2.1. The purposes behind international regulation of enforcement

There is a simple functional explanation for why international law might be concerned with how States enforce their laws: States’ desire to discourage self-interested enforcement that has negative effects on foreign actors.Footnote ¹⁰ In general, law enforcement imposes costs—including the costs of being placed at a competitive disadvantage—on both the enforcing State and the actors targeted for enforcement with the aim of inducing the targets to change their behaviour in ways that benefit, or at least reduce harm to, other actors. These costs and benefits will not necessarily be distributed equally across domestic and foreign actors.Footnote ¹¹ States generally will have little interest in pursuing enforcement strategies that create net costs for domestic actors, even if they produce net benefits for foreign actors. Correlatively, they will have an interest in strategies that generate net benefits for domestic actors, even if they impose net costs on foreign actors.

The focus here will be on situations involving enforcement aimed at commercial actors, including both firms and individuals. The nationality or location of an actor is not necessarily obvious. Consequently, for the purposes of the analysis that follows, whether an actor qualifies as foreign or domestic in relation to a particular State does not turn on the actor’s nationality or location but rather on the more elusive question of whether the State has an interest in promoting the actor’s welfare.

Foreign actors who are prejudiced by a State’s self-interested enforcement strategies have an interest in persuading the State to commit itself to more other-regarding strategies, commitments that might be embodied in international agreements of various sorts, including not only instruments focused on repression of transnational crime but also agreements concerned with trade and investment.Footnote ¹² The question then becomes: how might a State be persuaded to assume international obligations that require them to refrain from setting enforcement strategy based purely on self-interest?Footnote ¹³ The simplest reason why a State might agree to refrain from self-interested law enforcement is to induce other States to make reciprocal commitments. For instance, imagine two States that derive no direct benefit from sanctioning cross-border fraud committed by actors based in their territory—and in fact find it costly—but will each benefit from the other State’s cross-border enforcement efforts. Self-interest, narrowly defined, dictates that each State refrain from sanctioning cross-border fraud. However, it is easy to see that the two States may find it mutually beneficial to commit themselves to sanctioning frauds that affect the other State. In return for undertaking costly enforcement, each State will reap the benefit of the other State’s enforcement effort.

Promises of reciprocity will not always be sufficient to induce States to abandon self-interested enforcement. A State which finds it unusually costly to undertake enforcement that yields benefits for other States, or which derives relatively little benefit from other States’ enforcement efforts, will not be strongly moved to abandon self-interested enforcement by promises of reciprocity. In this case, other States, or perhaps even non-State actors, may have to sweeten the pot with additional benefits in order to induce enforcement. Alternatively, they may resort to coercion—i.e., threats to impose costs—in order to extract commitments to enforcement.Footnote ¹⁴

2.2. Functions of specific types of enforcement obligations

The proposition that treaties’ enforcement obligations are designed to prevent harm to foreigners, implies that the content of those obligations can be explained by identifying enforcement practices foreigners might be concerned about. Four broad categories of enforcement practices seem relevant: under-enforcement, over-enforcement, discriminatory enforcement and concealed enforcement.

First, States might fail to take actions that would diminish the welfare of domestic actors but enhance the welfare of foreigners (under-enforcement).Footnote ¹⁵ Provisions that demand effective enforcement seem to respond to this concern. In principle, States also can prejudice foreigners by taking affirmative steps to enforce their laws, as in the case where they enforce so vigorously that they drive criminals to other countries.Footnote ¹⁶ Interestingly though, it is difficult to find examples of international obligations designed to discourage this kind of over-enforcement.Footnote ¹⁷

A second concern is that States will take actions that diminish the welfare of foreign actors but enhance the welfare of domestic actors. In competitive settings, the most straightforward way to impose costs on foreign actors while simultaneously benefitting domestic actors is to systematically impose greater regulatory burdens on foreign actors than on their domestic competitors (discriminatory enforcement). In multilateral settings, the concept of discriminatory enforcement includes practices that impose greater regulatory burdens on some foreign actors than on others, thereby placing the burdened actors at a competitive disadvantage. A binding commitment to refrain from discriminatory enforcement has the direct effect of improving the welfare of actors who otherwise would be discriminated against. It also has the indirect effect of discouraging parties who fear discrimination from entering into costly and potentially futile competition to secure favourable treatment through rewards or coercion. Of course, a State which abides by such a commitment will have to incur the costs of ignoring credible rewards or threats. Non-discrimination provisions can address any or all of these forms of discriminatory enforcement.

There is a third reason why foreigners might be concerned about a State’s enforcement practices. States that practice under-enforcement or discriminatory enforcement sometimes will have an incentive to conceal their behaviour from the affected foreign actors—in other words, to engage in concealed enforcement. Incentives for concealed enforcement arise in at least two scenarios. First, consider the situation of a State which has been promised a reward in exchange for a commitment to refrain from these practices. One way to capture the reward is to honour its commitment. An even better option is to pretend to honour its commitment and claim the reward without incurring the costs of abandoning self-interested enforcement.Footnote ¹⁸ Second, consider the situation of a State which benefits from self-interested enforcement but fears that the harmful spillovers will provoke a powerful foreign actor, or a State which represents it, to resort to coercion. This is another situation in which a State considering enforcement will have an interest in surreptitious self-interested enforcement. Transparency provisions respond to concerns about concealed enforcement. These benefits of transparency will, however, have to be weighed against the costs. The most obvious costs of transparency are the direct costs of compiling and publishing information. Transparency in law enforcement also risks giving wrongdoers a strategic advantage by making it easier for them to calculate how to avoid sanctions.

2.3. Costs of alternative formulations

Like other legal obligations, enforcement obligations can be formulated as either obligations of result or obligations of conduct, and as being more like rules or standards. Obligations of result require a State to prevent a particular event from occurring without specifying the regulatory means by which that is to be accomplished, while obligations of conduct specify means but not results.Footnote ¹⁹ Meanwhile, the famous distinction between rules or standards turns on whether the content of the obligation is specified before (a rule) or after (a standard) the events it governs take place. In practice, many obligations take an intermediate form; some of their content in advance but leaving some matters to be defined when they are applied.

The literature on how parties draft ordinary contracts suggests that decisions about how to draft enforcement obligations in treaties will be based in part on the difficulty of successfully specifying and applying different formulations.Footnote ²⁰

Obligations of result often are relatively easy to draft, but it typically is difficult to determine whether they have been fulfilled, primarily because the results of law enforcement often are difficult to observe and verify. Regulated actors typically try to conceal prohibited activities such as drug trafficking, fraud, anti-competitive conduct, counterfeiting, pollution or violations of labour rights, and the effects of those kinds of activities may not be easy to observe, particularly when they are felt across borders. An additional concern about obligations of result is that the connection between actions and results is often uncertain and so States that make commitments to achieve specific results take on the potentially undesirable risk of being held liable for poor results for reasons beyond their control. Finally, there may be situations in which obligations of result are difficult to draft, in other words, when it will be difficult to define the activities the State ought to prevent. Noxious activities such as industrial-scale scamming are relatively easy to define, but defining the outer bounds of anti-competitive conduct and violations of labour rights may be more challenging.

As for obligations of conduct, drafting is complicated by the fact that it is difficult to specify the conduct required of enforcement officials in advance because enforcement is an extremely complex activity and the appropriate conduct varies depending on the circumstances, including the reactions of the targets of enforcement.Footnote ²¹ Imagine trying to specify everything a State ought to do to enforce its competition laws: the number of investigators, their qualifications, how much time they should devote to investigations in each industry, what penalties they should seek in each case, etc. The amount of detail involved would make the exercise prohibitively costly. And any static set of prescriptions would be vulnerable to strategic avoidance by wrongdoers, which would be particularly costly in the context of a treaty obligation since those are often difficult to revise. As a result, obligations of conduct often take the form of standards.

There are several ways of formulating obligations of conduct that avoid the need for detailed ex ante specification. For example, some enforcement obligations define the required conduct in terms of the results it is likely to achieve. An example is the Palermo Convention’s obligation to undertake ‘effective’ enforcement.Footnote ²² The effectiveness criterion is easy to specify but can be difficult to apply. People who disagree about causal relationships between enforcement strategies and results may disagree about whether a given enforcement strategy fulfils a commitment to aim at a particular result. Disagreement of this kind is commonplace.Footnote ²³ For instance, social scientists disagree about whether and how individuals and firms respond to the fear of State-imposed sanctions, social sanctions, social norms or perceptions of legitimacy. They also disagree about how States’ actions shape beliefs about sanctions and the other determinants of compliance with laws.Footnote ²⁴ These disagreements relate to how enforcement strategies influence firms’ and individuals’ decisions not only to comply with the law, but also to monitor, investigate, report or conceal the misconduct or potential misconduct of other actors. So, for example, a strategy of granting leniency to firms that have made efforts to investigate and report misconduct and which cooperate with investigations might be perceived in some quarters as ineffective because it undermines incentives to prevent misconduct. Meanwhile, in other quarters the same strategy may be seen as effective because it bolsters firms’ incentives to investigate, report and cooperate, facilitates prosecution of individuals, and bolsters the overall legitimacy of the system.Footnote ²⁵

Another way to avoid the need for detailed specification is to define the required enforcement strategy in relative terms, specifically by reference to a strategy employed by either another State or the same State in a different context, hoping that the reference strategy will be reasonably effective in the new context. A simple way to adopt a reference strategy is to copy enforcement obligations from another instrument.Footnote ²⁶ Another approach is to specify a benchmark enforcement strategy, for instance by making a commitment to non-discrimination, such as an agreement that cases involving foreign actors (whether as subjects or victims) will receive equal treatment to cases that involve domestic actors.

It is easy to specify a benchmark enforcement strategy in this way, but it may not be easy to determine whether a State has deviated from the benchmark. This is another ramification of the fact that enforcement strategies are often complex and context-sensitive. For example, suppose that the foreign firms operating in a country are typically, relative to domestic firms, larger, more likely to comply with consumer protection laws, and less likely to be managed by their owners. In this scenario it might be appropriate for the State to devote fewer enforcement resources to monitoring foreign firms, because they are less likely to engage in misconduct, but to impose higher penalties upon foreign firms because the penalties in cases involving domestic firms are aimed at the owners rather than the firms. Has the State treated foreign firms unequally? The answer depends on whether they have been subjected to the same enforcement strategy as large, relatively compliant, widely-held domestic firms. This is a complex inquiry, and it may involve a great deal of speculation if no domestic firms are comparable to foreign firms along the relevant dimensions.

Finally, some enforcement obligations regulate the process by which enforcement strategies are developed rather than attempting to regulate the content of the strategies. This includes obligations that bar States from considering certain factors when making discretionary enforcement decisions. These kinds of enforcement obligations can serve to impose liability on States whose enforcement authorities knowingly or deliberately seek to achieve a competitive advantage for domestic firms. Evidence of intention to cause a prohibited result (under-enforcement or discriminatory enforcement) can be a useful proxy for more direct evidence. Obligations to exercise enforcement discretion in good faith or in a non-arbitrary way also can serve to impose liability upon States whose decisions are not based on legitimate considerations. A drawback of these more standard-like procedural obligations is that they may place the onus upon States to incur the cost of articulating the reasons for their enforcement decisions. As noted above, this kind of transparency can be costly.

2.4. Why do enforcement obligations vary across States and areas of law?

Variations in the content and form of enforcement obligations can be explained by variations in States’ ability or willingness to commit to any given enforcement strategy. First, binding constitutional or resource constraints might rule out certain types of commitments for certain States. Constitutional prohibitions upon specific investigative techniques or upon excessive or disproportionate sanctions may limit the State’s ability to detect and sanction misconduct, and the obstacles to amendment of the relevant provisions often will be insurmountable. At the same time, shortages of skilled prosecutors, investigators or judges may limit the number of cases a State can detect or prosecute.

Second, when resources are available but scarce, the costs and benefits of redeploying those resources to focus on activities that harm foreigners will vary across States depending on the nature of the resources and the levels of purely domestic misconduct.Footnote ²⁷ For instance, the costs and benefits of deploying a State’s financial investigators to investigate corruption rather than the laundering of proceeds of foreign crimes will depend on both the qualifications of the investigators and the society’s level of corruption.

Third, the same enforcement strategy might yield different benefits in different States because of differences in either societal values or the substantive norms being enforced.Footnote ²⁸ For example, a State with relatively high environmental standards generally will find it much more costly to comply with an obligation to ‘effectively enforce’ its environmental laws than a State with weaker standards.Footnote ²⁹

Fourth, States also may vary in terms of the rewards or threats they have received from other States, or the value they assign to those rewards or threats. For all these reasons the commitments States are willing to make regarding enforcement practices are likely to vary, both across States and areas of law.

2.5. Purposive interpretation of enforcement obligations

Understanding why States assume enforcement obligations is an important step toward understanding how those obligations ought to be interpreted. Article 31 of the Vienna Convention on the Law of Treaties (VCLT) supports this step from positive to normative analysis to the extent it provides that, ‘[A] treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.’Footnote ³⁰ The claim here is simply that the object and purpose of many treaties that contain enforcement obligations is to limit the cross-border effects of self-interested enforcement, including through prevention of under-enforcement, discriminatory enforcement and concealed enforcement.

Interpreting enforcement obligations in light of a treaty’s object and purpose typically will entail paying close attention to the texts the parties have adopted to express their commitments.Footnote ³¹ As argued above, States’ purposes in assuming enforcement obligations are likely to vary across States and contexts. In order to enable States to make the widest possible variety of commitments they should be given a verbal palette that allows them to convey as many different shades of meaning as possible. In practice this means that interpretation should involve close attention to the text of each obligation, as well as the surrounding context, without any presumption that similar but slightly different texts were meant to be interpreted uniformly. However, when the meaning of the text is indeterminate, which is likely to be a common occurrence given the obstacles to specification detailed above, choices between alternative interpretations should be guided by the insight that enforcement obligations are generally intended to mitigate the cross-border effects of self-interested enforcement by preventing under-enforcement, discriminatory enforcement or concealed enforcement.

3.1. Overview of the OECD Anti-Bribery Convention

The main substantive provisions of the OECD Anti-Bribery Convention require parties to criminalise the bribery of foreign public officials (as well as related forms of improper accounting and, in certain circumstances, money laundering).Footnote ³² The adoption of the Convention in 1997 followed years of lobbying and, eventually, coercion, from the US, which had criminalised foreign bribery and related offences many years earlier, under the FCPA.

The fact that the US enacted and enforced the FCPA unilaterally suggests that there are certain situations in which prosecution of foreign bribery is consistent with a State’s self-interest. The US adopted the FCPA in order to make a moral statement to the rest of the world in the aftermath of the Watergate scandal, to avoid rewarding inefficient and unethical US companies, and, perhaps in part, to give its companies an excuse to resist demands for paying bribes for official favours that officials might be willing to provide anyway.Footnote ³³ In later years the US Government came to embrace the FCPA as a way of promoting the US’ enlightened self-interest in combatting corruption in foreign countries so as to promote democracy, the rule of law and economic development.Footnote ³⁴ The US regularly enforced the FCPA after its enactment, but the level of enforcement jumped dramatically after the adoption of the OECD Anti-Bribery Convention.Footnote ³⁵

As the history of the FCPA suggests, although enforcement of laws prohibiting foreign bribery laws generates some purely domestic costs and benefits, it may have different effects on foreign actors. First, to the extent the targets of enforcement are domestic actors, many of the direct costs of enforcement are borne domestically. In recent years, substantial penalties—ranging up to over a billion dollars—have been imposed in FCPA cases, and defendant firms typically also experience significant reputational harm.Footnote ³⁶ Domestic actors exposed to this kind of liability are at a competitive disadvantage to foreign actors that do not face similar legal risk. Second, to the extent that it targets foreign rather than domestic actors a State can use enforcement of foreign bribery laws to place the targeted foreign firms at a competitive disadvantage. Third, many of the benefits of enforcement, such as promotion of the rule of law and economic development, accrue to the State represented by the foreign official implicated in the bribery, and not all enforcing States will equate preventing corruption in a foreign country or providing compensation to foreign victims with their own self-interest. Fourth, some of the costs of enforcement, specifically, those which flow from deterrence of investment in jurisdictions where it is difficult to do business without bribing public officials, are borne by actors operating in those high-risk jurisdictions.

The first three factors listed above give States incentives to engage in under-enforcement or discriminatory enforcement of their foreign bribery laws, at least from the perspective of other States interested in stamping out corruption or maintaining a level competitive playing field. That seems like an apt characterisation of the interests of the parties to the OECD Anti-Bribery Convention.Footnote ³⁷ Accordingly, those States had interests in making reciprocal commitments to refrain from under-enforcement and discriminatory enforcement.Footnote ³⁸ The last factor listed above actually creates incentives for States to engage in over-enforcement, at least from the perspective of high-risk jurisdictions interested in attracting investment. However, none of the original parties to the OECD Anti-Bribery Convention fit that description and so it is safe to presume that the Convention was not intended to constrain over-enforcement. For similar reasons, there is no reason to presume that the parties’ commitments to refrain from under-enforcement extend to enforcement strategies that would be optimal from the perspective of non-parties to the Convention.Footnote ³⁹ Therefore, the following analysis will proceed on the assumption that the parties intended the Convention’s enforcement obligations to constrain discriminatory enforcement and certain forms of under-enforcement.

3.2. Enforcement obligations in the OECD Anti-Bribery Convention

The enforcement obligations created by the OECD Anti-Bribery Convention are found in multiple places, but the foundational principles are contained in Article 5:

The parties to the OECD Anti-Bribery Convention adopted it simultaneously with an official set of Commentaries. Article 27 of the Commentaries states that Article 5 ‘recognises the fundamental nature of national regimes of prosecutorial discretion’ and that ‘in order to protect the independence of prosecution, such discretion is to be exercised on the basis of professional motives and is not to be subject to improper influence by concerns of a political nature’.Footnote ⁴¹

In 2009, the parties to the Convention adopted a Recommendation (which was subsequently amended and expanded) that recommended that parties be ‘vigilant’ about ensuring compliance with Article 5 and then went on to provide that:

The OECD Anti-Bribery Convention also includes an important transparency provision, Article 12, which provides for monitoring and follow-up by the Working Group. The Working Group has interpreted its mandate boldly, most notably by developing an elaborate peer review system which results in periodic reports which assess various aspects of parties’ compliance with the Convention, including enforcement efforts.Footnote ⁴³ Each round of peer reviews is known as a ‘Phase’. Beginning with Phase 3, which was completed in 2014, the process has focused on enforcement.Footnote ⁴⁴ Some scholars believe that the peer review system has played a role in motivating countries to improve their enforcement by promoting transparency, mobilising pressure from peers and the public and stimulating learning and capacity building, but there is evidence that political will informed by cost-benefit calculations is of overriding significance.Footnote ⁴⁵

As will be demonstrated below, the OECD Anti-Bribery Convention’s provisions are unclear along several dimensions, thus bringing into play the VCLT’s invitation to refer to context, together with subsequent agreements and practices, as well as the treaty’s object and purpose, for interpretive guidance.Footnote ⁴⁶ Here referring to context means taking into account not only the text of the Convention but also contemporaneous instruments such as the Commentaries.Footnote ⁴⁷ In addition, since all of the parties to the Convention are members of the Working Group, its recommendations and statements can be regarded as evidence of subsequent agreements and practices of the parties. However, the Convention does not provide any mechanism for resolving disputes about its interpretation or application, and the parties’ agreements and practices provide limited guidance in those areas. Accordingly, it also will be helpful to refer to the Convention’s object and purpose.

The ambiguities in the OECD Anti-Bribery Convention’s enforcement provisions begin with the first sentence of Article 5: ‘[I]nvestigation and prosecution of the bribery of a foreign public official shall be subject to the applicable rules and principles of each Party.’ Does this provision permit a State to deviate from domestic standards in a foreign bribery case? For example, what if prosecutors in foreign bribery cases abandon a standard practice of granting leniency to first-time corporate offenders? Taken literally, the text of Article 5 seems to prohibit enforcement strategies that deviate from the domestic benchmark, even if they promise to be more effective. But there is no obvious reason why the parties would want to prohibit one another from investigating and prosecuting foreign bribery more effectively than they investigate or prosecute domestic bribery, at least when the target is a domestic firm. In other words, it is implausible that the domestic benchmark defines an upper bound on the permissible effectiveness of enforcement.

There are two alternative interpretations of Article 5’s first sentence that seem more plausible. The first possibility is that States are not required (but are permitted) to adopt enforcement strategies precluded by internal legal norms of general application, such as constitutional limitations on investigative practices. Second, States may be required to investigate and prosecute foreign bribery at least as effectively as comparable domestic cases. The first of these interpretations treats domestic standards as a safe harbour against claims of under-enforcement rather than the benchmark for an affirmative obligation. The second interpretation creates a commitment not to discriminate between foreign bribery cases and other cases. Both of these purposive interpretations of Article 5 seem preferable to the more literal interpretation, taking into account both the commentary on Article 5 and the reasons why treaties typically include commitments regarding law enforcement. However, the Working Group’s recent call for South Africa to take additional measures to ensure prosecutorial independence ‘whether through legislative amendment or constitutional reform’ implicitly rejects the idea that Article 5 contains a safe harbour for States that comply with domestic law.Footnote ⁴⁸

Turning to the second sentence of Article 5, the reference in the first clause to ‘national economic interest’ is both ambiguous and vague. The ambiguity lies in the fact that the concept of ‘national economic interest’ admits at least three distinct interpretations. First, the prohibition on being influenced by considerations of ‘national economic interest’ could reasonably mean that enforcement agencies are permitted to consider only foreign economic interests.Footnote ⁴⁹ This would permit a State to treat a firm leniently in order to, for example, preserve foreign jobs, but not to preserve domestic jobs. A second interpretation would avoid this odd result by reading the clause to bar consideration of any economic interests, foreign or domestic. This reading might be consistent with the moralistic notion that bribery is so reprehensible that it ought to be condemned at all costs but would require ignoring the provision’s explicit reference to ‘national’ economic interest, contrary to the principle of effectiveness. Alternatively, a third possibility is that while States are barred from considering exclusively national economic interests, they are permitted to consider ‘national and foreign economic interests’. Or to put it another way, States are barred from considering ‘nationalistic’ economic interests. Under this last interpretation a State could treat a firm leniently in order to preserve domestic jobs as long as it would grant the same leniency in order to preserve foreign jobs. In other words, the clause could be read as a commitment not to discriminate in favour of domestic economic interests.

Despite these ambiguities, in some cases it is clear what conduct is covered by Article 5’s prohibition on consideration of national economic interest. For instance, in 2025, the US adopted guidelines on FCPA enforcement that explicitly prioritise investigation and prosecution of conduct that undermines US national security and economic prosperity.Footnote ⁵⁰ Those guidelines seem to violate any plausible interpretation of Article 5. Other cases are, however, less clearcut.

Neither the text nor the context provide clear guidance on whether the second sentence of Article 5 bars all consideration of national economic interest or only prohibits exclusive consideration of national economic interests. Unlike the moralistic reading, both interpretations are consistent with the ordinary meaning of the Convention’s terms. The Commentary is of little assistance because it says that the intention was to rule out unprofessional, ‘improper’ and ‘political’ considerations, but it is unclear whether those include economic interests simpliciter.Footnote ⁵¹

Reference to subsequent practice of the Convention’s parties also fails to resolve the ambiguity. In its Phase 4 evaluation of Canada the OECD Working Group, referring to the SNC-Lavalin case, stated that taking the potential loss of Canadian jobs into consideration in deciding how to resolve a foreign bribery prosecution would be contrary to Article 5’s prohibition on consideration of national economic interest.Footnote ⁵² The Working Group did not suggest that it would have been acceptable to consider job losses in Canada alongside losses in other countries, implicitly ruling out the third interpretive option suggested above. But the Working Group’s earlier report on the US struck a different note. US federal prosecutors’ internal guidelines explicitly instruct them to consider the adverse impact of a criminal conviction upon innocent parties such as employees, investors, pensioners and customers when deciding whether to pursue criminal charges against a corporation and how to resolve cases involving corporate defendants.Footnote ⁵³ In its Phase 3 evaluation of the US, the Working Group concluded that this aspect of US enforcement policy was consistent with Article 5 because US prosecutors confirmed that they would ‘consider the potential harm to innocent third parties in all cases, including those involving non-U.S. companies and individuals’.Footnote ⁵⁴ This conclusion implies that Article 5 permits even-handed consideration of economic interests.

The functional account of enforcement obligations set out above provides a solution to this interpretive dilemma. Specifically, it suggests that the object and purpose of the Convention support interpreting Article 5 to permit non-discriminatory consideration of national economic interests. All of the possible interpretations of Article 5 provide protection against the most important forms of discriminatory enforcement, namely, those aimed at promoting domestic economic interests. The difference is that where foreign economic interests are not in play, the bar on all consideration of national economic interests would also serve as a crude way of restricting under-enforcement. For instance, it would bar a country from protecting a national champion simply to protect local jobs, regardless of whether there is evidence that a firm with weaker local ties would have been treated differently. It is unlikely that the parties to the OECD Anti-Bribery Convention intended for Article 5 to target under-enforcement with such a sweeping prohibition. While there is considerable evidence that they were interested in discouraging patterns of enforcement that would give some parties’ firms advantages over others (i.e. discriminatory enforcement), it is not clear that they were committed to incurring substantial economic costs to prosecute foreign bribery.Footnote ⁵⁵ More specifically, nothing in the text signals that the parties intended to commit themselves to incurring unlimited economic costs to prosecute bribery. A more plausible conclusion is that Article 5 was not intended to regulate under-enforcement at all. Rather, the parties addressed under-enforcement in the Commentary and the 2009 Recommendation by making commitments to take complaints ‘seriously’ and to provide ‘adequate’ resources to ensure ‘effective’ investigation and prosecution.Footnote ⁵⁶

Interpretation of Article 5’s prohibition on consideration of ‘national economic interest’ is complicated by vagueness as well as ambiguity. The text does not specify either what counts as an economic interest or when the economic interest of a firm or an individual counts as a ‘national’ interest. A purposive interpretation suggests, however, that the parties intended to include any economic considerations that might lead a State to enforce its laws in a way likely to prejudice foreign actors. Wealthy democratic States like the members of the OECD are typically motivated by a range of economic considerations. They are likely to be sensitive to economic benefits that flow directly to the State, such as revenues from State-owned enterprises or taxes collected from firms and individuals in the jurisdiction. They also are likely to respond to changes in the economic fortunes of potential voters or donors. Consequently, national economic interests should be interpreted broadly to include not only the interests of State-owned enterprises or government agencies, but also the economic welfare of citizens as well as firms with ties to the State.Footnote ⁵⁷

Moving on, Article 5’s second sentence also condemns States whose investigation or prosecution of foreign bribery is influenced by ‘the potential effect upon relations with another State’. The purpose of this provision is relatively clear. It seems to apply whenever a State’s enforcement is influenced by threats or offers, or the expectation of retaliation or rewards, from another State.Footnote ⁵⁸ Banning this form of discrimination presumably is meant to both encourage States to maintain a high level of enforcement and to discourage unproductive competition for favourable treatment. Ironically, this is the only part of Article 5 that has received substantial judicial attention. In December 2006, the UK closed a foreign bribery investigation aimed at BAE, a major British defence contractor, in response to a threat from Saudi Arabia to terminate cooperation in counter-terrorism and other matters.Footnote ⁵⁹ The Government was convinced that as a result of the Saudis’ threat, ‘British lives on British streets were at risk’.Footnote ⁶⁰ There is little doubt that the decision to terminate the investigation amounted to a violation of Article 5, at least if the text is given its ordinary meaning.Footnote ⁶¹ There were, however, reasonable arguments that i) it was inappropriate for the British courts to enforce the commitment, ii) there was no violation because the provision should be interpreted in light of the object of the parties to the Convention, and if they had turned their minds to the issue the parties would have excused compliance in a case where it would cause such hardship, or iii) compliance was such a hardship that it should have been excused under the doctrines of necessity or distress.Footnote ⁶² When the UK government announced that it was terminating its investigation of BAE, the OECD Working Group expressed ‘serious concerns as to whether the decision was consistent with the OECD Anti-Bribery Convention’.Footnote ⁶³ After the UK’s next full peer review the Working Group said, ‘The lead examiners are disappointed that the UK did not consider re-opening the Al Yamamah investigation.’Footnote ⁶⁴

The final clause of Article 5 is a bit mysterious. It bars States from being influenced in investigation or prosecution by ‘the identity of the natural or legal persons involved’. A person’s identity is a subset of attributes—name, size, location, national origins, social ties, age, etc—that serves to distinguish that person from other persons. This language was clearly intended to bar States from considering some identifying attributes in the course of decisions about investigation or prosecution, but it is unclear which ones. The commentary to Article 5 suggests that the purpose of this provision is to rule out ‘improper’ ‘political’ influences.Footnote ⁶⁵ Unfortunately, neither of these terms is self-defining. It also is unclear which persons are considered to be ‘involved’ in an investigation or prosecution. Is it improper to consider the identity of just the target or is consideration of the identity of the persons harmed by the conduct also proscribed? The ambiguity of this provision is problematic because States might have strong reasons for adopting enforcement strategies that discriminate on the basis of some identifying attributes. For example, it may be appropriate to treat large consumer products firms differently when it comes to initiation of criminal proceedings on the grounds that the stigma associated with criminal proceedings is relatively damaging for them. Therefore, a State could reasonably argue that differential treatment along this dimension is necessary to ensure that consumer firms are exposed to the same regulatory burdens as other firms. The purposive approach to interpretation allows room for consideration of these arguments.

A review of the OECD Working Group’s monitoring and reporting suggests that its members are deeply interested in regulating one another’s enforcement activities.Footnote ⁶⁶ The Working Group is particularly active in monitoring for under-enforcement. Its peer review reports occasionally criticise countries for imposing low financial penalties on corporate defendants.Footnote ⁶⁷ They also regularly compare the number of foreign bribery cases investigated or prosecuted to the reviewers’ perceptions of the number of cases that could have been investigated or prosecuted.Footnote ⁶⁸ The reports explicitly indicate that perceptions of the number of potential investigations or prosecutions are based on the size of the country’s economy and its economic ties to high-risk countries and sectors.Footnote ⁶⁹ The Working Group’s peer review reports also routinely evaluate the resources and expertise of bodies charged with enforcement, often concluding that they are inadequate.Footnote ⁷⁰ Data on prosecutions for domestic bribery are sometimes used to assess either a country’s relative commitment to sanctioning foreign bribery or its overall commitment to enforcement of anti-corruption laws.Footnote ⁷¹

The Working Group’s monitoring also captures discriminatory enforcement, or at least factors that might enable it. The peer review reports frequently call out institutional arrangements that might permit improper considerations to influence enforcement—usually ones which permit elected officials or Cabinet ministers to influence prosecutorial decisions.Footnote ⁷² The reports regularly raise concerns about the handling of specific foreign bribery cases and whether they may have been improperly influenced by ‘Article 5 factors’.Footnote ⁷³ Strikingly though, only a handful of reports—concerning Argentina, Italy Portugal and South Africa—refer to specific improper considerations.Footnote ⁷⁴ Accordingly, although the Working Group’s practices make it clear that the parties are concerned about under-enforcement and discriminatory enforcement, they shed little light on the meaning of the Convention’s most ambiguous provisions. The Working Group’s general reluctance to invoke the provisions on specific improper considerations also may be a sign that the language of those provisions is not particularly helpful.

3.3. SNC-Lavalin

The potential significance of Article 5 of the OECD Anti-Bribery Convention and its ambiguities were dramatically illustrated in the course of Canada’s SNC-Lavalin affair.

SNC-Lavalin, as it was then known, is a large engineering and construction company headquartered in Montreal. On 19 February 2015, the Royal Canadian Mounted Police (RCMP) charged SNC-Lavalin and two related companies with bribery of foreign public officials and fraud.Footnote ⁷⁵ The foreign bribery charges stemmed from allegations that SNC-Lavalin had paid bribes to Libyan public officials between 2001 and 2011. If convicted of foreign bribery the companies would be automatically barred from contracting with the government of Canada for at least 5 years and the Government would be entitled to cancel existing contracts.Footnote ⁷⁶ At the time, SNC-Lavalin held several substantial contracts with the federal Government.Footnote ⁷⁷

By 2016, SNC-Lavalin had begun to lobby several cabinet ministers, including Prime Minister Justin Trudeau, for favourable treatment. As a first step, the company pressed the Government to enact legislation that would permit organisations charged with certain offences, including foreign bribery, to avoid a criminal trial and conviction by negotiating a ‘remediation agreement’.Footnote ⁷⁸ Several other countries, including the US and the UK, permitted these kinds of arrangements, often known as ‘deferred prosecution agreements’, and by 2016 the US Department of Justice had begun to rely upon them extensively in enforcement of the FCPA.Footnote ⁷⁹ In 2018, after a period of consultation, the Trudeau Government introduced and passed legislation amending the Criminal Code to provide for remediation agreements.Footnote ⁸⁰ The amendments were included in an omnibus budget bill, a sign that the Government wished to have them enacted quickly.Footnote ⁸¹

One of the stated objectives of the amendments was ‘to reduce the negative consequences of the wrongdoing for persons—employees, customers, pensioners and others—who did not engage in the wrongdoing, while holding responsible those individuals who did engage in that wrongdoing’.Footnote ⁸² A remediation agreement would have the effect of staying charges against the firm while providing for an admission of wrongdoing, penalties, reparations, cooperation with enforcement authorities and, where appropriate, adoption of specified compliance measures and appointment of an independent monitor.Footnote ⁸³ Crucially, a firm that entered into a remediation agreement would avoid a criminal conviction and so would not be automatically barred from contracting with the federal Government.

The new provisions permitted a prosecutor to negotiate a remediation agreement with an organisation accused of foreign bribery and other specified offences if certain conditions were met. The conditions included the consent of the relevant Attorney General and that the prosecutor be ‘of the opinion that negotiating the agreement is in the public interest and appropriate in the circumstances’.Footnote ⁸⁴ The prosecutor was required to consider a long list of factors for the purposes of forming this opinion. More importantly for present purposes, the legislation provided that for organisations accused of foreign bribery, ‘the prosecutor must not consider the national economic interest, the potential effect on relations with a state other than Canada or the identity of the organization or individual involved’.Footnote ⁸⁵ This provision was clearly designed to track the language of Article 5 of the OECD Anti-Bribery Convention and similar language in the Code of Practice for deferred prosecution agreements issued by the UK enforcement authorities.Footnote ⁸⁶ The legislation provides no indication how to reconcile the explicit prohibition on considering the national economic interest with its stated objective of protecting innocent employees, customers and pensioners.

SNC-Lavalin’s hopes of benefitting from a remediation agreement were dashed by the Director Public Prosecutions (DPP), the head of Canada’s Prosecution Service, who, at the beginning of September 2018, decided not to negotiate a remediation agreement with the company.Footnote ⁸⁷ The Attorney General at the time refused to overrule this decision, citing the principle of prosecutorial independence.Footnote ⁸⁸ In the Canadian tradition, the Attorney General was an elected member of Parliament appointed to her cabinet position by the Prime Minister. The applicable quasi-constitutional principles provided that the Attorney General was solely responsible for decisions on whether to authorise criminal prosecutions. She was entitled to consult other members of the Government but was not to be put ‘under pressure’ by them.Footnote ⁸⁹ SNC-Lavalin sought judicial review of the DPP’s decision not to offer a remediation agreement, but its claim was dismissed on the ground that the decision was an unreviewable exercise of prosecutorial discretion.Footnote ⁹⁰

After being informed of the DPP’s decision, SNC-Lavalin redoubled its lobbying efforts. It led the Prime Minister to believe that if it was convicted and barred from federal contracting it would be compelled to reduce its operations in Canada, at the cost of approximately 9,000 Canadian jobs.Footnote ⁹¹ The Prime Minister and other members of his cabinet feared that this would have broad economic repercussions, especially in Quebec where the job losses would be concentrated. They also worried about the political ramifications, at both the federal and provincial levels of government. They were concerned that a negative economic shock of this kind could affect the upcoming provincial election in Quebec. It was also significant that the Prime Minister’s own riding, or parliamentary constituency, was in Montreal, in close proximity to SNC-Lavalin’s headquarters.Footnote ⁹² Over the course of several months, the Prime Minister and other cabinet ministers, with the support of SNC-Lavalin’s lobbyists, attempted to persuade the Attorney-General to reconsider her decision, but to no avail.Footnote ⁹³ In January 2019, the Prime Minister moved the Attorney General to another ministry. She eventually resigned from the Cabinet and was expelled from the governing Liberal Party’s caucus.

The following month, news of the lobbying effort was reported in the media.Footnote ⁹⁴ The next day, at the request of opposition politicians, Canada’s Conflict of Interest and Ethics Commissioner opened an investigation into whether Prime Minister Trudeau had violated conflict of interest legislation by seeking to influence the Attorney General ‘to improperly further another person’s private interests’.Footnote ⁹⁵ A Parliamentary committee also launched an inquiry and held a series of high-profile hearings.Footnote ⁹⁶ The OECD Working Group issued a press release saying that it was ‘concerned’ about the allegations and would monitor the proceedings.Footnote ⁹⁷ In August 2019, the Conflict of Interest and Ethics Commissioner issued a report concluding that the Prime Minister had broken the law.Footnote ⁹⁸ The next month the Government called an election. In October, Trudeau’s Liberal Party lost its Parliamentary majority but managed to form a minority Government. The scandal was widely perceived to have caused substantial political harm to the Liberal Party.Footnote ⁹⁹

In December 2019, two months after the election, a subsidiary of SNC-Lavalin pleaded guilty to a single count of fraud. It paid a fine of Canadian $280 million, by far the largest corporate penalty for fraud or foreign bribery in Canadian history, and was subjected to a three-year probation order.Footnote ¹⁰⁰ SNC-Lavalin issued a press release stating that it did not expect the sanction to affect its ability to bid on any future projects.Footnote ¹⁰¹ A few days before the company’s guilty plea, a jury convicted one of its former senior executives on related charges of fraud and corruption. He was sentenced to a term of imprisonment of eight years and six months, which was also of record severity for Canada.Footnote ¹⁰²

Was the OECD Working Group right to be concerned about whether the Trudeau Government’s handling of the SNC-Lavalin affair was compatible with Article 5? Did it correctly suggest that it would have been inappropriate for the DPP and the Attorney-General to consider potential job losses in interpreting the Canadian legislation based on Article 5?Footnote ¹⁰³ Perhaps. The purposive approach to interpretation suggests that the key issue is whether there was evidence of either under-enforcement or discriminatory enforcement.

At first glance, the evidence of under-enforcement is equivocal. On the one hand, the facts that liability was imposed on a subsidiary rather than the parent company and that the company was allowed to avoid the stigma of being convicted of bribery certainly reduced the impact of the Canadian Government’s sanctions.Footnote ¹⁰⁴ On the other hand, SNC-Lavalin ultimately was hit with record-breaking penalties.

The mere fact that a prosecutor seeks a negotiated resolution instead of a conviction does not necessarily mean that the level of enforcement has been reduced in any meaningful sense. In principle, for any given criminal conviction there ought to be a negotiated resolution—perhaps with a higher penalty to reflect the stigma of conviction—which has the same deterrent, rehabilitative and compensatory effects but with less harm to innocent parties. There is no reason to believe that the parties to the OECD Anti-Bribery Convention intended to rule out this kind of win-win option. The parties to the Convention have an interest in permitting prosecutors to pursue the objective of discouraging foreign bribery at the lowest cost to innocent parties. It is implausible that they would prefer an interpretation of Article 5’s prohibition on consideration of national economic interest that rules out such an approach simply because the innocent parties happen to be in the same country as the enforcement agency. Or to put it another way, none of the parties would have an interest in offering Canada inducements to substitute convictions for negotiated resolutions.Footnote ¹⁰⁵

A more difficult question is whether the SNC-Lavalin affair involved discriminatory enforcement. It is impossible to know for sure without being able to compare the handling of this case to the handling of an identical case involving a foreign firm. Although the Prime Minister’s lawyers argued to the contrary, consideration of potential job losses ought to qualify as consideration of economic interests.Footnote ¹⁰⁶ However, the fact that the Trudeau government tried to take economic interests into account in its efforts to influence the prosecution of SNC-Lavalin is not dispositive. A prosecutor who learns that 9,000 jobs will be lost if she obtains a conviction against a firm can take that into consideration and decide to negotiate an alternative resolution of the case without discriminating between foreign and domestic firms. As long as the prosecutor gives the same consideration to jobs at risk in Canada and jobs that might be lost elsewhere there is no discrimination against foreign interests.

To be clear, non-discrimination in this context implies that economic gains as well as losses should be weighed without regard to national interest. This might be relevant if a conviction would cause the firm to shut down its operations in Canada and reopen them in another location. The hypothetical Canadian prosecutor may wish to consider whether economic losses from conviction are offset by economic gains. If she does, then the offset should be applied without regard to whether those gains accrue to Canadians or foreign interests, especially if those foreigners are based in States which are party to the Convention.

All of this suggests that Article 5’s prohibition on being influenced by consideration of national economic interest should not necessarily have been an obstacle to offering SNC-Lavalin a remediation agreement. That language should have been interpreted to bar consideration of exclusively national economic interests, but not of all economic interests. Any domestic legislation intended to implement this aspect of Article 5 should have been interpreted in the same way.

The Trudeau Government may, however, have violated other provisions of Article 5. Recall that the opening sentence states that ‘[I]nvestigation and prosecution of the bribery of a foreign public official shall be subject to the applicable rules and principles of each Party.’ If this provision imposes an affirmative obligation (as opposed to creating a safe harbour for States that abide by standard practices), then to the extent the Trudeau Government violated Canadian norms of prosecutorial independence—which it did, according to the Conflicts of Interest and Ethics Commissioner—it violated Article 5. This is especially true if the violation was motivated by partisan political reasons, since the Commentary on Article 5 indicates that it is intended to protect prosecutorial independence against influence by ‘concerns of a political nature’.Footnote ¹⁰⁷ Similarly, if the Government was motivated by the desire to protect SNC-Lavalin in particular, and would not have been equally solicitous of another firm, then it might have violated Article 5’s bar on being influenced by ‘the identity of the natural or legal persons involved’. These kinds of deviations from standard decision-making processes represent at least circumstantial evidence of discrimination or under-enforcement. Accordingly, it is not difficult to understand why parties to the OECD Anti-Bribery Convention would want to prohibit these kinds of behaviours. This way of applying Article 5 would also track the terms of the legal and political debate in Canada, where critics of the Government’s actions generally framed them as a scandalous breach of norms of prosecutorial independence for partisan purposes.Footnote ¹⁰⁸