Filling in the Blanks

While the fragile ceasefire in the Gaza Strip brings a measure of hope, the international community collectively holds its breath. U.S. President Donald Trump’s 20-point Gaza peace plan – while light on detail and largely indifferent to the Palestinian right of self-determination – gives a glimpse of the future. As with its 2005 withdrawal, Israel may retain significant control over Gaza indefinitely.

States engaged in military occupation are bound by various obligations sourced from the Hague Regulations, the Fourth Geneva Convention (“GCIV” – Part III, Sections I and III), and customary International Humanitarian Law (“IHL”). Collectively, this is the law of occupation.

The temporal and geographic scope of the law of occupation is contested. Israel’s 2005 withdrawal from Gaza thrust this contest into the spotlight, culminating in the International Court of Justice’s (“ICJ”) embrace of the so-called functional approach to the law of occupation in a 2024 ICJ Advisory Opinion, reaffirmed in its recent UNRWA Opinion. The finer details of this functional approach remain unclear.

In search of these finer details, I will first explain the doctrinal dilemma that gave rise to the functional approach. Second, I outline an analogous debate regarding the extraterritorial application of states’ international human rights law (“IHRL”) obligations, with a focus on the case law of the European Court of Human Rights (“ECtHR”). Third, I extract four lessons from this equivalent IHRL debate that ought to be applied to the law of occupation.

The IHL community was doctrinally unprepared for the remote control that Israel exercised over Gaza following 2005 – there is no need to make the same mistake twice.

The Functional Approach to the Law of Occupation

Traditionally, the boundaries of the law of occupation were simple. The obligations apply when a state has effective control of a foreign territory. They do not when it does not. Notwithstanding debates around applying the law of occupation during invasions, the binary status quo prevailed for decades.

The binary approach has, however, become increasingly untenable. Following purported withdrawal in 2005, Israel retained control over Gaza’s air and sea, a no-go zone along the border, the local monetary market, customs duties, veto rights over large-scale construction, the population registry, and border crossings (see these findings, especially [20]).

Without Israeli forces administering Gaza with a physical presence, Israel did not appear to have retained effective control. Considering the substantial authority Israel retained, however, the complete non-application of the law of occupation left an unacceptable vacuum. The alternative – classifying Israel as a conventional occupying power – was hardly appealing either. Israel would, for example, have been duty-bound to enforce law and order in Gaza.

Inspired by academia (see, eg, this and this) and the ICRC (see [50]-[52] here), the ICJ reached the following solution (at [92]):

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.

Though Israel no longer satisfied the effective control threshold, it had retained key elements of authority. Thus, its obligations remained “commensurate” with its ongoing control ([93]-[94]).

The Court recently re-affirmed this approach in its UNRWA Opinion ([85]-[87]). The Court observed that “Israel’s effective control over the Gaza Strip has increased significantly” following reinvasion, and that Israel’s obligations “have also increased significantly, commensurate with” that increase in control ([86]. See also Judge Gómez-Robledo’s critique of the Court failing to explicitly declare Gaza under full-scale occupation, [5]-[6]).

The obvious uncertainty here is how to put this notion of “commensurate” obligations into practice. What “elements of … authority” does the ICJ have in mind? Can all obligations be applied commensurately after effective control is relinquished, or only some? Is there a difference in how the substance of obligations is assessed under the functional approach?

We lack detailed answers to these questions. To fill this gap, I will turn to an analogous IHRL debate.

Extraterritoriality at the ECtHR

Pursuant to Article 1 of the European Convention on Human Rights (“ECHR”), member states owe ECHR obligations to “everyone within [the states’] jurisdiction”. This jurisdiction is ordinarily limited to a state’s own territory, with two exceptions: 1) an agent exerting authority and control over others in a foreign territory, or 2) the state exercising effective control over a foreign territory.

The ECtHR applied these principles in Banković and Others v. Belgium and Others. In response to the conflict in Kosovo, NATO forces carried out air strikes against Federal Republic of Yugoslavia (“FRY”) targets between 24 March and 8 June 1999. On 23 April, a NATO missile hit Radio Televizije Srbije, killing 16 persons and seriously injuring 16 more. Six FRY nationals brought proceedings against 17 European NATO members. The Court held that the applicants fell outside the scope of Article 1 as they were not on land over which NATO exercised effective control ([74]-[77]). Moreover, they fell outside the ECHR’s espace juridique since they did not reside in an ECHR member state ([79]-[80]).

This case drew sharp criticism from a range of doctrinal and critical perspectives. How could the ECtHR turn a blind eye to member states’ rights violations just because the victims are foreigners in a foreign war? Furthermore, even if territorial jurisdiction could not be established, what greater control can a state exercise over an individual than the power to take their life?

The ECtHR’s jurisprudential arc has since bent towards greater universalism, at least regarding states’ use of force (cf, the Court’s environmental and migration jurisprudence – see, this and this).

In Al-Skeini and Others v. The United Kingdom – which concerned the killings of six Iraqi civilians by UK forces – the espace juridique limitation was rejected outright. The victims were deemed within the UK’s jurisdiction, due to “the [UK’s] exercise of some of the public powers normally to be exercised by a sovereign government” including, relevantly, “security operations in Basra” ([149]). By way of contrast, Georgia v. Russia (II) saw the Court embrace the notion of a “context of chaos” – a concept deployed with “decisive weight” to conclude that jurisdiction was not established during active hostilities in the five-day Russo-Georgian war ([137]-[138]).

In a triumph for universalism, however, the Court in Ukraine and The Netherlands v. Russia found Article 1 ECHR jurisdiction was established for Russia’s impugned uses of force, including the bombing and shelling of civilian targets ([361]). As Marko Milanovic has explained, however, this finding was narrowly framed. The Court was at pains to emphasise that Russia’s war of territorial conquest was antithetical to ECHR values ([177], [348]-[349], [360]). Moreover, even when Russia had shelled territory outside of its effective control, they had done so with “a view to acquiring and retaining effective control” ([361]). The weight assigned to this consideration is unclear, but the Court may have left itself the option of resurrecting Bankovic-style reasoning.

Plaguing the ECtHR case law are the arbitrary distinctions that must be drawn in determining territorial and personal jurisdiction. On the former, can a state exercise effective control over a building? A room within a building? The few metres between a member of the state’s military and a civilian? Regarding personal jurisdiction, as above, what greater control can a state exercise over an individual than deciding whether to drop a bomb on them? IHRL academia has offered many solutions to these problems, of which I will describe two.

In his book on extraterritoriality in IHRL, Marko Milanovic distinguishes between states’ obligation to respect human rights (negative) and their duty to secure/ensure respect (positive). Consistent with the critiques above, Milanovic argues that the duty to respect should be unlimited. The duty to secure/ensure respect should, however, be limited to territory over which a state exercises effective control. In other words, states must not perpetrate human rights violations anywhere, but must only be champions of human rights in the territory they control.

Yuval Shany has argued that this reliance on the positive-negative dichotomy is problematic, and that imposing territorial limits on positive obligations is still too restrictive (see also Milanovic’s reply). Shany prefers a functional approach whereby obligations are applied based on the intensity of power relations between the state and a purported victim, or the existence of special legal relations. On the former, Shany would apply human rights obligations wherever a state’s conduct has direct, significant, and foreseeable impact on individuals.

I will not attempt to resolve this debate here. Instead, I hope to illustrate its relevance to the law of occupation by extracting a few key lessons.

Lessons for the Law of Occupation

Before mapping IHRL doctrine on to an IHL debate, I must acknowledge that there are salient differences between the two fields. The law of occupation governs a (hopefully) rare state of exception – a state breaking the usual bind between sovereign title and effective control of territory. In contrast, a state ought to always be conscious of its IHRL obligations. Differences like this mean that my conclusions below rely on an implicit premise: that the drift towards functional application of IHRL obligations extraterritorially is a trend that should be replicated in the law of occupation. Many will reject that premise – perhaps from a position of sympathy for military forces in chaotic environments, or from a TWAIL-inspired suspicion of IHL’s role in legitimising western military dominance. Setting aside those legitimate grounds for caution, however, there are four features of IHRL extraterritoriality theorising that I would apply to the law of occupation.

First, with some exceptions (eg, Al-Skeini before UK courts – see Milanovic’s book), the consensus is that a state’s human rights obligations may be applied in part. The question is whether the circumstances of a case give rise to a particular obligation. It is irrelevant whether a state’s IHRL obligations should apply in full. Likewise, in applying the functional approach to occupation, one should avoid broad statements of “commensurate” obligations. Instead, the focus should be on whether a particular obligation is relevant on the facts. As a simple example, if a state requisitions a hospital despite otherwise withdrawing from a territory, the prohibition on the requisition of hospitals under Article 57 GCIV ought to be applied.

Second, and as illustrated by the hospital example, the functional approach is best suited to the law of occupation’s negative obligations. Regardless of whether a strict positive-negative delineation should be embraced (ie, whether one prefers Milanovic or Shany’s approach in IHRL), it is a useful rule of thumb. In the occupation context, it is the rules against forced displacement, destruction of property, etc that most obviously lend themselves to functional application.

Third, at [149] of Al-Skeini, the ECtHR talks of the UK seizing “some of the public powers normally to be exercised by a sovereign government”, the relevant one to that case being responsibility for “security operations in Basra”. This language suits the functional approach to occupation – where a rule within the law of occupation relates to a particular field of policymaking (eg, the judiciary), one should consider whether the withdrawing force has retained control of that public power to the exclusion of local authorities. This is consistent with the aforementioned purpose of the law of occupation – regulating a state’s exercise of public power in a foreign territory.

Finally, an obvious gap between IHRL extraterritoriality and the functional approach to occupation is the narrow circumstances under which the latter has been endorsed. The ICJ carefully limited the functional approach to Gaza-like circumstances “[w]here an occupying Power, having previously established its authority in the occupied territory, later withdraws its physical presence” ([92], 2024 OPT Advisory Opinion). I believe this limitation is normatively incoherent – a state exercising partial control over a territory before/without eventually assuming effective control should not be treated differently to a state exercising partial control after relinquishing effective control. Some like-minded scholars advocate for functional application of the law during the invasion phase, consistent with the so-called Pictet theory (see, eg, here). As I have argued elsewhere, however, I prefer a generalised functional approach akin to Shany’s model for the extraterritorial application of IHRL.

Ultimately, following the example of extraterritorial application of IHRL, the functional approach should mature into what it was always meant to be: a doctrine that follows power wherever it goes.