OF WAR AND INTERNATIONAL INVESTMENT LAW

Abstract

In the past decade, the practice of investor–State arbitral tribunals addressing investment protection in the context of armed conflict and military occupation has expanded. This has prompted a growing interest in the relationship between international investment law and international humanitarian law (IHL), two regimes with markedly different relationships to war—IHL more pragmatic and international investment law more idealistic. This article argues that, while its lack of pragmatism might render international investment law ineffective in changing how war is conducted, it is the regime under which States are most likely to be held liable for the conduct of war. This is a result of its more robust primary obligations, more effective enforcement mechanisms and large awards of damages. Nevertheless, comparing international investment law and IHL does also reveal some similarities—the legacy, it is argued, of a time when the laws of war were more about protecting private property and neutral commerce than civilians. Putting these two regimes together in this way exposes international law's uneven distribution of protection in war.

I. INTRODUCTION

In the last ten years, foreign investors have brought various claims against States in the context of armed conflict and military occupation.¹ This has prompted scholars to take an increasing interest in the relationship between international investment law and international humanitarian law (IHL).² These two regimes take markedly different positions towards war. The rules and vocabulary of IHL have permeated the military, and humanitarian lawyers are often positioned as ‘pragmatic insiders’ vis-à-vis the politics and practice of war.³ International investment law, in contrast, maintains a position external to war; its vocabulary is distinct from that used by the military; most international investment lawyers would find military professionals quite alien and vice versa. International investment law lacks pragmatic relationships with the military; it is idealistic in that it puts certain (liberal economic) ideals above the realities of war and State power in an uncompromising way. Yet while its lack of pragmatism might render international investment law ineffective in changing how war is conducted,⁴ this article argues that it is the regime under which States are most likely to be held legally responsible for the conduct of war and based on past arbitral practice it is likely to stay this way. This is a result of its more robust primary obligations, more effective enforcement mechanisms and large awards of damages.⁵ Nevertheless, comparing international investment law and IHL does also reveal some similarities—the legacy, this article argues, of a time when the laws of war were more about protecting private property and neutral commerce than civilians.

Putting these two regimes together in this way exposes international law's uneven distribution of protection during war. At times, international investment law and IHL frame war as a very different problem—international investment law as one for economic order and IHL as one for human beings and societies—protecting investments and civilians to different extents and in different ways, but also for different reasons.⁶ This is a fragmentation story. Yet there is also convergence. IHL is also implicated in economic order. While this is not so visible in the contemporary field, IHL in fact has a complex regime of property protections, dating back to before its more modern shift to focusing on protecting civilians.⁷

Existing literature on the relationship between international investment law and IHL tends to focus on how to resolve or minimize potential conflicts or tensions between the two regimes through various techniques such as the doctrine of lex specialis, judicial borrowing or systematic integration.⁸ In particular, scholars often argue that international investment law should be interpreted to take greater account of IHL.⁹ Building on and engaging with this existing scholarship, the analysis in this article makes the following contributions to the literature. The article deliberately takes a more descriptive approach, for two reasons: first, to enable reflection on what the relationship between international investment law and IHL as it currently stands tells us about the uneven distribution of protection during war under international law; and second, to give a better sense of how future arbitrators are likely to develop the relationship between international investment law and IHL based on past practice.¹⁰ The article also puts the relationship between international investment law and IHL into the context of the different positions that the two regimes take towards war (external v insider, idealistic v pragmatic), pays particular attention to property as a category of protection in IHL, and carefully distinguishes the rules applicable to combat operations, situations of occupation, the State's belligerent rights more broadly, and protection against non-State actors.

First, in Section II, the article compares primary obligations across international investment law and IHL. Tribunals have generally interpreted obligations found in bilateral investment treaties (BITs) to be much more robust than similar protections in IHL, especially in the context of military operations and against non-State actors. This is so even though there is some overlap when it comes to the taking of private property in wartime.¹¹ Section III compares enforcement mechanisms. Investor–State dispute settlement (ISDS) is most effective while IHL lacks a standing complaints mechanism. Rather, IHL is enforced by an assortment of direct and indirect mechanisms. Notably, ad hoc post-war mass claims processes have directly awarded individuals redress for losses resulting from war, including from violations of IHL, but they are less comprehensive and have a mixed record of effectiveness.¹² Finally, Section IV compares remedies. ISDS is unique in its substantial awards to single investors which are often enforceable in domestic courts, although post-war claims processes have also awarded significant compensation and restitution of property on occasion. The article concludes that international investment law is the regime under which States are most likely to be held liable for the conduct of war, even if it is unlikely to influence military decision-making on the ground.

II. COMPARING PRIMARY OBLIGATIONS

International investment law, which has its origins in the customary international law on the protection of foreign nationals, is today based on a patchwork of around 3,000 treaties, largely BITs but also multilateral agreements containing investment provisions.¹³ The first BIT was signed in 1959.¹⁴ Through these treaties, States agree to guarantee various protections to each other's investors, such as fair and equitable treatment and full protection and security. Investment treaties commonly provide that investors can directly submit disputes with host States about their investments to international arbitration without the need to exhaust local remedies.¹⁵ Arbitral development of treaty protection standards has been a significant feature of the regime.¹⁶

IHL is based on both treaties and customary international law. The first IHL treaties date back to the mid-nineteenth century. Earlier IHL treaties tended to focus on the means and methods of warfare and addressed situations of inter-State war and military occupation.¹⁷ The 1949 Geneva Conventions and their 1977 Additional Protocols—today the principal IHL treaties—marked a gradual shift to a more humanitarian approach focused on protecting civilians and the introduction of a treaty regime for non-international armed conflicts.¹⁸ IHL thus has rules for international conflicts, situations of occupation and non-international conflicts. The rules applicable to each increasingly converge, including through the development of customary IHL, but do not entirely. Customary IHL is of growing importance, especially since multilateral treaty-making processes have stalled.¹⁹ There are also various issue-specific treaties that prohibit certain weapons, like cluster munitions, or protect specific objects, like cultural property.²⁰

A. Comparing Protections During Military Operations and Against Non-State Actors

This section argues that, together, war losses clauses and the full protection and security obligation mean that investors receive greater protection under BITs than they would under IHL and BITs offer investments greater protection than IHL offers civilians, particularly during military operations and against non-State actors. War losses clauses and full protection and security are investment protection standards commonly found in BITs. War losses clauses are found in 70 per cent of BITs in different forms.²¹ Most commonly, they provide that States must compensate investors for losses resulting from war to the same extent as national or third-State investors.²² In their extended form, however, they provide that States must compensate investors for losses in war resulting from State conduct not required by the necessity of the situation.²³ Full protection and security clauses, also found in 70 per cent of BITs, require host States to exercise due diligence to protect investments from physical damage or destruction.²⁴

The robustness of these provisions is as much about how tribunals have interpreted the provisions as the text of the provisions themselves. While on their face these provisions are not entirely inconsistent with IHL despite being framed in very different terms, in practice tribunals have to a significant extent interpreted them to go beyond IHL. This can be related back to international investment law's external position to armed conflict. Its strong enforcement regime means that international investment law can have material effects without needing to forge strategic relationships with the military. International investment lawyers and arbitrators are thus insulated from concerns of practicality and realism that humanitarian lawyers often find themselves subject to. International investment lawyers are free to make ‘idealistic’ decisions that promote investment protection unencumbered by any idea of military imperative, backed up by large enforceable awards of damages.

Under IHL, investments are protected to the extent that they are civilian objects.²⁵ Where investments can be considered ‘objects which by their nature, location, purpose or use make an effective contribution to military action and whose total or partial destruction, capture or neutralization, in the circumstances ruling at the time, offers a definite military advantage’, they may be directly attacked during military operations.²⁶ The International Committee of the Red Cross (ICRC) has noted that this definition of military objects ‘will not always be easy to interpret’.²⁷ Ira Ryk-Lakhman Aharonovich explains that there has been debate about whether the ‘effective contribution’ must be direct or indirect and what it means to ‘offer a definite military advantage’.²⁸ At its broadest, some have interpreted military objects to include any object that has a future military use or generates revenue that supports the war effort.²⁹ Such an interpretation would open a very broad range of investments to legitimate targeting and arguably goes too far.³⁰ Nevertheless, the ICRC accepts that dual-use and economic targets can be military objects on a case-by-case basis.³¹

To the extent that investments are civilian objects they are protected from direct attack under IHL. However, incidental damage to a civilian object during an attack on a military objective is only prohibited to the extent that it is ‘excessive in relation to the concrete and direct military advantage anticipated’ (the principle of proportionality in attack) or that feasible precautions are not taken to avoid or minimize it (the principle of precautions in attack).³² The principle of proportionality in attack, like that of distinction, is hard to interpret and apply. There are disagreements about its scope. For example, while the ICRC has argued that in this context ‘a military advantage can only consist in ground gained and in annihilating or weakening the enemy armed forces’,³³ some States are of the view that a military advantage includes the security of the attacking forces.³⁴ Feasible precautions have been defined as ‘those precautions which are practicable or practically possible taking into account all circumstances ruling at the time, including humanitarian and military considerations’.³⁵ These circumstances include ‘possible measures to protect civilians (for example, fencing, signs, warning and monitoring)’, the availability and feasibility of alternative means or methods of attack, and short- and long-term military requirements.³⁶ Ultimately, these principles leave States a certain margin of appreciation and mean that even where they are protected from direct attacks, lawful damage or destruction to investments is still possible under IHL.

The IHL prohibition on the destruction and seizure of property—a rule separate from the principle of distinction despite certain overlap—is qualified by military necessity.³⁷ The best view today is that military necessity is not a general defence to violations of IHL but a principle that applies to limit the scope of specific provisions such as this one.³⁸ Fundamentally, for a measure to be lawful in accordance with military necessity, it must be taken for a specific military purpose and necessary to achieve that purpose.³⁹ Hayashi argues that such measures must also be otherwise in accordance with IHL and that the military purpose must be had in mind at the time and not only in hindsight.⁴⁰

All these rules (distinction, precaution, proportionality and the prohibition on the destruction and seizure of property other than in cases of military necessity) are broadly considered to apply in non-international armed conflicts, as well as international ones, as a matter of customary international law.⁴¹ It is also worth noting at this juncture that BITs and IHL are different in the territorial scope of their application. While BITs only apply in the territory of the State (or following recent awards in territory occupied by the State),⁴² relevant rules of IHL also bind States when they conduct military operations in territories that are not their own. So in relation to territorial reach, IHL provides a wider scope of protection than international investment law in international armed conflict. Otherwise, however, the argument is that the substance of full protection and security and war losses clauses is stronger.

In international investment law, there is no principle of distinction that permits targeting of military objectives or concept of military necessity that permits destruction of property. However, war losses clauses could achieve the same result, protecting investors only where States choose to pay compensation to others or where destruction and damage were not justified by necessity.⁴³ Indeed, under the customary international law of alien protection, States are not responsible to foreign nationals for losses suffered as a result of State conduct during military operations.⁴⁴ War losses clauses seem on their face to reflect or at least be compatible with this position. Likewise, when it comes to full protection and security, the due diligence standard could achieve a similar result to IHL's principle of distinction whereby legitimate targeting operations that result in incidental damage to investments will not breach the full protection and security standard, provided that due diligence is exercised to avoid or minimize investor losses as a result.⁴⁵ Security exception clauses, found in 15 per cent of BITs and which provide that the treaty does not preclude measures necessary for the maintenance of national security,⁴⁶ could also incorporate military necessity into BITs.⁴⁷ In practice, however, tribunals have not interpreted these provisions thus.

Turning first to war losses clauses, tribunals have applied them alongside rather than instead of full protection and security clauses, even though war losses clauses could be considered lex specialis provisions that displace more general provisions in times of war.⁴⁸ The result is that investors can use full protection and security claims to circumvent the limitations on liability found in war losses clauses. AAPL v Sri Lanka is a good example. This claim involved the destruction of the investor's prawn farm during a counterinsurgency operation in the context of the Sri Lankan civil war. The tribunal denied the investor's claim under the war losses clause for several reasons, including lack of evidence about whether the destruction resulted from State conduct or whether it was required by the necessity of the situation.⁴⁹ Full protection and security being a broader obligation that covers a failure to protect against actions of non-State actors and without an explicit necessity exception, these obstacles did not prevent the success of the full protection and security claim. Where war losses claims fail, full protection and security claims can often still succeed.

The reverse can also be true. Strabag v Libya was a claim concerning damage to the investor's road and infrastructure works during the Libyan revolution of 2011, including through looting and vandalism by both mobs and military forces. Here, although the full protection and security claim failed for insufficient evidence of a lack of due diligence on the part of the authorities, the war losses claim was successful because the State failed to establish necessity in respect of the destruction its forces caused.⁵⁰ By applying war losses and full protection and security clauses in tandem, tribunals have used these two provisions to broaden protection for investments beyond that provided for by IHL. On one hand, destruction of property considered necessary and thus lawful under IHL and under a war losses clause can still be a breach of the full protection and security obligation where a negligent failure to protect can be established. On the other, damage caused to an investment by an attack on a military objective with no evidence of lack of precautions or due diligence and thus lawful under IHL and under the full protection and security obligation can still be unlawful under a war losses clause if the State cannot prove it was necessary.

Furthermore, it cannot be assumed that tribunals will understand necessity in the context of war losses clauses as military necessity. This has also been the case for security exception clauses.⁵¹ For example, on the face of it the tribunal in AAPL v Sri Lanka equated the term ‘necessity of the situation’ in the war losses clause with military necessity.⁵² It framed the question, however, as one of ‘whether the destruction and losses were caused as an inevitable result of the “necessity of the situation” or, on the contrary, were avoidable if the governmental security forces would have been keen to act with due diligence’.⁵³ Yet for something to be justified by military necessity does not require that it is inevitable or unavoidable in the sense of a dictionary-definition of necessity. Military necessity is a matter of there being a connection to achieving a military purpose. While the framing of necessity adopted by the tribunal implied that if necessity had been found it would also have defeated the full protection and security claim, it also potentially makes military necessity almost as difficult to prove as necessity under the customary international law of State responsibility. Finally, the fact that some tribunals have found that the burden of proof is on the State to establish necessity rather than on the investor to prove a lack of it also equates necessity in war losses clauses with a defence rather than a limit on the primary obligation. In Strabag, for example, limited prima facie evidence consisting of some UN reports of civilian abuses by the State and photographs from the investor's insurer's loss surveyor was sufficient for the war losses claim to succeed.⁵⁴

Turning to full protection and security, in AAPL v Sri Lanka, the tribunal found the State responsible for the destruction of the investor's prawn farm despite a lack of evidence of how or by whom the farm was destroyed. Responsibility was based on the State's failure to take the ‘precautionary measures … essential to minimize the risks of killings and destruction when planning to undertake a vast military counterinsurgency operation in that area for regaining lost control’.⁵⁵ This amounted to a breach of full protection and security. Although the tribunal did not mention IHL in its award, the IHL principle of precautions in attack was effectively read into the full protection and security obligation in a way that extended both. What due diligence required under full protection and security was expanded to include significant precautions when carrying out military operations. Those precautions were expanded beyond what IHL requires by being unhooked from the principle of distinction or any meaningful assessment of feasibility or military considerations.

Zrilič argues that the award in AAPL v Sri Lanka is compatible with the IHL position that harm to civilian objects incidental to legitimate military operations is only prohibited to the extent that precautions are not taken to avoid or minimize such harm.⁵⁶ It is true that the award does not suggest that the State was not within its rights to conduct the operation as a matter of international investment law or that BITs prohibit military operations. However, it is argued that the full protection and security obligation is so robust that IHL is nevertheless exceeded. The breach in AAPL v Sri Lanka was based on the fact that the authorities had failed to ‘use peaceful … communication [with the farm's management] in order to get any suspect elements excluded from the farm's staff’ before launching the military operation.⁵⁷ If the authorities believed the management to be under suspicion, the tribunal found that they should have brought judicial proceedings against the management.⁵⁸ This arguably goes beyond the precautions that IHL would ever consider feasible.⁵⁹ Tillman Braun argues that AAPL v Sri Lanka thus understands force to be permitted only as a last resort.⁶⁰ In contrast in IHL, the logic is one of putting limits on the conduct of military operations, the fact of which is accepted.

While Zrilič is right that full protection and security could be interpreted to take into account IHL,⁶¹ nothing that tribunals have done so far suggests that they would be likely to do so. Military considerations are alien to international investment lawyers and arbitrators, who are insulated from them. The only published award to refer to IHL is Strabag, but this was to confirm State responsibility for acts of members of armed forces even if contrary to instructions, rather than to interpret BIT standards in line with IHL.⁶² A State has never raised military necessity as a defence to an alleged violation of full protection and security, at least not in any claim for which documentation has been made publicly available.⁶³

When it comes to the obligation to exercise due diligence to protect investments against non-State armed actors under full protection and security, international investment law has a well-developed jurisprudence, which tends towards a high and objective standard.⁶⁴ There is no equivalent obligation in IHL that is as strong or well-fleshed out in its requirements. Most comparable to full protection and security is the IHL obligation to take all feasible precautions to protect civilians and civilian objects under the State's control from the effects of attacks.⁶⁵ The primary obligations in this regard are to locate military objectives away from densely populated areas and/or to move civilians and civilian objectives away from military objectives.⁶⁶ The ICRC, taking a broad interpretation, argues that taking precautions against the effects of attacks could also include ‘constructing shelters, digging trenches, distribution of information and warnings, withdrawal of the civilian population to safe places, direction of traffic, guarding of civilian property and the mobilization of civil defence organizations’.⁶⁷ However, even at its broadest, precaution against the effects of attacks is weak compared to full protection and security.⁶⁸

For example, in Ampal-American v Egypt, Egypt was found to have breached its full protection and security obligations by failing to provide security to a gas pipeline in the North Sinai region of Egypt in the context of the revolution of 2011.⁶⁹ Although the tribunal acknowledged that Egypt could not have prevented the initial attacks on the pipeline by armed groups, it held that ‘the first attacks should have been seen as a warning that further attacks might be carried out if security measures were not taken’.⁷⁰ The relevant part of the pipeline was owned by the State gas company, not by the investor who brought the claim.⁷¹ The investor simply suffered the knock-on effects of the attacks because the attacks affected the gas supply to the investor's pipeline.⁷² This award thus seems to make full protection and security into an extremely broad obligation to protect investors against the negative effects of insecurity. However, even if it were about protecting the investor's pipeline, this award would still understand full protection and security as a stringent obligation, leaving the State little discretion regarding decisions about where and for what purposes to deploy limited troops. The tribunal did not engage with the fact that the Camp David Accords limited Egypt to 2,000 troops in North Sinai, who were presumably facing various other competing demands, and State police had abandoned the region.⁷³

In Cengiz v Libya, Libya was similarly held responsible for failing to provide security to an investor's construction projects in the remote South of the country during 2011.⁷⁴ Although the tribunal acknowledged that Libya could not have provided ‘dynamic protection’ to the projects due to the ‘means and resources and the general political and security situation in Libya’, it nevertheless held that Libya should at least have provided ‘static protection’ against looting.⁷⁵ According to the tribunal, this was a ‘simple task’ since Libya controlled several fighting units in the South.⁷⁶ Elsewhere the tribunal noted that Libya had actually sent a military unit to the projects at one point without any evidence of the result.⁷⁷

While Ackermann argues that due diligence under full protection and security is a context-sensitive standard that can incorporate IHL considerations,⁷⁸ it is contended here that, by requiring specific reallocations of limited security resources towards foreign investments, the awards in Ampal and Cengiz show how full protection and security is just as likely to go far beyond IHL. Even in international human rights law, which generally gives States less leeway than IHL, such an approach to protective obligations has been rejected, even in respect of the right to life. In Osman v UK, the European Court of Human Rights (ECtHR) held that:

bearing in mind the difficulties involved in policing modern societies, the unpredictability of human conduct and the operational choices which must be made in terms of priorities and resources, such an obligation [to take appropriate steps to safeguard the lives of those within its jurisdiction] must be interpreted in a way which does not impose an impossible or disproportionate burden on the authorities. Accordingly, not every claimed risk to life can entail for the authorities a Convention requirement to take operational measures to prevent that risk from materialising.⁷⁹

While the ECtHR is much more deferential to States than investor–State arbitral tribunals,⁸⁰ IHL is simply less able to rely on a specialist standing court or tribunal to flesh out what its obligations require. There are plenty of non-specialist or non-standing courts and tribunals that apply IHL, either directly or indirectly.⁸¹ They can produce thick interpretations of IHL but are more often deferential to military authority or vague, and overall, their practice is too piecemeal and fragmented to have such a great influence on interpretation.⁸²

The ICRC tries to fulfil this role of a standing specialist interpretive body and can be effective. For example, the ICRC's original commentaries to the Geneva Conventions have been described as ‘essential and well-respected interpretations’ and a ‘major reference’.⁸³ They are currently in the process of being updated. Despite criticism of the methodology behind its customary IHL database, a recent study by Milanovic and Sivakumaran suggests that the database is also increasingly authoritative.⁸⁴ Other ICRC efforts at expanding upon the particulars of IHL principles have received pushback.⁸⁵ Ultimately, the ICRC does not have the same teeth as ISDS tribunals. This, together with the nature of their work with States and non-State armed actors, would seem to require and encourage the ICRC to be more concerned about realism, practicality and legitimacy from a military perspective.

What is striking in Ampal and Cengiz, it is suggested, is the tribunals' complete lack of concern in substituting their judgment for that of the original decision-makers. Arbitrators often have few facts about what happened and little information about the situation on the ground.⁸⁶ Most probably have no military experience.⁸⁷ They do not, however, perceive this as a reason for deference or restraint and this goes back to the external position that international investment lawyers inhabit towards war and the military. There is something refreshingly idealistic about this ‘outsider virtue’,⁸⁸ and it is hardly naïve—these awards can be enforceable in the billions of dollars—even if its lack of pragmatism means international investment law has little influence over military conduct and decision-making. International investment law compensates for financial losses in war, whereas IHL seeks to change the behaviour of States through more pragmatic strategies. However, the responsibility provided for by international investment law is in the name of economic order rather than human life. In AAPL v Sri Lanka, the compensation was for the destruction of the farm, not for the loss of life of the dozen Sri Lankan employees who were killed, who barely received a mention in the award.⁸⁹

B. Comparing Protections Against Expropriation of Property During Wartime

The previous section argued that the full protection and security obligation and war losses clauses in BITs give investments more protection in military operations and against non-State actors than they would receive under IHL and that IHL provides to civilians. However, there is more convergence between international investment law and IHL in situations where the State interferes with property during wartime outside of direct combat situations or circumstances related to military operations.⁹⁰ International investment law and IHL both protect property against expropriation in a belligerent's own territory and in occupied territory to similar, although not identical, extents.

Of course, property as protected by IHL is not coextensive with investments as protected by BITs. Investment is broader in the sense that, as well as movable and immovable property and real rights, it includes contractual rights, concessions, licences, copyrights, trade secrets, shares and bonds, for example. At its broadest, the concept of investment can potentially cover any economic activity, including the sale of goods and services.⁹¹ IHL property protections were developed on the understanding that property is tangible, although there are now efforts to grapple with the inclusion, for example, of digital property and data.⁹² In a different way, however, property under IHL is broader than under BITs in the sense that it protects people's homes and personal possessions. So, while there is a shared middle ground, BITs and IHL extend protection out beyond this in different directions.

In international investment law, expropriation clauses, found in 95 per cent of BITs, provide that States may not directly or indirectly expropriate investments unless it is for a public purpose and non-discriminatory and they pay compensation and follow due process.⁹³ Extended war losses clauses, as well as requiring States to compensate investors for losses in war resulting from State conduct not required by the necessity of the situation, also provide that compensation must be paid where State forces requisition investment property.⁹⁴ Recent awards concerning Crimea confirm that these obligations apply to the occupying power in situations of occupation and not just in a State's own territory even though the scope of application of BITs is framed solely in territorial terms rather than including jurisdiction.⁹⁵

Present-day IHL does not say a great deal about property as such. The Additional Protocols to the Geneva Conventions do not make direct reference to property; rather, it falls under other protections to the extent that it is a civilian or cultural object.⁹⁶ Historically, however, regulating the relationship between war and especially military occupation on one hand and neutral commerce and private property on the other was central to the development of the laws of war.⁹⁷ The Paris Declaration of 1856, for example, one of the earliest laws of war instruments, agreed that enemy goods on neutral vessels and neutral goods on enemy vessels were protected from seizure and that privateering (effectively legal piracy) was abolished.⁹⁸ While today IHL textbooks rarely address property in depth, older treatises do so extensively.⁹⁹ During roughly the century leading up to the Geneva Conventions, there developed a complex system of (treaty and customary) rules of protection depending on distinctions between enemy and neutral, public and private, and moveable and immoveable property, and the location of property and its amenability to military use.¹⁰⁰ This section will examine the extent of the convergence between these rules and the international investment law rules on expropriation.

Direct expropriation (where legal ownership is transferred or property physically seized) is a relatively rare occurrence; it is not commonly alleged in ISDS claims.¹⁰¹ Nevertheless, several recent awards concerning the nationalization of Ukrainian investments in Crimea confirm that a BIT prohibition on direct expropriation without fulfilling the requisite conditions applies as usual during military occupation.¹⁰² IHL similarly provides that the confiscation (expropriation other than for a public purpose and without compensation) of private property is prohibited in occupied territory.¹⁰³ Only limited categories of private moveable property (telecommunications equipment, transport, munitions) can be seized for military use during occupation and, even then, it must be restored with compensation at the occupation's end.¹⁰⁴ IHL also prohibits pillage (forcible taking of private property by an invading or conquering army for private or personal use) during occupation.¹⁰⁵

IHL also prohibits confiscation of private property during international armed conflict.¹⁰⁶ The Ethiopia–Eritrea Claims Commission (EECC), for example, confirmed that ‘[w]here aliens’ property is taken for State purposes in wartime, the obligation to provide full compensation continues to operate, even if the payment of that compensation may be delayed by the interruption of economic relations between belligerents’.¹⁰⁷ In addition, IHL prohibits pillage during international and non-international armed conflict.¹⁰⁸ However, according to the ICRC, there is no rule of IHL that prohibits (or permits) confiscation in non-international conflict. Rather, the matter is left to domestic law.¹⁰⁹ This also means that for foreign investors, any relevant BIT would apply.¹¹⁰

Strabag is the only known ISDS claim to consider requisitions. The investor was awarded around €6 million under an extended war losses clause in respect of the uncompensated requisition of ‘vehicles, generators, fuel tanks and other valuable … equipment’ by Libyan forces in 2011.¹¹¹ Similarly to extended war losses clauses, IHL provides that requisitions must be compensated in international armed conflict and occupation.¹¹² IHL would seem to put even stricter limits on requisitions than BITs at least on the face of the relevant treaty provisions. IHL provides that requisitions are only permitted ‘for the needs of the army of occupation’, must be ‘in proportion to the resources of the country’ and can only be demanded by the local commander. They must be paid for in cash wherever possible and if this is not possible, ‘a receipt shall be given and the payment of the amount due shall be made as soon as possible’.¹¹³ In Strabag, Libya argued that the investor did not have the proper receipts to show that the requisitions were validly made by an authorized officer.¹¹⁴ The tribunal did not, however, engage with this argument so this award does not provide much of a clue as to whether tribunals might read these additional IHL requirements into extended war losses clauses.

More difficult is the scope of indirect expropriation, which is contested at the best of times and has hardly been explored in the conflict context. Tribunals have taken different approaches to the scope of indirect expropriation depending on whether they focus on the effect or purpose of the questioned measures. Some tribunals have interpreted indirect expropriation broadly to include any ‘covert or incidental interference with the use of property which has the effect of depriving the owner, in whole or in significant part, of the use or reasonably-to-be-expected economic benefit of property’.¹¹⁵ At the same time, other tribunals have read exceptions into expropriation clauses that narrow the scope of indirect expropriation. The best-established exception, which has been applied by various tribunals, is the police powers doctrine.¹¹⁶ As classically restated in Saluka v Czech Republic, ‘[s]tates are not liable to pay compensation to a foreign investor when, in the normal exercise of their regulatory powers, they adopt in a non-discriminatory manner bona fide regulations that are aimed at the general welfare’.¹¹⁷ A less well-established exception is where a deprivation results ‘from a valid exercise of belligerent rights’.¹¹⁸ This exception was noted by the tribunal in Saluka, referencing the 1961 Harvard Draft Convention on the International Responsibility of States for Injuries to Aliens but does not seem to have been applied by any tribunal or much discussed by scholars.¹¹⁹ Tribunals have also taken different approaches to the question of duration—how long measures must last for there to be an indirect expropriation. Some have taken a broad approach, holding that there will be an indirect expropriation where the requisite deprivation is ‘not merely ephemeral’.¹²⁰ Others have taken a narrower approach and found that the deprivation must be permanent.¹²¹ Still others have found that while an expropriation is usually ‘lasting’, a temporary deprivation can ‘in some contexts and circumstances’ amount to an indirect expropriation.¹²²

Both the issues of scope and duration are significant because, under IHL, during international armed conflict belligerents have the right to take temporary restrictive security measures, such as ‘freezes and other discriminatory controls’, against the property of enemy nationals in their territory.¹²³ The only known ISDS award about indirect expropriation in a conflict context, Mitchell v Congo, took a broad approach to indirect expropriation. The claimant's law firm had its premises searched and put under seals, documents seized and two lawyers imprisoned by military forces.¹²⁴ Even though these measures were only temporary, eventually being reversed by a domestic court decision, the tribunal found that they amounted to an expropriation since they resulted in a ‘definitive’ and ‘total loss of the firm's clients’ and thus the firm, which closed down.¹²⁵ The State argued that the measures were justified by security concerns given what the tribunal referred to as the ‘rebellion against the government’ ongoing in the Congo at that time.¹²⁶ The tribunal held that it did not have to decide on this point since there had in any event been no compensation paid to the claimant and there was not enough evidence to assess the argument.¹²⁷

Despite the broad approach to expropriation, which has been criticized by scholars,¹²⁸ the decision in Mitchell v Congo does not necessarily reflect a higher level of protection than IHL would provide. The first reason for this is the limitations on belligerent rights to deal with enemy property. Article 38 of the Fourth Geneva Convention on Civilians provides that, except for security measures authorized by IHL, ‘the situation of protected persons shall continue to be regulated, in principle, by the provisions concerning aliens in time of peace’.¹²⁹ Thus, the EECC held that, despite belligerent rights regarding enemy property, ‘the basic international legal rules regulating expropriation nevertheless continue to apply’.¹³⁰ In Eritrea's Civilian Claims, the EECC ruled that:

a belligerent is bound to ensure insofar as possible that the property of protected persons and of other enemy nationals are not despoiled and wasted. If private property of enemy nationals is to be frozen or otherwise impaired in wartime, it must be done by the State, and under conditions providing for the property's protection and its eventual disposition by return to the owners or through post-war agreement.¹³¹

The EECC found that Ethiopia had failed to fulfil its obligations in this respect. A series of measures against Eritrean property, including ‘rapid forced real estate sales, confiscatory taxes on sale proceeds, vigorous loan collections, [and] expedited and arbitrary collection of other taxes’, while not all unlawful in themselves, had the cumulative effect of depriving Eritreans of their property.¹³² Similarly, in Loss of Property in Ethiopia Owned by Non-Residents, the EECC found that the cumulative effect of a series of individually lawful measures, including expelling agents and employees of non-resident businesses and refusing to issue visas or allow the appointment of new agents, was to deprive Eritrean nationals of their property unlawfully.¹³³ The EECC held that ‘even though the operation of businesses owned by nationals of an opposing belligerent and the distribution to owners of business profits and of rental payments for real property may lawfully be suspended … measures making a business property a res derelicta without compensation are not acceptable’.¹³⁴ On this basis, it seems that the EECC would also have impugned the measures in Mitchell v Congo on the basis of IHL and customary international law.

The second reason that Mitchell v Congo is not inconsistent with IHL is because it concerned a non-international armed conflict. In such a conflict, IHL does not prohibit or permit security measures against enemy property; indeed, there is no such concept of an enemy national during internal conflict. The matter is left to domestic law and, for foreign investors, any applicable BIT, just like the issue of confiscation of private property. As a related aside, the protection of the property of neutral nationals in a belligerent's territory is also largely not regulated by the laws of war. For example, neutral nationals are not covered by most of the Fourth Geneva Convention on Civilians’ provisions ‘while the State of which they are nationals has normal diplomatic representation in the State in whose hands they are’.¹³⁵ In this way, the matter is left to customary international law or any applicable BIT. Belligerents do not have the same rights to take security measures in respect of neutral property as they do enemy property.

Nevertheless, the question of whether control and security measures against property of enemy nationals would in certain circumstances amount to indirect expropriation under a BIT remains unresolved given the uncertainty surrounding indirect expropriation.¹³⁶ Such measures are only temporary, but they are likely to be of more than ephemeral effect.¹³⁷ Ackermann argues for the police powers doctrine as a corrective to an overly broad concept of indirect expropriation in the armed conflict context.¹³⁸ However, the police powers exception is not necessarily available since IHL permits measures to be discriminatory as they are specifically against the property of enemy nationals. Ackermann argues that not all measures against enemy property would necessarily be discriminatory if, for example, they were based on legitimate security concerns relating to a particular investor and not just on that investor's enemy character as such.¹³⁹ However, this still leaves those measures that are applied to enemy property in a blanket manner. The belligerent rights exception is untested. Tribunals taking a broad approach to expropriation might find security measures against enemy property unlawful whereas those taking a narrow approach would probably reach the same conclusion as under IHL.

III. COMPARING ENFORCEMENT MECHANISMS

A. The Advantages of ISDS

As well as having more robust primary obligations, international investment law also has effective enforcement mechanisms. States have given broad consent in advance to arbitrate disputes about the protection of foreign investments in thousands of BITs.¹⁴⁰ Depending on whether their location and nationality mean that their investment is covered by a BIT with an ISDS clause, investors have direct access to investor–State arbitration without having to exhaust local remedies. Although coverage is thus not universal, the sheer number of BITs, nearly all of which do provide for ISDS, mean that it is widespread. The concept of investment has been understood broadly, as has nationality and jurisdiction.¹⁴¹ The fact that consent is given before the event means that there is some certainty about whether ISDS will be available, and investors do not have to wait until the conflict is over to bring a claim. The 1965 Convention on the Settlement of Investment Disputes between States and Nationals of Other States (ICSID Convention) means that many ISDS awards are directly enforceable in domestic courts.¹⁴²

B. The Enforcement of IHL

There is currently no standing complaints mechanism through which individuals can directly file claims for loss or damage resulting from violations of IHL. While there are various international courts and tribunals that apply and enforce IHL, some only do so indirectly, like human rights courts, and others do not enable individuals to bring claims, like the International Court of Justice and the International Criminal Court (although the latter does have a victim reparation scheme).¹⁴³ Individuals can bring claims for reparation for violations of IHL directly on an ad hoc basis where States consent to claims resolution after a conflict or, less commonly, a Security Council resolution provides for it. Such consent is far from guaranteed and a Security Council resolution even less so. Nevertheless, there are several examples of international mass claims processes established after wars that have enabled individuals (and legal persons) to bring claims for losses resulting from war, including from violations of IHL, particularly since the early 1990s.¹⁴⁴ These international mass claims processes, relatively analogous to ISDS with their direct individual access, are the focus of this section. Compared to ISDS, the coverage of these post-conflict mass claims mechanisms is inconsistent and somewhat arbitrary. Although ISDS is also patchwork and relies on consent, ISDS is a more comprehensive patchwork, and the consent is given before the event instead of after. Where these mechanisms are established, they can be effective in enforcing and/or remedying violations of IHL but they have limitations and a mixed record.

The most effective example is the United Nations (UN) Compensation Commission (UNCC), an ‘innovative mass claims mechanism’,¹⁴⁵ set up after the First Gulf War by UN Security Council Resolution 687, made under Chapter VII of the UN Charter. Resolution 687 provided that Iraq was liable for any direct loss caused by the invasion and occupation of Kuwait.¹⁴⁶ Compensable losses included property destruction or damage as well as personal injury, mental suffering and even environmental harm.¹⁴⁷ Although Christine Evans has described the UNCC as ‘the first mechanism created by the UN whereby individual victims could claim compensation for violations in armed conflict’,¹⁴⁸ claimants did not actually have to prove a breach of international law. They only had to show their loss and that it was caused by the invasion or occupation. The standard of proof was to ‘demonstrate satisfactorily’ that a claim was eligible rather than prove the facts, with simple expedited procedures for smaller individual claims.¹⁴⁹ Its establishment by the Security Council under Chapter VII gave the UNCC greater authority than such a body might otherwise have and, over a period of three decades, it was able to resolve 2.7 million claims and the compensation it awarded was paid out in full.¹⁵⁰ However, the circumstances of the establishment of the UNCC are perhaps unlikely to be repeated.

Two examples of a different type of commission are the Commission for Real Property Claims of Displaced Persons and Refugees (CRPC) established by the Dayton Accords in 1996 and the Housing and Property Claims Commission (HPCC) set up by the UN Interim Administration Mission in Kosovo in 1999.¹⁵¹ They were much more limited in scope than the UNCC, dealing only with deprivation of property claims. The CRPC resolved claims seeking to recover property dispossessed during the 1992–95 Bosnian War. The HPCC settled residential property claims arising from the events of 1989–99 in Kosovo, including the North Atlantic Treaty Organization (NATO) intervention.¹⁵² Like the UNCC, these commissions were quasi-legal; claimants did not have to establish a violation of IHL, just that they had been ‘deprived [of their property] in the course of hostilities’ at the CRPC or had lost ‘possession of the[ir] property in circumstances surrounding the NATO air-campaign’ at the HPCC.¹⁵³ Like the UNCC, the CRPC and HPCC were very flexible about evidence.¹⁵⁴ Between them, they resolved around 340,000 claims, with most of the decisions executed.¹⁵⁵

A commission that more strictly applied IHL was the EECC, which was set up after the 1998–2000 war between Ethiopia and Eritrea in accordance with the Algiers Agreement that ended the conflict. The EECC arbitrated claims for loss, damage or injury related to the conflict and resulting from violations of IHL or international law more broadly.¹⁵⁶ Rules of evidence were similar to those of the UN Commission on International Trade Law (UNCITRAL) Arbitration Rules.¹⁵⁷ Although the EECC was designed to offer a mechanism for individuals to bring claims, it ended up operating largely as an inter-State arbitration. Only five claims were brought on behalf of named individuals. While the EECC produced a rich set of awards on IHL and the customary international law applicable to international armed conflict, it did not actually end up resolving mass claims like other mechanisms did.

In sum, while ad hoc mass claims processes can be effective, in some respects more so than ISDS particularly in terms of the number of claimants redressed and the flexibility of evidentiary rules, they have a mixed record. The UNCC was set up by a binding Security Council resolution which was perhaps a one-off occurrence, the CRPC and HPCC addressed only a limited category of claims, and the EECC resolved very few individual claims. The next section will consider the remedies awarded by ad hoc mass claims processes, which reinforce the variable nature of their contribution.

IV. COMPARING REMEDIES

A. ISDS and Massive Compensation

Awards of compensation are the norm in ISDS, something that is not necessarily the case for other international dispute settlement mechanisms.¹⁵⁸ Even in comparison to other international courts and tribunals that do provide monetary damages, ISDS is famous for its massive awards of compensation to individual investors. The largest ever ISDS award is US$40 billion.¹⁵⁹ Although these relatively rare enormous awards can distort the overall impression,¹⁶⁰ the mean award is still in the hundreds of millions.¹⁶¹ ISDS awards can be so big because they compensate for the loss of the whole value of the investment and for loss of profits. For example, in Yukos, the Permanent Court of Arbitration awarded the investors compensation for the loss of the entire value of their shareholdings as well as dividends that would have been paid out.¹⁶² The ECtHR, in contrast, in a case arising out of effectively the same facts, awarded the claimant compensation only for the particular acts (tax penalties and enforcement fees) that violated the right to property (although it still amounted to nearly €2 billion).¹⁶³

In the conflict context, there have not been awards of amounts like that seen in Yukos. Awards have generally not compensated for loss of profits. Nevertheless, significant amounts have been awarded. Even though in AAPL v Sri Lanka the investor received compensation for the loss of the full value of its shareholding, this amounted to US$460,000, a modest sum in the ISDS context.¹⁶⁴ In Cengiz v Libya, the investor was awarded US$50 million for actual damage suffered. The investor's claim for loss of profits was considered too speculative given its lack of track record in Libya, the remoteness of the projects which were a long way from completion, and the security and business environment in Libya at the time.¹⁶⁵ In another recent conflict-related claim against Egypt, Union Fenosa, the compensation awarded amounted to US$2 billion, based on the cost to the investor of sourcing gas from elsewhere after the gas supply to its plant was suspended by a State-owned enterprise during the 2011 revolution.¹⁶⁶

Many of these awards are enforceable in domestic courts under the ICSID Convention, which has been ratified by over 150 States.¹⁶⁷ However, in practice it is not necessarily straightforward to do so. It can be a long and expensive process to pursue enforcement proceedings in various jurisdictions. As a result, claimants can end up accepting smaller settlements, simply using the award as a bargaining chip.¹⁶⁸ Union Fenosa is a good example. After a series of additional proceedings seeking both to enforce the award and stay its enforcement, the claim was finally settled without any of the US$2 billion compensation being paid.¹⁶⁹ More controversially, claimants can also end up selling their awards to a third-party investor at a discount (who will then go on to try to enforce them).¹⁷⁰ So ISDS is effective, but it is worth asking for whom.

B. Remedies and Ad Hoc Mass Claims Processes

The level of damages and the enforceability of awards vary widely among ad hoc post-war claims resolution mechanisms.¹⁷¹ For example, the UNCC paid out more than $50 billion to 1.5 million claimants.¹⁷² This was enabled by the fact that it was established by the UN Security Council and, controversially, had access to Iraqi oil revenue from which to fund claims.¹⁷³ However, the UNCC shows that massive compensation can, in the right (albeit unusual) circumstances, be paid outside of the investment context. While individual claimants were generally paid relatively low amounts (in the thousands of dollars), corporate claimants received awards into the hundreds of millions.¹⁷⁴ Fixed amounts were awarded for certain lower-value individual claims for forced displacement, personal injury and death.¹⁷⁵

In contrast, no compensation has ever been paid out by the CRPC or the HPCC.¹⁷⁶ While the Dayton Accords provided for the CRPC to pay compensation to claimants whose property could not be restored, the compensation fund was never set up due to lack of funds (in part because of a view among potential donors that a compensation option would distract from the goal of restoring property).¹⁷⁷ While the lack of compensation was a limitation, other remedies were nevertheless available. The CRPC made over 300,000 awards restoring property rights, estimated to benefit up to one million people.¹⁷⁸ As with the CRPC, the focus of the HPCC was on the restitution of property rights rather than compensation.¹⁷⁹ The few decisions awarding compensation did not end up being enforced,¹⁸⁰ but nearly 30,000 decisions awarding repossession, restitution or registration of property were successfully implemented.¹⁸¹ This distinction between remedies relates back to the differences between investments and property discussed above.

At the EECC, the individual claimants were awarded approximately US$2 million in compensation in total, while Eritrea received around US$160 million and Ethiopia US$175 million (leaving Eritrea owing a balance of about US$15 million).¹⁸² Due to lack of evidence, the amounts awarded were estimates (at best). The EECC stated that, even though the lack of evidence was such that ‘in a commercial arbitration … [it] might warrant dismissal of a damages claim for failure of proof’, it was not appropriate to dismiss the claims here given that ‘significant violations of international law causing harm to many individual victims’ had been established.¹⁸³ However, the Algiers Agreement did not stipulate any procedure for making payments to individual claimants and, in the end, no payments have ever been made between Ethiopia and Eritrea.¹⁸⁴ Apparently, since Ethiopia was unhappy with the decision of the parallel Eritrea–Ethiopia Boundary Commission, they did not want to be seen to legitimize either commission by demanding that Eritrea pay the difference between the EECC's awards.¹⁸⁵ This has led some to conclude that ‘reparation has not been achieved through this process and justice for individuals was largely symbolic’.¹⁸⁶ In sum, although the ad hoc mass claims processes can deliver effective and flexible remedies in the right circumstances, they can be hit and miss, particularly when it comes to compensation.

V. CONCLUSION

This article has argued that international investment law's robust primary obligations, effective enforcement mechanisms and large awards of damages make it the regime under which States are most likely to be held liable for the conduct of war, even if it is unlikely to influence military decision-making on the ground. This is evident from the current trend of ISDS claims in this context, and based on past arbitral practice it is likely to stay this way. Particularly, war losses clauses and the full protection and security standard provide greater protection than similar provisions in IHL, especially during military operations and against non-State actors. This is so even though there is some overlap when it comes to the taking of private property in wartime. Although ad hoc mass claims processes mean that individuals are not always left without redress for violations of IHL during conflict, these mechanisms are less effective than ISDS.

It is thus argued that international law distributes protection unevenly in war. International investment law and IHL protect against confiscation relatively equivalently, but otherwise investors are more likely than civilians to have their war losses redressed. How should this be interpreted? Ackermann argues that the arbitration of investment claims arising out of conflict is a good thing since it ‘contribute[s] to the perseverance of the rules-based international order’ and is a step towards increased accountability in war.¹⁸⁷ Others have expressed concern that it could be a threat to post-conflict peace building.¹⁸⁸ Would it be good if international investment law had a ‘chilling’ effect on war?¹⁸⁹ Would this incidentally benefit civilians and communities too? It has been suggested above that international investment law's lack of pragmatism means that it is unlikely to influence how war is conducted. For now, States have IHL lawyers, not investment lawyers, advising their military decision-making and international investment law looks likely to remain an outsider to the policies and practices of war. Claims that the possibility of investor–State arbitration incentivizes States to undertake broad governance reform not just for the benefit of investors but for society more broadly have so far proved unfounded.¹⁹⁰ International law's uneven distribution of protection in war would seem to be a feature of the international legal order then, rather than a temporary bug on the way to stronger humanitarian limits in conflict.