As I noted in my first post on the ICJ Palestine AO, the Court has held that the law of occupation, at least to some extent, continues to apply in Gaza. In this post I will examine the Court’s holding and reasoning more closely.
Readers will be aware that the question whether Gaza remained occupied after Israel’s 2005 disengagement has been very controversial. Painting with a very broad brush, many academic experts, states, various international bodies and NGOs, including the ICRC, have argued that Gaza remained occupied, due to Israel’s control of Gaza’s borders and other ways of projecting its power within the territory. Other experts, including many from military backgrounds, have argued that occupation requires boots on the ground and that, after its disengagement, Israel had no effective control over Gaza, with its adversary, Hamas, eventually assuming such control. (For more background, see this post of mine from 15 years ago examining the debate, another post looking at a relevant judgment of the European Court of Human Rights, and this more recent post by Mike Meier over at Articles of War).
While this debate has centred on Gaza, it also raised important questions of principle, especially as to whether occupation is a binary status, that either exists or not, or whether it can instead exist functionally, on a sliding scale, with the quantum of obligations of the occupying power being variable and dependent on the degree of control that it exercises in the territory. This functional approach has echoed, to some extent, the so-called ‘Pictet theory’, which argued that the law of occupation could also apply at the invasion stage, before an occupation was fully established, with no intermediate stage between the two. Similarly, an important question of principle is whether the notion of effective control should be understood differently at the end rather than at the beginning of an occupation, especially in a situation such as Gaza where the occupying power withdraws unilaterally but maintains some measure of influence over the territory. (For more background on the functional approach in particular, see the work of Aeyal Gross, who originated the term, and see also this post explaining how the ICRC adopted and mainstreamed it; see also the proceedings of a 2012 expert meeting on occupation convened by the ICRC, and see also the ICRC new Common Article 2 commentary, reiterating its position).
So, with this background in mind, what has the ICJ decided in its new AO? Briefly, the Court has clearly embraced some variant of a functional approach. But, it also mentions factors that do not necessarily factor into that approach. And while it is clear that the Court’s view is that the law of occupation governs at least some of Israel’s actions, it is not clear what exact obligations the Court regards to be engaged. It is moreover unclear whether the Court actually regards Gaza to have been occupied, fully or partially, and in what precise period after 2005, or rather whether it regards some obligations to apply residually, even after the occupation had ended. In other words, the Court’s approach is quite ambiguous. This ambiguity is again most likely due to the need to maintain unity within the Court.
After providing some background (AO paras 86-89), the Court begins its discussion as follows (para 90):
In these circumstances, the Court must determine whether and how Israel’s withdrawal of its physical military presence on the ground from the Gaza Strip in 2004-2005 affected its obligations under the law of occupation in that area. As the Court observed above (see paragraph 86), territory is occupied when it is actually placed under the authority of the hostile army. A State occupies territory that is not its own when, and to the extent that, it exercises effective control over it. A State therefore cannot be considered an occupying Power unless and until it has placed territory that is not its own under its effective control (see Armed Activities on the Territory of the Congo (Democratic Republic of the Congo v. Uganda), Judgment, I.C.J. Reports 2005, p. 230, para. 173). (emphasis added)
Two points are worth noting here. First, the ‘to the extent’ language clearly implies that the Court regards effective control as a variable criterion, rather than as a binary one. Second, its citation to a paragraph of its own Congo v. Uganda judgment does not support that proposition – on the contrary, that paragraph was widely criticized for holding, or appearing to hold, that an occupying power actually had to establish its authority as a matter of fact, i.e. take over the running of an occupied territory, for an occupation to commence, even though it could deliberately refrain from doing so while fully having the capacity to do so, thus leaving the area in chaos (not, perhaps, unlike the situation in Gaza today), or indeed by leaving the running of the territory to the pre-existing local government (as with Denmark under German occupation during World War II).
Then the Court says (para 91):
Where a State has placed territory under its effective control, it might be in a position to maintain that control and to continue exercising its authority despite the absence of a physical military presence on the ground. Physical military presence in the occupied territory is not indispensable for the exercise by a State of effective control, as long as the State in question has the capacity to enforce its authority, including by making its physical presence felt within a reasonable time (for example, see United States Military Tribunal, USA v. Wilhelm List and others (Hostage case) (19 February 1948), Trials of War Criminals before the Nuremberg Military Tribunals under Control Council Law No. 10, Vol. XI, p. 1243; International Criminal Tribunal for the former Yugoslavia, Prosecutor v. Mladen Naletilić and Vinko Martinović, IT-98-34-T, Trial Chamber, Judgement, 31 March 2003, para. 217).
The Court here is clearly saying that control over an occupied territory can be maintained without boots on the ground. But the cases that it cites – including the notion of the occupying power having the capacity to make its presence felt – dealt with a very specific, and not exactly analogous scenario: one where, due to resistance in the occupied territory, the occupying power does not physically control every atom of that territory at every moment in time, without the occupation terminating and resuming with each flux in control. Thus, for example, when the US and the UK occupied Iraq in 2003, they were the occupying powers in all of Iraq despite the fact that their troops were not present in every single village or could not stop all kinds of insurgent activity. It is unclear whether this approach can be applied to post-disengagement Gaza, bearing in mind the strength and durability of Hamas’ power in the area (see my old post for a discussion of this). And it’s unclear whether the Court is saying here that Gaza was still occupied because Israel could re-enter the territory to fight Hamas, as it has now done – the Court just doesn’t engage with this directly. To my mind, this approach is more relevant for whether Gaza is (wholly or partly) occupied today, with Israeli troops on the ground but not in complete control of every single square metre of Gaza, than as to the question whether it was occupied 2005-2023.
Then the Court says the following (para 92):
The foregoing analysis indicates that, for the purpose of determining whether a territory remains occupied under international law, the decisive criterion is not whether the occupying Power retains its physical military presence in the territory at all times but rather whether its authority “has been established and can be exercised” (Article 42 of the Regulations Respecting the Laws and Customs of War on Land annexed to the Fourth Hague Convention of 18 October 1907; hereinafter the “Hague Regulations”). Where an occupying Power, having previously established its authority in the occupied territory, later withdraws its physical presence in part or in whole, it may still bear obligations under the law of occupation to the extent that it remains capable of exercising, and continues to exercise, elements of its authority in place of the local government. (emphasis added)
Note the subtle shift in language here – the Court starts this paragraph as an inquiry into whether Gaza remains occupied, but then pivots to determining whether, after a withdrawal, the occupying power ‘may still bear obligations under the law of occupation.’ That is, the Court could here be read as saying either that the occupation continues, if in a different, lesser form, or instead as saying that that some obligations under the law of occupation persist residually even after the occupation has ended.
Finally, the Court holds that (paras 93-94):
Based on the information before it, the Court considers that Israel remained capable of exercising, and continued to exercise, certain key elements of authority over the Gaza Strip, including control of the land, sea and air borders, restrictions on movement of people and goods, collection of import and export taxes, and military control over the buffer zone, despite the withdrawal of its military presence in 2005. This is even more so since 7 October 2023.
In light of the above, the Court is of the view that Israel’s withdrawal from the Gaza Strip has not entirely released it of its obligations under the law of occupation. Israel’s obligations have remained commensurate with the degree of its effective control over the Gaza Strip.
(emphasis added)
These concluding paragraphs elevate the Court’s analytical ambiguity to the level of an art form. Note, first, how the Court’s reference to the situation after 7 October, which it makes despite previously holding that the scope of the questions before it excludes Israel’s conduct after the Hamas attack (para 81), can only relate to the fact that Israel today has a massive military presence in Gaza, i.e. does have boots on the ground. To put this differently, it is entirely possible to accept that Israel occupies all or parts of Gaza today without embracing a functional approach (see, in that regard, this expert opinion by nine eminent Israeli scholars, arguing in April this year that northern Gaza was occupied). Second, the Court never says that Gaza was occupied after 2005, or whether that occupation was at any point interrupted. Rather, it speaks of Israel not being entirely released of its obligations – which again could be read as holding that some such obligations apply residually, even after the occupation has ended. The ‘commensurate with the degree of its effective control’ language could equally be read as saying that the occupation exists, on a functional basis, or that residual obligations exist functionally even though the occupation may have, at some point, stopped (and then perhaps even resumed).
Some of this uncertainty is explored in the separate opinions of Judge Iwasawa and Judge Cleveland. Judge Iwasawa’s basic point is that the Court has embraced a functional approach to the law of occupation, but that this does not necessarily entail that it has held that Gaza remained occupied (para 8):
Thus, while the Court makes clear that Israel continues to be bound by certain obligations under the law of occupation, it does not take a position as to whether Gaza remained “occupied” within the meaning of the law of occupation after 2005. […] The situation in Gaza has drastically changed since 7 October 2023. However, events taking place after that date are beyond the temporal scope of the Court’s inquiry (para. 1 above).
This is an entirely plausible reading of the AO, but it is not the only possible such reading – the Court does not expressly say that the law of occupation can apply functionally even if the occupation has ceased, which is the implication of Judge Iwasawa’s position.
As for Judge Cleveland, she summarizes the Court’s approach as follows (para 10):
Otherwise the Opinion says very little about Gaza. In identifying the applicable law in paragraphs 88 to 94, the Court observes that after its withdrawal in 2005, Israel continued to exercise certain key elements of authority with respect to the Gaza Strip, including “control of the land, sea and air borders, restrictions on movement of people and goods, collection of import and export taxes, and military control over the buffer zone” (para. 93). In this regard, it concludes that aspects of the law of occupation continued to apply with respect to the Gaza Strip, commensurate with Israel’s degree of effective control (para. 94). However, the Court does not identify which obligations continued to bind Israel after 2005, nor does it find any violations of such obligations. In fact, the Court’s determination that the law of occupation continued to apply with respect to the Gaza Strip plays no subsequent role in the Court’s analysis. (emphasis added)
Judge Cleveland is herself somewhat coy here – she, like the Court, avoids directly pronouncing on whether Gaza remained occupied, but talks about the continued application of ‘aspects’ of the law of occupation. Whether she reads the AO in the same way as Judge Iwasawa is difficult to say (and again, probably deliberately so). Judge Cleveland is completely right when she says that the Court’s determination plays no subsequent role in its analysis, with one important exception – the Court’s holding that Israel has to withdraw from all occupied Palestinian territories, a term which it bundles Gaza into without expressly pronouncing on its status.
Judge Cleveland is also perfectly right in saying that the functional approach the Court adopted leaves substantial uncertainty, without further clarification from the Court, as to which obligations Israel was bound to comply with. She tries to address this uncertainty at paragraph 24 of her opinion:
Second, with respect to Israel’s obligations under the law of occupation regarding Gaza after 2005 (see para. 10 above), it is clear that Israel did not exercise effective control over most of the day-to-day government administration of the Gaza Strip — a responsibility which, after 2007, was under the control of Hamas. Israel, therefore, did not generally possess the effective control necessary, for example, to incur the obligation under Article 43 of the 1907 Hague Regulations to maintain public order within Gaza. Nevertheless, the Court could have found that Israel’s control over the sea and air space of the Gaza Strip, as well as land crossings (which it shared in part with Egypt (Advisory Opinion, para. 89)), and its severe restrictions on, for example, imports of food, exports, and activities such as fishing in Gaza’s maritime space (contrary to Israel’s commitments under the Oslo Accords), brought with it, inter alia, aspects of the duty “to ensur[e] the food and medical supplies of the population” under Article 55 of the Fourth Geneva Convention, as well as the duty to facilitate humanitarian relief under Article 59 of that Convention.
This makes a whole lot of sense to me, whether the functional approach is seen as one relating to the existence of the occupation itself, or as one relating to residual obligations stemming from an occupation that may have been terminated. I do have to say, though, that it is regrettable that a point of this importance is addressed only by Judge Cleveland – no matter how excellent her opinion is, the other fourteen judges have essentially remained silent on this point, and that silence exacerbates the ambiguity of the Court’s position.
To conclude, it is clear that the Court has endorsed some variant of a functional approach to (the law of) occupation. But it has also left many points uncertain. As I have explained, the core uncertainty is whether the Court regards occupation itself as a sliding-scale concept, rather than as a binary one, which could have implications not just at the end but also at the beginning of an occupation; or, rather, whether the Court believes that certain obligations under the law of occupation can apply residually, even after an occupation has terminated, to the extent that the occupying power still has some measure of control over the territory. Note that this kind of residual application theory – that the law of occupation can apply after the occupation has ended – would have some precedent in other parts of IHL, e.g. in those rules of the Third Geneva Convention protecting prisoners of war who remain detained even after the international armed conflict had ended (see more here). But again, it is only Judge Iwasawa who expressly endorses the possibility of such a theory.
Similarly, the AO says nothing about which obligations exactly applied or apply in the post-disengagement Israel-Gaza relationship. Only Judge Cleveland offered some thoughts on this question. While not central to the ICJ’s AO, these issues may well be tested before the International Criminal Court, as it decides on cases now pending before it regarding Gaza, or indeed before the High Court of Justice in Israel itself as it deals with Gaza-related litigation.
