 27 is the limit attached to that denial. The court wrote that “the principle of proportionality does not allow us to continue to assume forever that choosing the drastic option of house demolition, or even of house sealing, achieves the desired purpose of deterrence, unless all of the data that properly confirms that hypothesis is presented to us for our review,” and held that state authorities must examine the measure’s utility from time to time, including through follow-up research.
Justice Hayut added a qualification about the people who bear the measure: where family members whose home is to be demolished “can convince, by means of sufficient administrative evidence, that prior to the terrorist attack they tried to dissuade the terrorist from carrying it out, that factor should be given very significant weight, which may, in suitable cases, rule out a decision to demolish.”
The judgment carries one count of use, and it is the State’s own, given as a respondent in the case: “since 2005, the military commander has exercised the contemplated authority only several times.” That count measures orders and buildings, not people. The court also quoted the security establishment’s own 2005 committee presentation, which found that “within the context of deterrence, the measure of demolition is ‘eroded’,” and recorded that systematic demolition for deterrence was to be stopped and reserved for extreme cases.
Both halves of the judgment stand together. The petition failed, and Regulation 119 remains part of Israel’s positive law, in Justice Hayut’s words, by virtue of the Preservation of Laws clause in Basic Law: Human Dignity and Liberty. The same judgment named collective punishment as the line disproportionate use crosses and conditioned continued deference on effectiveness data the State had not yet produced.
How do Israeli courts test the stated purpose of a measure?
Israel’s Supreme Court ruled 5 to 0 on 22 October 1979, in HCJ 390/79 Dweikat, the Elon Moreh case on requisitioning private land for a settlement near Nablus rather than on Gaza, that a measure resting on a stated military need is judged by its dominant purpose. Adopting a test from De Smith’s Judicial Review of Administrative Action, Deputy President M. Landau wrote that “the legality of the act is determined by reference to the dominant purpose.”
The full test, as Landau quoted it, reads: “What was the dominant purpose for which the power was exercised? If the actor pursues two or more purposes where only one is expressly or impliedly permitted, the legality of the act is determined by reference to the dominant purpose.” Applied to the requisition order, Landau found that “the process was inverted: the initiative came from the political echelons, which then reached out to the Chief of the General Staff for his professional opinion,” and that “the military consideration was subordinate to the primary, political decision.”
Justice A. Witkon set the burden of proof: “The law does not give the commander’s assertion that the taking of possession is required for military needs the force of a presumption,” and “We need not be convinced of the sincerity of the consideration, but rather of its correctness.” The court annulled Order of Possession 16/79 as to the petitioners’ plots, about 125 dunams of an order covering about 700, and left a second order on the access road untouched.
The ruling did not bar requisition for settlement in principle. Landau wrote that taking private land for a civilian settlement “is potentially justified under Article 52 of the Hague Regulations” where military needs in fact drove the decision. The petition succeeded on the facts of one requisition. Landau decided the case against the Beit El judgment of 1978, which had upheld a requisition on the same Article 52 reasoning, and distinguished it on the facts rather than overruling it. The same test therefore produced opposite results in two cases a year apart.
What the 1979 judgment held is confined to one requisition order: where the decision’s dominant purpose was political, a stated military need did not sustain it, the burden sat with the authority, and the court asked for correctness rather than sincerity. The judgment concerns land requisition near Nablus and says nothing about the Gaza closure.
Did entries into Gaza stop entirely, or vary by day?
COGAT’s truck-level workbook, Israel’s own count at its own crossings, records 467 trucks on 9 April 2024, the highest single day between 21 October 2023 and 31 May 2024, against 9 trucks on 8 May 2024, the lowest. Across that span entry varied from day to day rather than stopping; the dated period of nil entry sits in 2025, in COGAT’s own dashboard text.
The workbook holds one row per truck: 32,946 rows across 193 operating days inside a 224-day span (COGAT, 2026). Of the 31 dates with no rows, 24 are Saturdays, when the crossings do not operate. On the Institute’s arithmetic, that is 170.7 trucks per operating day or 147.1 per calendar day; COGAT publishes neither rate, and the two differ by the choice of denominator alone.
Food trucks account for 22,281 of the 32,946, or 67.6%. The truck count is a physical tally COGAT makes during inspection; the weight column is headed “Estimated weight (tons)” and takes only eight banded values, so tonnage from this record is a sum of shippers’ declarations, not a measurement.
The same distinction governs the later period. COGAT’s ceasefire situation report (2026) records 1,778,363 metric tons of food entering Gaza between 10 October 2025 and 7 June 2026. That tonnage is COGAT’s own reported figure, based, in its own footnote, on “the self-declaration of humanitarian organizations and private sector providers,” and it does not share the physical basis of the 2024 truck count.
Set end to end, the party’s own records describe three different regimes: a day-level flow between October 2023 and May 2024 that ranged from 9 to 467 trucks, a period from March 2 to May 19 2025 in which the dashboard states no aid entered, and a ceasefire period reported in self-declared tonnage. The truck-level workbook ends on 31 May 2024 although it remained linked on the portal in September 2026. The Institute’s reading of access to humanitarian aid sets these entry counts against the household enumeration.
How many goods crossings did Israel operate before and after October 2023?
COGAT stated on 15 December 2025 that “Prior to October 7th, only two crossings operated for transferring goods into the Gaza Strip,” and listed, as of that date, five inspection points and three transfer and offloading points. Both counts are COGAT’s own account, published by a party to the events, and the pre-war figure is the restricting party’s own baseline for how narrow the pre-war crossing regime was.
The five inspection points COGAT listed are Kerem Shalom, Ashdod Port, Nitzana, Sha’ar Efrayim and Tarkumiya. The three transfer and offloading points are Kerem Shalom for southern Gaza, Kisufim for central Gaza and Zikim for northern Gaza. The 2024 truck-level workbook uses a different label set, and its four crossing values are not four entry points into Gaza.
| Label in COGAT’s record | Trucks, 21 Oct 2023 to 31 May 2024 | Operating window | Source |
|---|---|---|---|
| Kerem Shalom | 18,371 | 12 Dec 2023 to 31 May 2024 | COGAT, 2026 |
| “Nitzana\Rafah” | 13,359 | 21 Oct 2023 to 5 May 2024 | COGAT, 2026 |
| Erez | 1,111 | 17 Apr to 31 May 2024 | COGAT, 2026 |
| JLOTS | 105 | 17 to 24 May 2024 | COGAT, 2026 |
“Nitzana\Rafah” fuses an Israeli inspection point in the Negev with the Egypt-Gaza crossing, so a truck in that row was inspected at one place and entered at another. JLOTS is the temporary United States pier, not an Israeli crossing. The 2024 workbook names Erez for the north, and COGAT’s December 2025 document lists Zikim; no COGAT document equates the two.
The International Court of Justice, in its order of 28 March 2024 in South Africa v. Israel, indicated unanimously, in paragraph 51(2)(a), that the State of Israel shall take all necessary and effective measures to ensure the unhindered provision at scale of humanitarian assistance, “including by increasing the capacity and number of land crossing points and maintaining them open for as long as necessary.” The French text of that order is authoritative and the English is a translation. The order states that it is without prejudice to any findings concerning compliance with the Court’s earlier order, and nothing in it establishes whether any measure was or was not complied with. COGAT’s December 2025 list postdates the order by more than 20 months, and the two documents measure different things: one records what the Court indicated, the other what a party reports it operates.
How does Israel’s government say its conduct in Gaza should be judged?
Israel’s Ministry of Foreign Affairs, in a legal paper updated to 2 November 2023, states that “compliance is conduct-oriented, not result-oriented,” so that civilian harm beyond what was expected “does not necessarily indicate a violation.” The paper, “Hamas-Israel Conflict 2023: Key Legal Aspects,” is a party’s statement of its own legal position on how harm in an attack is to be assessed, not a finding of compliance.
The full sentence reads: “If the attack results in more collateral damage than expected, this does not necessarily indicate a violation: compliance is conduct-oriented, not result-oriented.” The same paper states that “Estimating expected civilian harm can be very challenging, especially when operating in a complex and dense urban environment where civilian presence is dynamic and often hard to ascertain.”
The Ministry’s frequently asked questions, updated to 8 December 2023, apply the same frame to proportionality, which is “to be applied to each and every attack independently,” and illustrate it: “one attack causing 500 casualties should logically, and legally, be seen differently than 500 attacks against military objectives resulting overall in the same number of casualties.” The same document reports Israel’s own counts of evacuation warnings.
Those counts are Israel’s own, cumulative to 8 December 2023, and each is a floor. They cannot be added into a number of warnings or of people warned, since one person may have received several.
Both Foreign Ministry documents address the conduct of attacks and evacuation. Neither sets out a legal basis for the closure regime itself, and both predate the March 2 to May 19 2025 period recorded on COGAT’s dashboard. The two framings therefore measure different objects. The collective-punishment characterisation, as its holders state it, runs on outcomes across a population: who bears a measure and to what effect. Israel’s stated framework runs on decisions taken attack by attack, judged by what was known and expected at the time. The record holds no document in which the two are applied to the same measure.
Methodology and limitations
Evidence was drawn from four source types. Official records of a party to the conflict: COGAT’s “Humanitarian Aid” workbook (one row per truck, 21 October 2023 to 31 May 2024, downloaded and parsed directly), its live Aid Data dashboard read on 10 September 2026, its July 2026 ceasefire situation report, its 15 December 2025 crossings document, and two Israeli Ministry of Foreign Affairs papers from November and December 2023. Every figure from these is that party’s own count or statement and is not independently verified.
Domestic court judgments: two Supreme Court of Israel judgments, HCJ 8091/14 (2014) and HCJ 390/79 (1979), read in the Cardozo Israeli Supreme Court Project’s English translations, with the Hebrew as the governing text. Paragraph numbers are cited because the 2014 PDF’s page numbers are garbled. The 1979 translation misprints its own date; 22 October 1979 is the date that matches its Hebrew calendar date.
A humanitarian coordination body’s field enumeration: the Site Management Cluster’s verification of 1,856 targeted and 1,591 confirmed sites, fieldwork 3 February to 19 April 2026, extraction 16 May 2026. Its population figures are estimates, its household count is derived, and its demographic data cover a subsample of sites. A university research report: Crawford (2025), not peer-reviewed, which compiles the Gaza Ministry of Health’s counts from a United Nations dashboard and relays the Palestinian Central Bureau of Statistics pre-war population; neither figure is a Costs of War measurement.
COGAT’s truck counts are a physical tally at inspection; its tonnages rest on shippers’ self-declaration, and the two are never combined. All sums, shares and per-day rates from the truck record are the Institute’s arithmetic on published rows, not figures COGAT states. Dashboard figures depend on a date filter.
Amnesty International and the International Court of Justice are named as holders of dated positions or as the authors of recorded text and votes; their conclusions are not used as measurement. Whether the closure meets the legal definition of collective punishment is a contested legal characterisation. The two Israeli judgments concern punitive demolition and land requisition; neither concerns the Gaza closure.
Conclusion
So why is the Gaza blockade called collective punishment? Because the measures at issue reach everyone in the Strip at once, and the restricting party’s own record says as much in its own words: COGAT’s dashboard, a party’s own published text, states that no aid entered between March 2 to May 19 2025 by directive of the political echelon.
The record built around that line is specific. In COGAT’s own 2024 truck count, entry swung from 9 trucks on one day to 467 on another; it varied by day rather than stopping, until the 2025 directive. The Site Management Cluster estimated 1,708,822 people in displacement sites in May 2026. Israel’s Supreme Court has held, in 2014, that disproportionate use of a punitive power amounts to collective punishment, and, in 1979, by 5 to 0, that a stated military purpose is tested for correctness rather than sincerity, with the burden on the authority. Israel’s Foreign Ministry, in its own stated position, judges conduct attack by attack, not by result.
Three tests sit in that record: who bears a measure, what purpose drove it, and whether each decision was reasonable when taken. The collective-punishment label turns on the first two; the Foreign Ministry applies the third, and its documents are silent on the closure. When the clearest evidence of reach comes from the restricting party’s own dashboard, and the record on purpose is a directive with no stated reason, which of those tests is a measure aimed at a whole population to answer to?
Frequently Asked Questions
How much fuel entered Gaza between November 2023 and May 2024?
COGAT’s “Fuel” sheet records 335 fuel tanker-trucks and 598 gas tanker-trucks between 1 November 2023 and 31 May 2024, on the Institute’s sum of Israel’s own published entries (COGAT, 2026). Both are floors, because the sheet mixes dashes and empty cells for nil days. The dashboard notes separately that its aid totals “do not include fuel.”
How reliable are COGAT’s published aid tonnages?
COGAT’s July 2026 ceasefire report states that its tonnage “is based on the self-declaration of humanitarian organizations and private sector providers,” and that comparisons at security clearance detect “Only negligible divergence (around 5%)” day to day (COGAT, 2026). That 5% is COGAT’s own assessment of its own process. Pallets are counted at inspection; their contents are not weighed.
How many people in Gaza were sleeping without any shelter in 2026?
The Site Management Cluster estimated 7,700 households, about 38,500 people, sleeping outdoors with no shelter, present in 19% of surveyed sites (Site Management Cluster, 2026). The household figure rests on an assumed household size of five. The same survey recorded an average of about six people per tent, against tents designed for five.
How many people has Gaza’s Ministry of Health reported killed?
The Gaza Ministry of Health reported 67,075 killed and 169,430 injured in the Gaza Strip as of 3 October 2025, figures compiled by Crawford (2025). These are the ministry’s own counts, a party to the conflict, not independently verified. Crawford’s combined 236,505 adds killed and injured and is not a death toll.
Did medical supplies enter Gaza during the ceasefire?
COGAT reports 18,000 tons of medical supplies and medicines entering Gaza since the ceasefire began on 10 October 2025 (COGAT, 2026). The figure is COGAT’s own reported total, a party to the events, and rests on self-declared tonnage. Hospital bed counts in the same report are World Health Organization figures.
How much water did COGAT report reaching Gaza during the ceasefire?
COGAT’s July 2026 situation report states that more than 70,000 cubic meters of water per day was facilitated through external water lines and desalination facilities (COGAT, 2026). The figure excludes local wells. It is COGAT’s own reported figure, not an independent measurement.
Sources
- COGAT (Coordinator of Government Activities in the Territories), Israel Ministry of Defense, 2025. Aid to Gaza: Common Misleading Claims vs. Facts. 15 December 2025. Gaza Aid Data portal, https://gaza-aid-data.gov.il/mainhome/. Official record (a party’s own account).
- COGAT (Coordinator of Government Activities in the Territories), Israel Ministry of Defense, 2026. Gaza Aid Data portal: “Humanitarian Aid” workbook and Aid Data dashboard. https://gaza-aid-data.gov.il/media/d15pw0ii/humanitarian-aid.xlsx and https://gaza-aid-data.gov.il/mainhome/. Read 10 September 2026. Official record (a party’s own operational count).
- COGAT (Coordinator of Government Activities in the Territories), Israel Ministry of Defense, 2026. Humanitarian Situation Report During the Ceasefire, October 2025 to June 2026. 9 July 2026. https://gaza-aid-data.gov.il/media/e2gjueei/humanitarian-situation-report-during-the-ceasefire-cogat-july-2026.pdf. Official record (self-reported).
- Crawford, Neta C., 2025. The Human Toll of the Gaza War: Direct and Indirect Death from 7 October 2023 to 3 October 2025. Costs of War project, Watson School of International and Public Affairs, Brown University. https://costsofwar.watson.brown.edu/paper/HumanTollGaza. University research report, not peer-reviewed.
- Geneva Convention relative to the Protection of Civilian Persons in Time of War (Fourth Geneva Convention), 1949. Article 33. 12 August 1949. https://ihl-databases.icrc.org/en/ihl-treaties/gciv-1949/article-33. Treaty text.
- Israel Ministry of Foreign Affairs, 2023. Hamas-Israel Conflict 2023: Frequently Asked Questions. Updated to 8 December 2023. https://www.gov.il/en/pages/swords-of-iron-faq-6-dec-2023. Official record (a party’s own position and counts).
- Israel Ministry of Foreign Affairs, 2023. Hamas-Israel Conflict 2023: Key Legal Aspects. Updated to 2 November 2023. https://www.gov.il/BlobFolder/generalpage/swords-of-iron-faq-6-dec-2023/en/English_Documents_Hamas-Israel%20Conflict%202023%20-%20Some%20Factual%20and%20Legal%20Aspects%20-%20Israel%20Ministry%20of%20Foreign%20Affairs%20(2%20NOV%202023).pdf. Official record (a party’s statement of its own legal position).
- Israeli Supreme Court, 1979. HCJ 390/79 Dweikat v. Government of Israel (the Elon Moreh case). Judgment of 22 October 1979. English translation, Cardozo Israeli Supreme Court Project, https://versa.cardozo.yu.edu/sites/default/files/upload/opinions/Dweikat%20et%20al.%20v.%20State.pdf. Official record (court judgment, unofficial translation).
- Israeli Supreme Court, 2014. HCJ 8091/14 HaMoked: Center for the Defense of the Individual v. Minister of Defense. Judgment of 31 December 2014. English translation, Cardozo Israeli Supreme Court Project, https://versa.cardozo.yu.edu/opinions/hamoked-center-defense-individual-v-minister-defense. Official record (court judgment, unofficial translation).
- Regulations concerning the Laws and Customs of War on Land, annexed to Hague Convention (IV), 1907. Article 52. 18 October 1907. https://ihl-databases.icrc.org/en/ihl-treaties/hague-conv-iv-1907/regulations-art-52. Treaty text.
- Site Management Cluster in the Occupied Palestinian Territory, 2026. Gaza Displacement Site Verification Assessment: Baseline Findings for May 2026 Humanitarian Planning and Coordination. 22 June 2026. https://www.cccmcluster.org/resources/gaza-displacement-site-verification-assessment-baseline-findings-may-2026-humanitarian. Humanitarian coordination body, primary field enumeration.