1 Introduction

Proof that an investment has been infected with corruption will probably deliver a death blow to any investment-treaty claim relating to such investment.1 Various arbitral tribunals2 and scholars3 have recognised that this outcome creates an asymmetry in international investment law. This asymmetry arises because when an investor engages in a corrupt act with a state official, there is always state involvement in the corruption, yet the state is not sanctioned for such wrongful conduct in investment-treaty arbitration.4

What is not as well recognised is that this asymmetry can worsen (or potentially disappear) depending on the form of the corruption in question.5 For example, if an investor devises a plan to corrupt an until-then honest state official who works in a corruption-free government department, this is rather different from a situation where the investor confronts a state official who only performs his or her job on the payment of bribes. In the first case, denying the investor’s claim6 is probably a proportionate response, but it looks rather disproportionate in respect of the second case.7 But the jurisprudence on corruption in international investment law does not usually distinguish between forms of corruption:8 if there is any corruption, the ordinary result will be a denial of the investor’s claim.9

What is the relevance of this jurisprudence for investment-treaty arbitrations involving Asian states? Because of the prevalence of corruption among state officials working in Asian governments,10 and the relatively high amounts of foreign direct investment that finds its way into Asian economies,11 Asian states stand to benefit from this jurisprudence in future investment-treaty arbitrations. A foundation stone of this chapter is that where investor–state corruption takes the form of systemic corruption, they should not so benefit. Going forward on that foundation, a proposal is put forward on how, even when the investor has participated in systemic corruption, arbitral tribunals might avoid inflicting a death blow on the investor. Very concisely, the key to that result is to show that arbitration agreements for investment-treaty arbitrations are subject to the doctrine of duress, and that doctrine might be invoked to nullify the legal effects of the investor’s participation in an instance of systemic corruption.

To arrive at that destination, the first port of call is to show how systemic corruption differs from other forms of corruption, particularly because the state should be apportioned most of the blame for it. Illuminating this difference is the concern of Sect. 7.2. On account of this difference, it follows that systemic corruption should be treated differently in investment-treaty arbitrations; in other words, investors should not be denied on the basis of systemic corruption. How arbitral tribunals can avoid such denials is assessed in Sect. 7.3. To that end, it is demonstrated how investment-legality requirements can be applied differently. Section 7.4 considers the issue of how arbitral tribunals might respond to investors’ participation in systemic corruption, including, for example, by referring them to anti-corruption agencies. Section 7.5 contains the conclusion.

2 The (Limited) Wrongfulness of Investor Participation in Systemic Corruption

This chapter demonstrates how the asymmetry12 that arises when arbitral tribunals deny investors’ claims on account of their participation in systemic corruption can be eliminated. That change in the jurisprudence, however, will live or die according to the reasons that underpinned it. In Sect. 7.2.2, these reasons are put forward, but, before getting there, a prerequisite is to have an understanding of what systemic corruption is.

2.1 The Nature of Systemic Corruption

‘Systemic corruption’ is a well-known concept in both government-produced literature13 and scholarship, but there is no authoritative definition of the term.14 Indeed, it is apparent that many authors presume that their audience have an instinctive understanding of what systemic corruption is or are otherwise ready to work with fairly basic definitions, such as ‘corruption is not the exception to the rule, but the rule’.15 Michael Johnston, however, has ventured into some turbulent waters by attempting to corruption16 and has put forward this definition for systemic corruption:17

Systemic corruption is not a special category of corrupt practice but rather a situation in which the major institutions and processes of the state are routinely dominated and used by corrupt individuals and groups, and in which many people have few practical alternatives to dealing with corrupt officials.

This definition identifies two defining characteristics of systemic corruption: first, corrupt practices by the state official are routine18 and, second, the other party has no or little alternative but to participate in the corrupt practice. These two characteristics distinguish systemic corruption from its counterpart, what might be called isolated corruption, which is sometimes called individual corruption.19 As its name suggests, it arises when the state official engages in corruption as a one-off occurrence or otherwise infrequently: the corruption is the product of the individuals concerned, not the system that they are part of.20 Finally, systemic corruption should be distinguished from institutional corruption, the basic idea of which is that an institution becomes corrupted because its members habitually (and lawfully) act contrary to its purpose.21

What does systemic corruption look like, particularly in respect of investment-treaty disputes? Consider a case where the investor is constructing an apartment complex, which is its investment. The land has been purchased for 1,000,000 ducats. The investor now sets about procuring the various permits that it needs to build the apartment complex and transform it into a profit-making investment. These permits need to be procured from various governmental departments. In some of these departments, the culture of corruption is such that bribes are not asked for, but simply expected. If they are not paid, then the application for the relevant permit is not acted on. This is the advice of the investor’s local consultant. The investor pays the bribes. Its permits are granted. The investor spends 6,000,000 ducats on its investment, of which about 200,000 ducats is dedicated to paying bribes.

2.2 Systemic Corruption: Wrongfulness of the Conduct of States

If good evidence of this corruption was produced in an investment-treaty arbitration involving this apartment complex, it would be likely that the arbitral tribunal would deny the investor’s claim.22 And while the corruption in this case might be seen as low-level corruption, as opposed to some high-level corruption involving high-ranking state officials or their associates,23 the general position in international investment law is that all corruption should be met with a response of this nature.24

This approach is fairly simplistic. As long recognised in the political-science literature on corruption, not all corruption is created equal,25 which is also the case with legal concepts under which many different crimes fall, such as theft: shoplifting is a different beast compared to creating a long-running Ponzi scheme. The problem for arbitral tribunals in investment-treaty arbitrations is that they only see that they have one tool to sanction corruption, namely denying the investor’s claim. Such a response might be, at times, disproportionate to the wrongfulness of the corruption. But if an arbitral tribunal ever came to doubt the proportionality of declining jurisdiction, it could comfort itself with the thought that it was upholding the rule of law. This sentiment is most evident in the first investment-treaty arbitration where the arbitral tribunal declined to exercise jurisdiction on account of corruption,26 specifically Metal-Tech v. Uzbekistan:27

As a result of the foregoing analysis, the Tribunal lacks jurisdiction over Metal-Tech's treaty claims … the Tribunal is sensitive to the ongoing debate that findings on corruption often come down heavily on claimants, while possibly exonerating defendants that may have themselves been involved in the corrupt acts … The idea, however, is not to punish one party at the cost of the other, but rather to ensure the promotion of the rule of law, which entails that a court or tribunal cannot grant assistance to a party that has engaged in a corrupt act.

But are arbitral tribunals promoting the rule of law with this approach? Apparently the theory is that when an arbitral tribunal declines jurisdiction over an investor’s investment-treaty claim, that investor and other investors are disincentivised from engaging in corruption in respect of their investments in the future. As every act of corruption undermines the rule of law,28 then with this disincentivisation, the end result should be a stronger rule of law.

This is not a strong argument in respect of systemic corruption. In this context, it must be remembered that systemic corruption is a product of the state, hence the designation ‘systemic’ corruption. When an arbitral tribunal declines jurisdiction over an investment-treaty claim on account of systemic corruption, what incentive does the state have to change its systems? None. In fact, this outcome (declining jurisdiction) might act as an incentive to continue its corrupt practices29 because it means that it does not have to face investment-treaty claims that have been infected with its systemic corruption.

A final point is that the theory is naive about the realities of combatting systemic corruption. Evidence shows that systemic corruption is very difficult to root out.30 A root-and-branch transformation of government bureaucracy is needed,31 something which a decision in an arbitral award cannot hope to achieve. Because of this, the practice of denying investment-treaty claims based on systemic corruption looks like, at best, virtue signalling and, at worst, further entrenching corruption.

Practically speaking, therefore, this rule of law–based argument fails. Nonetheless, there is still the moral contention;32 specifically: all corruption is wrongful33 and so the arbitral tribunal should do what it can to sanction it. It is considered that all corruption is wrongful, but where most (or all) fault lies depends on the form of the corruption in question. In respect of systemic corruption, most fault has to be placed at the feet of the state.34 As its name suggests, systemic corruption is a product of the ‘system’ within a state, as opposed to the greed of particular individuals.35 Because of this comparatively higher degree of fault compared to the investor, the state should not benefit from an arbitral tribunal’s response to a finding of corruption, thereby meaning that there is no good moral basis underlining the practice of declining jurisdiction.

But there is still the lingering issue of the investor’s participation in the corrupt dealing. Is this conduct also wrongful? Starting from the foundation that all corruption is wrongful,36 it follows that any participation in a corrupt dealing is wrongful. An investor might respond, however, that ‘the system made me do it’,37 which is effectively a duress-derived argument that it was coerced into the corrupt dealing. Duress is a moral game-changer: if a person’s act is the product of coercion, he or she will be relieved of moral responsibility for it.38 Accordingly, this the-system-made-me-do-it argument has some potential in the moral sphere.39 The question is now whether it can be used in the legal sphere, specifically for the purpose of giving arbitral tribunals an alternative to ruling that the investor’s investment-treaty claim should be rejected (on account of its participation in an instance of systemic corruption).

3 Current Jurisprudence on Application of Investment-Legality Requirements in Cases Involving Corruption

Before assessing whether an investor might make use of a duress-based argument to keep its investment-treaty claim alive, that assessment must be properly contextualised. To this end, it is helpful to begin by outlining the current doctrine that informs how arbitral tribunals respond to situations where there is proven investor–state corruption.

3.1 Current Doctrine on Corruption in Investment-Treaty Arbitration

Investor–state corruption becomes particularly relevant in investment-treaty arbitration via ‘investment-legality requirements’. They are found in the definitions of ‘investment’ in most investment treaties, although some do not have them.40 In this latter case, arbitral tribunals have nonetheless sometimes found ways to read them in, meaning that investment-legality requirements are an omnipresent feature of investment-treaty arbitrations. In respect of these implicit investment-legality requirements, there is a question regarding whether they are relevant to jurisdiction, admissibility of the claim or its merits.41 The same question does not arise in respect of explicit investment-legality requirements because it is clear that they are conditions of states’ consent to arbitration.42

As an explicit investment-legality requirement will likely be applicable in any investment-treaty arbitration, they will be in focus for the remainder of this chapter. Although their exact wording varies between investment treaties,43 they usually stipulate that an investment must be made in accordance with host state law. Whenever an investor has participated in corruption in developing its investment, it should always fail to satisfy this requirement, noting that even the most corrupt states have laws against corruption.44

To take some of the bite out of investment-legality requirements, some jurisprudence has been developed that softens their sharp edges.45 Perhaps because of the wrongfulness associated with corruption, these softeners have not been applied in cases involving corruption, with the possible exception of the case of Gavrilović v. Croatia. ‘Possible’ because the arbitral tribunal was reluctant to call the relevant conduct ‘corrupt’, but rather preferring to see it as an ‘irregular dealing’ between the investor and Croatia (including a government minister), apparently fearing that it might assume the role of a criminal court if it did make a finding of corruption.46 In any case, the investment-legality requirement was not breached because the state (through its courts and a government minister) orchestrated the dealing.47 Presumably another factor that influenced the arbitral tribunal was that even though the various state officials involved in this dealing were abusing their offices, they did not (apparently) directly benefit from the dealing. Rather, the benefit of the investor’s part in the dealing went to Croatia (for its war effort against Serbia).48

The point is that if there is investor–state corruption, the investment-legality requirement will operate to bite the investor, unless the case can be brought within the narrow exception carved out by Gavrilović v. Croatia. As noted in Sect. 7.1, this doctrine is bound to deliver asymmetrical outcomes, particularly in cases involving systemic corruption, because the wrongfulness of the investors’ participation in such corruption is limited. But how can this outcome be avoided? The plain words of investment-legality requirements stipulate that investment must be lawfully made, which is clearly breached if an investor participates in corruption. It could be argued that ‘law’ in an investment-legality requirement assumes a different meaning in systemically corrupt societies; that is, that it should mean ‘the law in action’ as opposed to ‘the law in the books’, though it is considered that this argument has no prospects of success. The better approach is not to concentrate on the meaning of the words in investment-legality requirements, but rather to think more broadly about their application.

3.2 The Application of Explicit Investment-Legality Requirements in Cases Involving Systemic Corruption: Non-ICSID Arbitrations

Enter the doctrine of duress. Its promise is that it can stop the application of investment-legality requirements in cases involving systemic corruption. Very concisely, it can operate in this way because it can undo the legal effect of an act that was the product of duress. Applied to cases involving investor–state systemic corruption, the argument would be that if the investor’s participation is the product of duress, then it should not have any legal effect, meaning that it could not breach any investment-legality requirement.

To make this argument work, two pillars need to be put in place: (i) investment treaty–based arbitration agreements are subject to the doctrine of duress and (ii) an investor’s participation in systemic corruption could be the product of duress. Whether each of these pillars can be laid is assessed in the next three sub-sections.

3.2.1 The Applicability of the Doctrine of Duress to Investment Treaty–Based Arbitration Agreements

The basic issue is: are investment-treaty-based arbitration agreements subject to duress? In order to start answering that question, it needs to be determined what the applicable law is on the validity of such arbitration agreements. From here on, this will be referred to as the law of the arbitration agreement. If the doctrine of duress is part of this law, then it follows that investment treaty–based arbitration agreements are subject to it.

The process of making this determination starts with recognising that there are two types of investment-treaty arbitrations, namely ICSID and non-ICSID.49 There are two different regimes that underlie each of them. As regards ICSID arbitrations conducted under the ICSID Arbitration Rules (2022),50 they are entirely underpinned by the international legal regime,51 most especially the ICSID Convention, with other international law playing a supplementary role. For present purposes, this means that the law of the arbitration agreement for ICSID Convention arbitrations is international law. The follow-up question is then whether there is a doctrine of duress in international law that might be applicable, which is a question addressed later on.

Turning to non-ICSID arbitrations, they are grounded in domestic law,52 specifically the law of their arbitral seats. The law of an arbitration agreement for a particular arbitration, however, can be different from its arbitral seat.53 As an example, in Ecuador v. Occidental, which is a case that derived from the London-seated Occidental v. Ecuador (I), the English High Court held that while English law was the law of the arbitration, the law of the arbitration agreement was international law.54 Whether this still represents good law is open to doubt,55 but it illustrates the point that domestic law or international law might be applicable.

How can a final determination be made? It is ultimately a matter of applying the choice-of-law rules on the law of arbitration agreements in the law of the forum; for example, and with reference to Ecuador v. Occidental, an English court (the forum) would apply English law to make this determination. The good news is that the basic choice-of-law rule in almost all jurisdictions is that if the parties explicitly choose a governing law for their arbitration agreement, that will be the law of the arbitration agreement.56 The bad news is that such a choice is seldom expressed in investment treaty–based arbitration agreements, either directly in the treaty or indirectly via the arbitration rules that the investor is allowed to select for its claim.57 If there is no explicit choice made, then it is somewhat of a lottery as to which law will end up being applicable because a number of different methods (to make this determination) have developed across different jurisdictions.58 While many jurisdictions opt for the law of the arbitral seat and see it as the primary indicator of the parties’ intention regarding the law of the arbitration agreement, other jurisdictions look to the substantive law governing the relationship between the parties.59

What does all of this mean for present purposes? In respect of non-ICSID arbitrations, either domestic law (law of the arbitral seat) or international law (substantive law between the investor and the state) might be applicable as the law of the arbitration agreement. Does it matter whether one or the other is applicable? Probably not, because both almost certainly contain a doctrine of duress. This is certainly the case if domestic law is applicable: the rules on the validity of arbitration agreements are informed by the corresponding rules in contract law (which are operative in the relevant jurisdiction);60 all well-developed bodies of contract law contain a doctrine of duress,61 thereby entailing the conclusion.62

To get to the same conclusion in respect of international law is somewhat trickier. It might be thought that if international law is the law of the arbitration agreement, then the Vienna Convention on the Law of Treaties (VCLT) is applicable, which indeed contains duress-related provisions.63 The reason for the apparent application of the VCLT is that the validity of an investment treaty-based arbitration agreement is in question. But this is a misapprehension. An investment treaty–based arbitration agreement is a separate agreement that derives from a treaty, much in the same way that arbitration agreements are separate from the contracts that they are often found in.64 This separateness excludes the application of the VCLT because it only applies to state–state treaties,65 with the result that other international agreements between states and other subjects of international law are explicitly excluded from its scope.66

Because the VCLT is not applicable, recourse should be made to other sources of international law, namely custom and general principles of law. The latter is particularly helpful for present purposes. Generally speaking, a principle qualifies as a general principle of law if it is a feature of well-developed domestic legal systems,67 particularly if it forms part of private law.68 As previously noted,69 a doctrine of duress is a part of every body of contract law, meaning that it fits well within this definition. It follows that duress is a general principle of law, thereby making it part of international law for the purposes of the law of an arbitration agreement (governed by international law). And this conclusion does not only concern non-ICSID arbitrations, but also ICSID arbitrations. In this context it should be noted that while Article 25 of the ICSID Convention generally controls issues of jurisdiction, it only performs that role in respect of the jurisdictional requirements that it mentions,70 such as ‘legal dispute’, ‘investor’ and ‘investment’.71 The question of the application of other jurisdictional requirements, such as investment-legality requirements, is governed by general international law.72

3.2.2 Temporal Application of the Doctrine of Duress

Putting all of this analysis together, the conclusion is that all investment treaty-based arbitration agreements are subject to a doctrine of duress. The next question is whether the doctrine can save investors’ corruption-infected claims. Answering that question has several dimensions, one of which is whether an investor’s participation in systemic corruption can legally be seen as a product of duress, while another aspect looks at whether the doctrine temporally applies. The second aspect is examined in this sub-section.

The issue of the temporal application of the doctrine of duress arises because of how an investor would use it. To recap, if an investment-legality requirement is activated by an investor’s participation in systemic corruption, then the arbitration agreement will be terminated; but, to avoid that outcome, the investor will argue that it should be revived because its conduct was the product of duress. A potential problem with this argument is that the relevant conduct (systemic corruption) occurs before the arbitration agreement is formed. It might be questioned whether the tentacles of the doctrine of duress spread this far. When duress applies preformation, it ordinarily acts to render contracts void,73 not revive them.74 The revival of contracts is something that can be achieved if there is duress in the post-formation phase.75

To respond to this problem, a basic precondition is to know the content of the applicable doctrine of duress—its temporal application has to be determined. As explained in Sect., in respect of investment treaty–based arbitration agreements, this doctrine could come from domestic law or international law. Considering that arbitration agreements for ICSID arbitrations are governed by international law, and most investment-treaty arbitrations are ICSID arbitrations,76 it will be assumed that international law is applicable. This creates some difficulty because there is no authoritative source on the content of the international doctrine of duress, as it applies to agreements between states and other subjects of international law. To fill the void, domestic doctrines of duress can be looked to, but their content will also differ between jurisdictions. Because it is representative of the doctrine of duress throughout the common law world, reference will be made to English law. This is not to suggest that English law serves as the dominant body of law upon which Asian legal systems were built, although, acting with French law and German law, English law has been influential in Asia. But moving beyond that point, the broader purpose here is to demonstrate that arbitration agreements do not exist in a legal vacuum, but rather have some governing law that regulates their application. To achieve this purpose, English law serves as an illustrative tool.

Under the English doctrine, could an ‘unformed agreement’ be revived? Without direct jurisprudence on this point, no firm conclusions can be drawn, but such revival may be possible. Investment-legality requirements can be seen as conditions precedent to investment treaty-based arbitration agreements. If one party coerced another party not to perform a condition precedent, with the result that the contract was not formed, it is difficult to see how this is different from a situation where an existing contract is terminated with duress, noting that such a contract will be revived under English law.77 A further enhancement to this argument is the opening comment of Lord Burrows in his doctrinal exposition on the English doctrine of duress in Pakistan International Airlines v. Times Travel, which was that ‘duress may take various forms’.78

Finally, it might be argued that, at the time of the investor's participation in the systemic corruption, an arbitration agreement may have already formed, assuming that the investor is already developing its investment. This argument butts up against the conventional wisdom that investment treaty–based arbitration agreements form when the investor initiates arbitration.79 Another view, and the one that is a foundation of the theory developed in this chapter, is that, at this moment, the investor's obligation under such arbitration agreements is perfected, with the result that the state's obligation becomes due for performance. This ‘moment of perfection’ should not be confused with the ‘moment of formation’.80 If the contract is a unilateral contract,81 which investment treaty-based arbitration agreements arguably are,82 formation occurs when the promisee (the investor) starts performing.83 When does an investor start performing? The foundational case of Carlill v. Carbolic Smoke Ball Co is instructive. The offeror promised as follows: if any person used its ‘smoke ball’ and became infected with influenza, then such person could make a claim against the offeror for GBP100.84 In a subsequent case, Soulsbury v. Soulsbury, it was held that as soon as a promisee inhaled a smoke ball, the contract was formed.85 What is the equivalent of inhaling a smoke ball when thinking about investment treaty-based arbitration agreements? It is the act of making an investment in the host state. It is the first step that has to be made in order for the state to perform its obligation.

3.2.3 Is Investor Participation in Systemic Corruption a Product of ‘Duress’?

Accordingly, even if it was conceived that the doctrine of duress only applies to revive contracts that are already formed, it is likely that, when an investor makes an investment, an investment treaty–based arbitration agreement is already formed. But the temporal application of the doctrine of duress is not the real challenge, however. The real challenge is determining whether a state’s conduct relating to an investor’s participation in systemic corruption can be substantively described as ‘duress’.

Again, the question hinges on the content of the international doctrine of duress, which it has been assumed that English law might provide some inspiration for. In Pakistan International Airlines v. Times Travel, Lord Burrows posited that duress comprised three elements.86 When they are reformulated for cases where the investor argues that an investment treaty-based arbitration agreement should be reactivated, this three-element test provides that the state must make (i) an illegitimate threat towards the investor that (ii) causes the investor to perform the relevant invalidation act and (iii) the investor had no reasonable alternative but to perform such an act.

Regarding the first element, ‘illegitimate threat’, one basic rule is that if the threat is unlawful, then it is always illegitimate.87 Thus, if a state official indicates that he or she will only perform his or her duty upon the payment of a bribe by the investor, then this element will be satisfied. But it will be a rare case where a state official openly asks for a bribe. In systemically corrupt government departments, it is likely that a bribe will be expected,88 meaning that the state official will not have to ask for it. When an investor encounters such an environment, does the state official ‘make’ an illegitimate threat? There are two options here. First, if the investor can show that it honestly and reasonably believed that, unless it paid a bribe, the state official would not act, then this might constitute ‘making’ an illegitimate threat. Because this idea stretches the notion of ‘making’, a second option would be to add that the investor has to ‘pay under protest’;89 in other words, the investor has to indicate that it does not wish to participate in systemic corruption. As regards the second element, the causal link between the investor’s act of participating in systemic corruption and the illegitimate threat, this will ordinarily be satisfied because investors will not pay bribes unless they are made to.90 For the case where an investor intends to get the government service that it seeks via corruption, this element will act to disqualify it from relying on duress.

Turning to the third element, specifically that the investor had no reasonable alternative, its application will be generally contestable. Its application involves a balancing act.91 If the illegitimate threat is unlawful, this element requires the weighing of different factors, particularly the value of the potential loss (of not submitting), the chances of nullifying the threat with legal action and the amount of the bribe.92 For investors, the value of the potential loss will hinge primarily on the amount that it has invested; in other words, an investor who has already invested 1,000,000 ducats faces greater pressure than the investor who has invested 1,000 ducats. Additionally, the centrality of the permit for the investor’s plans will be relevant: a permit for the investor to perform the basic investment activity has a different meaning compared to one that makes doing business easier.

As regards the second factor, this examines whether the investor could have successfully applied for an injunction-like order against the state (to make it withdraw its demand for the investor’s participation in the corrupt dealing). The ideal venue for this purpose would be the courts in the host state, but certain realities stand in the way of this ideal. Factually speaking, if the investor faces systemic corruption in one government department, then it might reasonably assume that the courts also lack independence.93 Theoretically, the investor might also have the option of initiating emergency arbitration against the state to obtain such an order, but this option is not available under the ICSID Arbitration Rules (2022), nor the UNCITRAL Arbitration Rules (2021),94 which are the most commonly applicable rules for investment-treaty arbitrations. And while many arbitral institutions offer emergency arbitration, one of the popular arbitral institutions,95 namely the International Chamber of Commerce, has specifically excluded the possibility of emergency arbitration in respect of investment-treaty cases.96

4 The Lingering Issue of the Investor’s Corrupt Conduct

In summary, investment treaty–based arbitration agreements are subject to a doctrine of duress and investors might use it to save their investment-treaty claim, if such claim will otherwise be disqualified on account of its participation in systemic corruption. This outcome addresses the asymmetry that arises when a state effectively benefits in investment-treaty arbitration from systemic corruption in its governmental departments. Nonetheless, the fact is that there is no sanction brought down on the investor. How might arbitral tribunals deal with the lingering issue of the investor’s corrupt conduct?

There is always the option of considering the investor’s corrupt conduct on its merits,97 although it is considered that this should be a stop-gap solution. The basic reason underlining this view is that arbitral tribunals are not the ideal corruption fighters,98 which is, in turn, a premise informed by two other reasons. First, there is the issue as to whether arbitration is a suitable venue to adjudicate on criminal (or quasi-criminal) matters, particularly given the public interest in such matters, remembering that the general public is the ultimate victim of corruption. Second, because of the secretive nature of corruption, it is especially difficult to prove—and this problem is only exacerbated in arbitration because arbitral tribunals lack powers to compel the production of evidence.

It has to be emphasised that arbitral tribunals should not stop adjudicating on corruption. When systemic corruption has been proven, arbitral tribunals should take action. The point is rather that the toolkit of arbitral tribunals needs to be expanded (to take such action). It is submitted that the best way for an arbitral tribunal to deal with instances of systemic corruption is to refer them to the proper law-enforcement body, which will probably be the anti-corruption body in the investor’s home state. It is important to emphasise that the most important building-block for this idea is already in place: the prevalence of extrajurisdictional laws prohibiting corruption in international business. Under these laws, the home states of investors criminalise the corrupt conduct of their investors in host states.99 The impetus for the development of these laws was the Convention on Combating Bribery of Foreign Public Officials in International Business Transactions.100 All OECD members states, plus Argentina, Brazil, Bulgaria, Peru, Russia and South Africa, have ratified this treaty.101 Accordingly, whenever an investor from one of these states, which includes all of the major capital-exporting states, invests abroad, that investor can be prosecuted for its corrupt conduct in its home state.

The policy underlying this submission that allegations of corruption should be referred to anti-corruption specialists—rather than letting arbitral tribunals guess how to sanction, it is better to hand over the task to people who know how to do it. New legal infrastructure would have to be constructed for that purpose in order to ensure that arbitral tribunals feel secure in making such referrals,102 specifically some kind of ‘referral mechanism’. That has to be the procedural innovation of the future.

5 Conclusion

If that future never eventuates, arbitral tribunals for investment-treaty arbitrations will have to use the only tool that they currently have, the investment-legality requirement. This works like a sledgehammer—it can drive a crucifixion-sized nail into the heart of an investment-treaty claim. But this punishment should fit the crime, which it probably does not if the investor’s crime is its participation in systemic corruption. The problem for arbitral tribunals is that investment-legality requirements are rigid. Although some arbitral tribunals have sought to temper them by subjecting them to a test of proportionality,103 which requires that the consequence (denying the investor’s claim) must be proportionate to the unlawful conduct, this limitation has not found its way into the wording of investment-legality requirements. Formally speaking, investment-legality requirements are simple: if there is unlawful conduct in the establishment of an investment, then arbitral tribunals are given a sledgehammer with which they can strike at the investor’s claim.

What this contribution has shown is that arbitral tribunals do not always have to pick up the figurative sledgehammer on proof of systemic corruption. This is because, like any arbitration agreement, investment treaty–based arbitration agreements need a governing law. That law will usually be international law, which must contain a doctrine of duress, as it is a general principle of law. Accordingly, investment treaty–based arbitration agreements are subject to the doctrine of duress. The content of this international doctrine of duress is unclear. To fill the void, indicatively, English law was referred to. As demonstrated in Sect., a feasible argument could be constructed that the investor’s participation was the product of duress. If that argument was successful, then the legal effect of the investor’s conduct would be nullified, specifically the breach of the investment-legality requirement. With its nullification, the investor could continue to push its investment-treaty claim.104 With that outcome, the asymmetry that results when a state succeeds in an investment-treaty arbitration on account of the investor’s participation in systemic corruption would be addressed.

An obvious limitation to this proposal is that the case must involve systemic corruption. Does this limit the effectiveness of the proposal put forward in this chapter? No. Almost by definition, systemic corruption has to be far more commonplace than opportunistic corruption. Moreover, arbitral tribunals have good reason to punish investors’ participation in opportunistic corruption because an investor chooses to participate. Accordingly, in rebalancing the asymmetry caused by the application of investment-legality requirements, there is still room for arbitral tribunals to apply investment-legality requirements in cases involving corruption where the investor is deserving of punishment.


  1. 1.

    Llamzon 2014, para. 7.10.

  2. 2.

    See Fraport AG Frankfurt Airport Services Worldwide v. Republic of the Philippines, ICSID Case No. ARB/03/25, Dissenting Opinion of Mr. Bernardo M. Cremades (19 July 2007) (Fraport v. Philippines), para. 37; and Metal-Tech Ltd. v. Republic of Uzbekistan, ICSID Case No. ARB/10/3, Award (4 October 2013) (Metal-Tech v. Uzbekistan), para. 389.

  3. 3.

    Brower and Ahmad 2018, para. 18.02.

  4. 4.

    Reid and Zamour 2022, p. 275.

  5. 5.

    But see Bishop 2010, p. 66.

  6. 6.

    ‘Denial of the investor’s claim’ refers to, one, an arbitral tribunal’s declining of jurisdiction and, two, rejecting the investor's claim on the merits on account of corruption. Exactly which of these two options the arbitral tribunal takes depends on whether the investment-legality requirement in question is relevant to jurisdiction or merits; see further Sect. 7.3.1 below.

  7. 7.

    Bishop 2010, p. 66.

  8. 8.

    The major exception here is the case of Georg Gavrilović and Gavrilović d.o.o. v. Republic of Croatia, ICSID Case No. ARB/12/39, Award (25 July 2018) (Gavrilović v. Croatia). There, as the arbitral tribunal found that high-level Croatian officials had orchestrated the corruption, it did not make an adverse finding against the investor.

  9. 9.

    Christofi 2021, p. 135.

  10. 10.

    In the most recently released Global Corruption Barometer on Asia (covering Bangladesh, Cambodia, China, India, Indonesia, Japan, Malaysia, Maldives, Mongolia, Myanmar, Nepal, the Philippines, South Korea, Sri Lanka, Taiwan, Thailand and Vietnam), the first ‘key finding’ was that ‘nearly three out of four people think that government corruption is a big problem in their country’; see Transparency International 2020, p. 5. See also Chapter 1 in this volume.

  11. 11.

    UNCTAD 2022a, b.

  12. 12.

    By way of note, this asymmetry is an interesting counterpoint to the asymmetry that is often highlighted in international investment law, specifically that only investors can bring claims (but states ordinarily cannot counterclaim); see generally Kürtz 2020. For scholarship that challenges this asymmetry narrative, see Ishikawa 2022 (reviewed by Nottage 2023).

  13. 13.

    See, for example, Public Safety Canada 2022, p. 2.

  14. 14.

    Although it has been defined for legal purposes under Australian (federal) law, see Law Enforcement Integrity Commissioner Act 2006 (Cth), s 5 (‘“systemic corruption” means instances of corrupt conduct (which may or may not constitute serious corruption) that reveal a pattern of corrupt conduct in a law enforcement agency or in law enforcement agencies’).

  15. 15.

    Persson et al. 2013, p. 449 (original emphasis). See also Stefes 2007, p. 6 and Johnston 1998, p. 85.

  16. 16.

    See Kurer 2015, p. 30.

  17. 17.

    Johnston 1998, p. 89.

  18. 18.

    Kahana and Qijun 2010, p. 83.

  19. 19.

    Ceva and Bagnoli 2021, p. 4.

  20. 20.

    Klitgaard 2004, p. 1.

  21. 21.

    Lessig 2021, p. 553. See also generally Thompson 2018.

  22. 22.

    Llamzon 2014, para. 7.10.

  23. 23.

    It is recognised that systemic corruption can be both low level and high level; see Stefes 2007, p. 6.

  24. 24.

    Jarrett 2023, pp. 228–230.

  25. 25.

    Klitgaard 1988, p. 46. See also Chap. 1 in this volume.

  26. 26.

    Reid and Zamour 2022, p. 273.

  27. 27.

    Metal-Tech v. Uzbekistan, para. 389 (emphasis added).

  28. 28.

    See Fuller 1969, p. 81.

  29. 29.

    Bulovsky 2019, p. 132.

  30. 30.

    Stefes 2007, p. 8.

  31. 31.

    See Rothstein (2018), p. 45 (‘The main goal should be to convince the population that the basic social contract is about to change and to give them a stake in the existence of a well-functioning public sector that can deliver important goods to them in an honest and competent manner.’).

  32. 32.

    See, for example, Glencore International A.G. and C.I. Prodeco S.A. v. Republic of Colombia, ICSID Case No. ARB/16/6, Award (27 August 2019) (Glencore v. Columbia) where the arbitral tribunal described corruption as ‘morally odious’; see Glencore v. Columbia, para. 663.

  33. 33.

    Karklins 2005, pp. 6–8. Note, however, that some political scientists contend that some forms of corruption are beneficial for economic growth; see Aidt 2019, p. 618.

  34. 34.

    Reflects standard view that if a person causes a consequence, he or she is ordinarily morally responsible for it; see generally Sartorio 2007, p. 750.

  35. 35.

    Puckett 2010, pp. 818–819.

  36. 36.

    Karklins 2005, pp. 6–8.

  37. 37.

    Phrase taken from title of Karklins’s book; see Karklins 2005.

  38. 38.

    Sartorio 2022, p. 355.

  39. 39.

    See further Ceva and Bagnoli 2021.

  40. 40.

    According to UNCTAD statistics, approximately 66% of all investment treaties contain explicit investment-legality requirements; see UNCTAD, Mapping of IIA Content. See also Chap. 1 in this volume.

  41. 41.

    There are at least four investment-treaty arbitrations where this was a live issue; in chronological order: Plama Consortium Limited v. Republic of Bulgaria, ICSID Case No. ARB/03/24, Decision on Jurisdiction (8 February 2005) (Plama v. Bulgaria); Bear Creek Mining Corporation v. Republic of Peru, ICSID Case No. ARB/14/21, Award (30 November 2017) (Bear Creek v. Peru); Cortec Mining Kenya Limited, Cortec (Pty) Limited and Stirling Capital Limited v. Republic of Kenya, ICSID Case No. ARB/15/29, Award (22 October 2018) (Cortec v. Kenya); and Fynerdale Holdings BV v. Czech Republic, PCA Case No. 2018–18, Award (29 April 2021) (Fynerdale v. Czechia). In the first two, the illegality was held to be relevant to the merits, while, in the latter two, it was held to be relevant to jurisdiction; see Plama v. Bulgaria, paras. 126–130; Bear Creek v. Peru, paras. 318–324; Cortec v. Kenya, para. 319; and Fynerdale v. Czechia, para. 554 (by majority).

  42. 42.

    Obersteiner 2014, p. 268.

  43. 43.

    For an overview of the various types of ‘in accordance with host state law’ clauses, see Reinisch 2018.

  44. 44.

    Rothstein 2018, p. 39.

  45. 45.

    For an overview of these ‘softeners’, see generally Risso and Amaro 2020.

  46. 46.

    See Gavrilović v. Croatia, para. 399.

  47. 47.

    Ibid. paras. 383 and 396.

  48. 48.

    See ibid. para. 325.

  49. 49.

    Kriebaum et al. 2022, p. 342.

  50. 50.

    Only these ICSID arbitrations are governed by international law; see ibid. p. 345.

  51. 51.

    See Steingruber 2012, para. 12.37.

  52. 52.

    Commission and Moloo 2018, para. 1.04. Note that as domestic law is formally applicable, it will often be the case that international law is applicable because the domestic law is informed by the New York Convention; see Paulsson 2016, p. 33.

  53. 53.

    Blackaby et al. 2015, p. 157.

  54. 54.

    See Republic of Ecuador v. Occidental Exploration and Production Company [2005] EWHC 774 (Comm); Occidental Exploration & Production Company v. The Republic of Ecuador [2005] EWCA Civ 1116, para. 13.

  55. 55.

    A new test for determining the law of the arbitration agreement was subsequently pronounced in Sulamerica v. Enesa Engenharia.

  56. 56.

    Born 2021, p. 617.

  57. 57.

    Although, interestingly, if the LCIA Rules are applicable, then they include a deeming provision to the effect that the law of the arbitral seat is the law of the arbitration agreement; see LCIA Rules, Article 16(4). LCIA Rules are also potentially applicable in investment-treaty arbitrations; see Onwuamaegbu 2018, para. 3.03.

  58. 58.

    For an overview of these methods, see Blackaby et al. 2015, pp. 158–165.

  59. 59.

    A smaller minority (including Switzerland) default to whichever law validates the arbitration agreement, while another even smaller minority of jurisdictions (which are French law dominated) jump directly to divining the ‘common intentions of the parties’ as the applicable law of the arbitration agreement; see generally Scherer and Ole Jensen 2021.

  60. 60.

    Born 2021, p. 895.

  61. 61.

    Smith 1997, p. 343. For an example of the global acceptance of duress as an invalidating factor, see UNIDROIT Principles of International Commercial Contracts (2016), Article 3.2.6. Additionally, the doctrine forms part of bodies of contract law in various Asian jurisdictions; see Beale 2022, pp. 510–513.

  62. 62.

    Indeed, duress is already used to challenge the validity of arbitration agreements; see Born 2021, pp. 871–872.

  63. 63.

    See VCLT, Arts. 51 and 52.

  64. 64.

    On the separability principle, see generally Blackaby et al. 2015, pp. 104–107.

  65. 65.

    VCLT, Article 1.

  66. 66.

    VCLT, Article 3.

  67. 67.

    Statute of the International Court of Justice, Article 38(1)(c).

  68. 68.

    Crawford 2019, p. 32.

  69. 69.

    See note 61.

  70. 70.

    For a list, see Schill et al. 2022, p. 100.

  71. 71.

    Note that although some authorities advocate the view that Article 25 contains an implicit legality requirement, the better view is that it does not; see ibid. pp. 258–262.

  72. 72.

    Ibid. p. 99.

  73. 73.

    In English law, duress-procured contracts or changes to contracts are voidable, not void; see McKendrick 2013, para. 8.203.

  74. 74.

    Lucy 2007, p. 345.

  75. 75.

    See Pakistan International Airlines v. Times Travel [2021] UKSC 40, para. 62.

  76. 76.

    UNCTAD 2022a, b, p. 1.

  77. 77.


  78. 78.


  79. 79.

    Schreuer 2012, p. 836.

  80. 80.

    McKendrick 2013, para. 8.16.

  81. 81.

    ‘Unilateral contract’ as conceived under English law, which comprises a promise (offer of arbitration) and an act (acceptance of offer of arbitration by performing various acts, such as making an investment and initiating arbitration); see Martin and Law (2006), p. 555. It is noted that in other legal systems ‘unilateral contract’ assumes a different meaning; for example, in Japanese law, it refers to a gift; see Sono et al. 2018, para. 47.

  82. 82.

    See Steingruber 2012, para. 11.45 and Paulsson 1995, p. 232.

  83. 83.

    Davison-Vecchione 2014, p. 22; although the jurisprudence is not uniform on this point. For example, in Australian contract law, the case law points in different directions; see Veivers v. Cordingley [1989] 2 Qd R 278 (indicating that part performance forms a unilateral contract) and Mobil Oil Australia Ltd v. Wellcome International Pty Ltd & Anor [1998] FCA 205 (indicating that part performance does not form a unilateral contract). For a critique of Mobil Oil v. Wellcome, see Clark 2000.

  84. 84.

    Carlill v. Carbolic Smoke Ball Company [1892] EWCA Civ 1, p. 256.

  85. 85.

    Soulsbury v. Soulsbury [2007] EWCA Civ 969, paras. 49–50.

  86. 86.

    Pakistan International Airlines v. Times Travel, para. 78 (citations omitted).

  87. 87.

    Ibid. para. 82.

  88. 88.

    Kahana and Qijun 2010, p. 83.

  89. 89.

    For inspiration behind this second option, Occidental Worldwide Investment Corporation v. Skibs A/S Avanti (The Siboen and Sibotre) [1976] 1 Lloyds Rep 293, p. 336 (per Kerr J: ‘One relevant factor would be whether the party relying on duress made any protest at the time or shortly thereafter’).

  90. 90.

    This assumes that the test of causation is the ‘but for’ test. English courts have held that this test is applicable to test causation; see Kolmar Group AG v. Traxpo Enterprises Pty Ltd [2010] 2 Lloyd’s Rep 653, para. 92.

  91. 91.

    For an example of how it is done, see Carillion Construction Ltd v. Felix (UK) Ltd [2000] ADR. L. R. 11/06, paras. 41–43.

  92. 92.

    See, for example, Modular Windows v. Command Construction (1984), 11 CLR 131, para. 28.

  93. 93.

    Systemic corruption is ordinarily the symptom of deeper problems that spread their tentacles throughout society; see Johnston 1998, 88.

  94. 94.

    Under the ICSID Arbitration Rules and UNCITRAL Arbitration Rules, only ‘expedited arbitration’ can be pursued, which is quite different from ‘emergency arbitration’; see Sim 2021, para. 2.39.

  95. 95.

    For a list, see ibid. para. 2.55.

  96. 96.

    See ICC Rules of Arbitration 2021, Art 29(6)(c).

  97. 97.

    See Jarrett 2023, p. 231.

  98. 98.

    Rose 2014, p. 186.

  99. 99.

    Pacini et al. 2002, p. 390.

  100. 100.

    See generally Miller 2000, p. 139.

  101. 101.

    OECD 2018.

  102. 102.

    Particularly bearing in mind the duty of confidentiality that arbitral tribunals are subject to.

  103. 103.

    Subjecting investment-legality requirements to the principle of proportionality was developed by the arbitral tribunal for Vladislav Kim and others v. Republic of Uzbekistan, ICSID Case No. ARB/13/6, Decision on Jurisdiction (8 March 2017). For an analysis of that arbitral tribunal’s approach to investment-legality requirements, see Mouawad et al. 2021, pp. 71–75.

  104. 104.

    Note that when an investor uses duress to strike down a state’s attempt to terminate their arbitration agreement, the investor does not seek to void this arbitration agreement, but rather void its termination. That duress can be used for the purpose of re-establishing a contractual relationship is well established in the case law; see, for example, North Ocean Shipping Co. Ltd. v. Hyundai Construction Co., Ltd. [1979] QB 705 (where a duress-induced amendment to a contract was voided) and Pao On v. Lau Yiu [1979] 3 All E.R. 65 (where a duress-inducted amendment to a contract was not voided, but the Privy Council recognised that, if duress was proven, it could have operated in this way). See further England and Rafferty 1980, pp. 633–638.