International investment law and repressive state power: Rethinking a relationship

Racial justice movements, such as Black Lives Matter, have sparked domestic and international debates about the role of policing in perpetuating racism and its inability to guarantee the safety of the most vulnerable in society. Even though police and prison abolitionism remain minority positions, their renewed presence in the public debate can encourage more reserved critics of state repression to rethink the appropriate role of police or the army in society. This rethinking will not be easy, not least because state violence enjoys a privileged position in domestic and international legal systems. In this brief contribution, we want to highlight an unexpected synergy between international investment law and state repression that has generally escaped scrutiny. To this end, we discuss the “full protection and security” (FPS) standard, which constitutes a commonplace clause in bilateral investment treaties. In particular, we argue that even though international investment lawyers have been defending their field by arguing that it restricts arbitrary and authoritarian state power, investment tribunals have used FPS to demand that states use repressive violence to protect investments. These demands have been put forward even in conditions of social upheaval and resource scarcity and when investors themselves have contributed to violence directly or indirectly. For this reason, we argue that investment lawyers should rethink the relationship between the field and state violence and that those concerned about the role of policing in society should pay closer attention to international investment law.

Ours is not an evident position. In Asian Agricultural Products Ltd.[AAPL] v. Sri Lanka—the first case that dealt with FPS—the tribunal found the then-embattled state to have violated FPS due to its military forces causing extensive damage to a shrimp farm in the process of its violent counter-insurgency campaign against Tamils. Given Sri Lanka’s abysmal record of violence during the civil war, this ruling appears justified, and, if anything, it offers some counterweight to the state’s excessive violence. In fact, criticisms of AAPL v. Sri Lanka have generally adopted the viewpoint of the state, criticizing the tribunal’s refusal to assess if the army had reasonable alternatives and whether its actions were consistent with international humanitarian law. These criticisms identify real flaws in the reasoning of the arbitrators, but they do so in order to defend the state’s repressive capabilities. Of course, these are valid arguments from a state sovereignty perspective insofar as they correctly identify that investment tribunals encroach—often without proper reasoning—into the state’s protected sphere of legitimate violence. In particular, the willingness of arbitrators unfamiliar with the laws and realities of war and in possession of minimal evidence about the facts on the ground (see, for example, Cengiz v Libya [218], [219]; AAPL v Sri Lanka [64]) to find violations of international investment law evidences the field’s bias in favour of investors, who are often required to do very little beyond showing that they suffered some adverse effect in order to be awarded compensation.

Nevertheless, our own research highlights that the tribunal jurisprudence on FPS is more complicated.[1] In fact, even AAPL v Sri Lanka does not rule out violent state repression as long as it is deliberate and controlled rather than wanton or reckless. Other awards go even further. It is commonly accepted that FPS encompasses both positive and negative obligations. AAPL v Sri Lanka focused on the latter, namely the obligation that the state does not inflict through its organs direct physical damage upon the investment. Lately, though, cases rely on the positive obligations of the state, that is, its obligation to prevent non-state actors from damaging the investment. Most such cases fall under one of two categories: first, the host state is facing extensive upheaval, civil war, or popular uprising. As security forces become stretched or even defect to the opposition, it becomes difficult to guarantee the security of any investment, domestic or foreign. When damage eventually occurs, investors bring cases under BITs—and they often win. A series of cases against Middle Eastern and North African states in the aftermath of the so-called Arab Spring exemplify this trend. The second type of case concerns violence and/or unrest targeted specifically against the investment due to labour disputes, adverse environmental impact, or other social conflicts, for example, disputes over land in the aftermath of formal colonialism. In such instances, the state appears reluctant to use violence—at least beyond a certain threshold—especially when there is a generalized local uprising against the specific investors and/or investors have contributed to the tensions, for example, by hiring private security firms who go on to commit crimes against those resisting.

In general, both scholars and tribunals have evaluated both scenarios from the perspective of the state. The main disagreement in literature and practice, then, concerns whether FPS is a variable standard, which depends on a state’s resources and circumstances (see Pantechniki v Albania), or a universal standard, measured against “a reasonably well organized modern State” (this was the benchmark adopted in AAPL v Sri Lanka, at 77). The former is seen as more flexible and accommodative of states’ needs and interests, especially when it comes to relatively under-resourced states in the Global South facing internal upheaval. The latter has been described by UNCTAD as placing “a clear premium on political stability, and the obligation of host countries to ensure that any instability does not have negative effects on foreign investors, even above the ability to protect domestic investors” (p. 41). Our research has led us to question whether this distinction between the variable and objective standards is as important as often assumed. This is due to tribunals’ failure to take into account either the capacity of the state in the particular context or the conduct of investors in a meaningful way.

First, even when tribunals proclaim that they adopted the variable standard, they tend not to provide a reasoned and detailed account as to how the state had failed not in the abstract but in light of the prevailing circumstances and the resources at its disposal. Ampal-American et al. v Egypt is a good example. This claim concerned a series of attacks on a gas pipeline by armed groups in the context of the 2011 Egyptian Revolution. The tribunal held that the state’s failure to respond to the initial attacks by providing security to the pipeline was a breach of FPS. Despite acknowledging that Egypt was facing difficult circumstances, the tribunal did not engage in a meaningful analysis of Egypt’s capacity to respond. For example, the tribunal did not address the fact that the Camp David Accords limited Egypt to 2,000 troops in the area, which the state police had abandoned. This does not mean that Egypt was in the right, but the failure of the tribunal to engage with these circumstances effectively blurs the variable and objective standards of protection. What is more, the pipeline was not even owned by the investor, an issue overlooked by the tribunal in its analysis of FPS. Rather, the investor relied on the supply of the gas from the pipeline to supply its own pipeline. This shows how FPS, even when framed in less strict terms, can operate as a broader security guarantee to ensure that conflict does not negatively impact investors. States are, in practice, required not only to protect the physical integrity of particular investments but also to repress any non-state violence that might affect investor profitability.

Cengiz v Libya is another award where the tribunal paid lip service to assessing FPS in light of the “means and resources and the general political and security situation” in the respondent state yet ended up applying a stricter standard. Although the tribunal conceded that Libya could not in the circumstances have provided “dynamic security” to the investor’s construction projects, there was nevertheless a level of “basic security” against looting by militias that could and should have been provided since Libya had troops in the area. In this way, FPS can effectively require an active deployment of security resources in favour of investors regardless of competing demands and interests. Although investors have not always been successful with their FPS claims against Libya (see Oztas, Güriş, and Strabag), there is a distinct body of arbitral jurisprudence that prioritizes security for investors even under extreme circumstances of conflict and state collapse.

Unless the reasoning of arbitrators changes in practice to take into account both resource restraints and also the explosive environment caused by the conduct of some investors, the distinction between objective and subjective standards will remain of conceptual but not of practical importance. The reasoning of the International Court of Justice provides some guidance in that direction. In its 1989 case, Elettronica Sicula, which relied on a Treaty of Friendship, Commerce and Navigations between the United States and Italy, the court applied a genuinely subjective version of the standard of protection (“no less than that accorded to the nationals, corporations and associations of the other High Contracting Party or any third country”) including by emphasizing the fact that some social unrest was to be expected in the light of 800 workers being laid off. The judges, then, adopted an analysis that relied not only (or even primarily) on resources dedicated to state repression, but they also took into account social frictions and resistance against unpopular investment decisions, eventually finding that Italy had not violated its “constant protection and security” obligations.

This observation takes us to our second concern about the current state of FPS-related jurisprudence. Especially in cases of unrest specifically directed at the investment, tribunals have explicitly or implicitly required that the state use as much repressive violence as is necessary for the safety of the investment to be secured. As noted already, this has also been the case in instances when the investor’s behaviour has been shown to have at least contributed to the escalation of violence. For example, in the case of Copper Mesa v the Republic of Ecuador, the state was found in violation of its FPS obligations despite the fact that, first, the tribunal acknowledged that there was a generalized uprising in the region against the mining project that would have required the state to “declare war against its own people” in order to suppress opposition to the mine; and second, the arbitrators admitted that the investor had contributed to the escalation of violence by hiring private security who went on to commit a multitude of crimes against locals. The only consequence of the latter was a reduction of 30% to the awarded compensation. Similarly, in the case of von Pezold v Republic of Zimbabwe, the arbitrators found the respondent in violation of its FPS obligations due to its failure to repress protesters who occupied white-owned farms in an effort to force land redistribution in light of unequal ownership along racial lines that was the direct consequence of the country’s history of white supremacy. The respondent noted the widespread nature of the protests and the fact that land redistribution is a politically loaded issue in Zimbabwe in an effort to persuade the tribunal that suppressing the protests would have required extensive violence. The arbitrators rejected this assertion and ruled that violent suppression was both possible and required under the FPS clause.

Admittedly, the jurisprudence of investment tribunals on the issue is neither perfectly clear nor perfectly uniform (compare, for example, Tecmed v Mexico with von Pezold and Copper Mesa). Nevertheless, what is concerning is that tribunals have repeatedly construed FPS to require swift and violent suppression of local resistance against investments. This has been the case even when popular resistance was motivated by environmental or human rights concerns or was directly linked to the socially destructive effects of open white supremacy and formal imperialism and colonialism. In fact, both cases discussed above involve instances that explicitly (in the case of von Pezold) or implicitly (in the case of Copper Mesa) concerned racial justice. In the case of Zimbabwe, protesters were directly challenging racially skewed patterns of land ownership that were attributable to Rhodesian policies of white supremacy and to British colonialism before that. In the case of Ecuador, the region where mining was taking place was extremely poor, and over 60% of its residents were Indigenous or Afro-descendants and had resisted extractive businesses for decades, concerned about the health and environmental effects of open-pit mining.

The impact of arbitral awards on state behaviour is an open empirical question that does not lend itself to easy answers. For this reason, we do not argue that there is a straight line between the demands of international investment law and real-world repressive statehood. Nevertheless, the normative bias of the case law that we mapped above makes FPS yet another terrain of struggle for those who want to rethink the purpose of the state for the benefit of both racial justice and social justice more broadly.


Kathryn Greenman (Senior Lecturer, UTS Faculty of Law) & Ntina Tzouvala (Associate Professor, ANU College of Law).


[1] See: Greenman, K. (2021). Protecting foreign investments in revolution and civil war: Critiquing the contemporary arbitral practice. London Review of International Law, 9(3), 293–318; Tzouvala, N. (2022). Full protection and security (for racial capitalism). Journal of International Economic Law, 25(2), 224–241.