The Dis-Embedded Arbitrator: Releasing Arbitration from Corruption-Shaped Environments in the Wake of the Odebrecht Arbitral Ordeal in Peru

Abstract

Despite local instances of single arbitrators’ corruption not having proven completely absent from arbitration chronicles over the last decades, one may safely argue that until very recently, no scandal had ever been severe enough to shake the foundations of arbitration communities on a regional, let alone global, level. However, this eventually occurred in 2019 in Peru as the outcome of one of the countless parallel investigations stemming from the 2016 Odebrecht corruption saga, propagated from Brazil to the whole of Latin America, the Caribbean, sub-Saharan Africa, and beyond, and labelled by many as the largest scandal of its kind in recent history. Peru’s vicissitudes revolved around a number of corrupted arbitrators who systematically accepted bribes and political favours from Odebrecht in return for favourable awards upholding the repricing of public-procurement contracts. This story can teach us about more than the simple evidence that arbitrators, too, might fall for corruption; criminologically, it suggests that arbitration as a dispute-resolution mechanism can find itself embedded within regionalised networked systems of corruption-prone regulatory capture, and even play an active role in their normalised perpetuation. To prevent this, while having regard for safeguarding the independence and confidentiality of arbitral proceedings to the highest possible extent, the enactment of context-sensitive binding regulation is advised.

1. Introduction

In 2016, the gravest corruption scheme in recent history surfaced in Brazil, after having socialised its criminal practices all throughout Latin America and the Caribbean; the Odebrecht scandal had begun, and at the time of writing, in July 2021, its effects are yet to be realised as to the whole range of their legal, economic, and sociocultural implications. Stemming, in turn, from another corruption saga traversing the same region (the Petrobras scandal), the Odebrecht scandal unveiled the normalisation of corruption endeavours successfully applied by a construction conglomerate, Odebrecht, throughout the region and far beyond, over several years, vis-à-vis thousands of corporate-captured policymakers, regulators, bureaucrats, and other professionals.

Complicit to these extensive networks of bribery, arbitrators in Peru, Colombia, and other jurisdictions have been convicted for accepting money and favours in return for favourable awards aimed at contract levelling-up repricing. And yet, ‘[f]or all of the economic and political upheaval that the corruption scandals have created, there has been surprisingly little academic analysis of how these crises have changed the policy landscape in South America’ (Brewster and Ortiz 2021, p. 107), including the regulatory environment for arbitrators and the way the law addresses arbitration as a dispute-settlement system alternative to domestic courts in context. This work is premised on filling this lacuna, by introducing and conceptualising the figure of the “embedded arbitrator”, who is not simply an arbitration professional who occasionally or even systematically accepts bribes on an individual level, but rather one who is so embedded within a well-oiled system of mutual capture between state and corporate entities within a specific jurisdiction that they themselves become essential links-in-the-chain for the perpetuation of the mentioned capture. Although these phenomena of mutual capture reflect domestically the cross-continental reality of transnational elites, as evidenced, e.g., by Odebrecht’s exchanges with Angola’s rulers through Brazilian ones (Alencastro 2019), the embedded commercial arbitrator operates on a narrower scale and tends to turn into a component of the well-engineered machinery of domestic regulatory capture.

Consequently, building on the eye-opening scandal-within-the-scandal of Peruvian arbitrators accepting bribes from Odebrecht, this work investigates their embeddedness within Peru’s regulatorily captured institutions, their public significance, and the limits of treating these phenomena as instances isolated from one another. Criminology and sociolegal studies meet at the confluence between public capture and arbitral practices, the salient question being how to prevent not simply the act of corruption, but the far more complex (and worrisome) embeddedness of arbitrators within corrupted logics of power and mutual favouritism that discredit arbitration as an alternative to courts within the constellation of dispute-settlement mechanisms. The present paper accepts the challenge of offering initial views on this uncapturing exercise through the enactment of provisions customised for the seriousness and “nature” of corruption in each relevant jurisdiction. One disclaimer is due from the outset: Corporate-to-corporate arbitrations fall outside the scope of this endeavour, and international—whether commercial or investment—arbitrations will not be directly addressed, either (unless for comparative or illustrative purposes). The final policy suggestions as formulated hereinafter will exclusively concern domestic, state-to-corporation, commercial arbitrations, and mostly ad hoc rather than institutional ones.

The next section, Section 2, inspects the regional features of corruption in Latin America, emphasising the collusion between private entities and public bodies and its normalisation beyond the “white-collar” phenomenology of crime.1 Section 3 follows up with a factual account of Odebrecht’s embedded corruption within the region, touching upon its systemic consequences for the capture of Brazil’s politically driven judicial activism. The ensuing section, Section 4, temporarily steps into the investment arbitration’s domain to make a point which conceptually introduces the upcoming sections on commercial arbitration: when it comes to captured state-to-corporation arbitrations, no one’s hands are clean, meaning that to understand and untangle the phenomenon criminologically as well as legally, one shall focus on both sides’ embeddedness rather than confining themselves to noting which party was bribed by the other, and whether arbitrators could play a role in preventing or at least not aggravating said corruption. Section 5 delves into the gist of this work: the sub-scandal in Peru. It explains how it worked, its main characters, and especially the misdirected overreaction by Peruvian prosecutorial authorities. The (perhaps hypocritical) rush for emergency executive measures in Peru to address arbitrators’ corruption is criticised in Section 6, my argument being that most of such measures will prove either naive, ineffective, or both, exactly because they fail to complexify “corruption instances” as part of broader double-edged professional embeddedness within regulatorily captured jurisdictions. At this stage, one could hypothesise that establishing civil-liability regimes for arbitrators, a somewhat trendy move in recent years, would provide a welcome deterrent; Section 7 argues that this would stand as a short-sighted approach because, again, it is centred on the individual arbitrator as such rather than (also) on its embeddedness within the formal and informal system of legal and societal codes and expectations it abides by. My three embryonic proposals to address arbitrators’ embeddedness, with their foreseeable promise and shortcomings, are advanced in Section 8. Section 9 succinctly concludes this work.

Building on a Latin American case study which embodies the most serious (and yet not the only) instance of systematic corruption of commercial arbitrators, the overarching aim of the present work is to expose and “correct” the naivety of legislatively addressing the mentioned corruption as a context-independent phenomenon which accounts for neither the customary practices of the arbitral field, nor their embeddedness within certain well-engineered machineries of mutual quasi-necessary parasitic survival between hypertrophic private actors and the frangible state which is captured thereby.

2. Regionalising Criminology of White-Collar Crime in Latin America

While corruption, in different shades and to uneven degrees, embodies a typical human endeavour at all latitudes, no contention exists against evidence that for a number of economic and historical legacies, Latin America, in its overall regional complexity, is particularly prone to corrupt practices in the delivery of public services. The “public sector”, however, is not to be understood as a collection of realms of daily life which are distinct from market-regulated ones, but rather as a conjunctural manifestation of powers placed on a continuum with private forces, which not infrequently end up “capturing” the public discourse and its purportedly consensus-based policy priorities. Embeddedness is key for corruption to thrive; as remarked by Lambsdorff (2007, p. 210),

corrupt arrangements go along with high transaction costs. Corrupt relationships are unstable. They entail that sticking to one’s word may be dominated by betrayal and fraud. Corrupt agreements may even end up in mutual denunciation. Preexisting legal relationships can lower these transaction costs and serve as a basis for the enforcement of corrupt arrangements.2

To some extent, this insight had already been advanced by Granovetter (2007) through his account on the “social construction of corruption”, which has since been deployed to describe clientelist networks of societally embedded malversation across different regions. For instance, building on Granovetter’s theorisation, Szántó et al. (2012, p. 163) observed that

[a]s we have every reason to assume that the actors in corruption transactions are capable of gauging the risks and estimating the costs, we can therefore expect that they will attempt to minimize these using all means at their disposal, thereby increasing the net profit produced by the transactions. Among these means, it would seem advisable to consider the establishment, maintenance, and expansion of various types of […] interpersonal [… and …] institutional (business, contractual, political, etc.) networks.

In this region particularly, the so-called “regulatory capture” is also enabled by the absence of an impartial, super partes bureaucracy; in fact, Latin American governmental experts who advise on the technical feasibility of public projects and financial convenience of public tenders are not necessarily civil servants, being rather recruited through personalistic, unaccountable political appointments tapping into revolving doors (Murillo 2009, pp. 250–52). One should specify, however, that personalistic appointments are not translated into stronger informal accountability; even though appointments per se are customised for specific individuals, the latter, once chosen, will still operate within the bureaucratic machinery in the same depersonalised, detached, “anonymised”, responsibility-dodging manner of faceless bureaucrats—so that two negatives do not make a plus. Posts are personal and tailored, while relationships are cultivated with insiders but kept impersonal externally, and thus responsibility-relieved and ultimately unaccountable in a democratic sense (Ruggiero 2015, pp. 35–36). This is not exclusive to Latin America, of course, but while it represents somewhat the exception (ratione personae and/or ratione temporis) in more consolidated democracies, it seems to be accepted as “normal” in weaker ones and thus no longer questioned. As stereotyped as it might sound to some, this also helps explain why ‘[f]or many Latin Americans, corrupt […] bureaucrats are simply accepted facts of life’ (Warf 2019, p. 21); in this region, as ‘corruption is deeply entrenched, it may be tolerated simply as another part of doing business; private gain at public expense is not inevitably seen as an intolerable sin’ (Warf and Stewart 2016, p. 136).

Once the above is accepted, then it makes sense to discuss systemic corruption and different degrees of embeddedness of public and private actors alike within a colluded system that, to a greater or lesser extent, keeps working also through necessary corruption. The pervasiveness of the issue redefines the prima facie identification of elites with corrupted practices, in that what one would label as “white-collar crime” in other contexts is here extremely intertwined with and embedded within “street-level” forms of porous, low-end, unsophisticated corruption, to such an extent that the “distinguishness” of the former gets easily lost in pseudo-normalcy. This is why, in these contexts, speaking of institutional embeddedness seems most appropriate; corruption in institutions is not the exception, but the very symptom of a way of “doing the social”.

While comparative corruption statistics (Warf 2019, pp. 23–28) are often grounded in perceptions (i.e., surveys) rather than “hard data”, there are probably no figures that would highlight the intricacies of Latin America’s regional corruption more satisfactorily than a recent scandal named after its main legal-person protagonist: Odebrecht.

3. Causa Perit Iusta si Dextera non sit Onusta: Odebrecht and “Corruption as Embeddedness”

Oftentimes, when a natural person decides to change their name, the underlying will is to break up with their past and hide conduct they are no longer comfortable with. Similar reasons are subsumed under the move by a legal person, the family-owned Odebrecht, to rebrand itself as “Novonor” in the aftermath of the massive Odebrecht scandal. Indeed, even more dramatically than the corporate restructuring made necessary by the USD 3.5 billion fine settled with prosecutorial agencies in the US, Brazil, and Switzerland (Kim et al. 2017, p. 37; Morales and Morales 2019, p. 10; Low and Prelogar 2020, p. 215), the brand itself had become the epitome of “white-collar” (but see supra) crime regionally—and, one could claim, worldwide, as the protagonist of ‘the largest corruption case ever prosecuted under the US Foreign Corrupt Practices Act in its 40-year history’ (Campos et al. 2021, p. 186).

Odebrecht SA “was” a Brazil-headquartered conglomerate—one of the biggest worldwide—mainly operating in the fields of civil engineering, construction, and petrochemicals; in 2015 and 2016, it came under the spotlight for its involvement in two of the gravest regional corruption plots ever devised in Latin America: the one orchestrated by the Brazilian state-owned oil-industry multinational Petrobras,3 investigated under Operação Lava Jato/Operation Car Wash (Brun et al. 2021, p. 63), and its own, partly stemming therefrom (Campos et al. 2021, p. 172; Morales and Morales 2019, p. 2). As for the first, one crucial passage involved the incorporation of one hub of Odebrecht’s subsidiaries in the British Virgin Islands, Belize, and Uruguay, laundering money in another string of subsidiaries incorporated in Panama and Antigua and Barbuda due to shadow-banking operations through Swiss accounts, having ‘Petrobras employees as the final beneficiaries’ (Brun et al. 2021, p. 79). In what follows, however, I mainly focus on the second scandal, which sent its shockwaves across the region but also far beyond, involving politicians and businesspeople around the globe and raising fundamental questions of law and society.

Besides Brazil, investigations were initiated inter alia in Andorra, Argentina, Chile, Colombia, the Dominican Republic, Ecuador, Mexico, Panama, Peru, Portugal, Switzerland, and Venezuela, with bribes having been paid to officials in Angola, Guatemala, and Mozambique as well. In Colombia alone, the scandal led to ‘eleven convictions, 93 proceedings, 39 investigations, three indictments, eight charges, and 20 incarcerations’, Odebrecht’s debarment from the World Bank Group’s tenders,4 and domestic-law amendments so as to ‘criminalize bribes offered and received from the directors of the multinational company and public officials in charge of infrastructure works in the country’,5 and ‘[i]n Panama alone, Odebrecht was accused of paying more than US$59 million in bribes to government officials and intermediaries working on their behalf’ (Brun et al. 2021, p. 80). In December 2016, after being cornered by Brazilian authorities, Odebrecht admitted to having handed approximately USD 788 million in bribes to politicians, public managers, bureaucrats, and lawmakers across the aforementioned jurisdictions, in order to secure the tenders for a hundred projects. It ‘had created a unit called Division of Structured Operations to disguise improper payments. To conceal its activities, the unit used an off-grid communications system, called “Drousys”, which allowed members to communicate through secure e-mails and instant messages with codenames and passwords’ (Barclay 2019, p. 13). In order to pay the bribes, Odebrecht seldom made recourse to cash (Morales and Morales 2019, p. 7); instead, it chartered a sophisticated ‘web of offshore entities via tax havens with strong banking secrecy laws’ (Campos et al. 2021, p. 174), funnelling ‘hundreds of millions of dollars in secret payments to companies and banks in countries including the [US], China, the Netherlands, the United Arab Emirates, Panama and Antigua’ (Shiel and Chavkin 2019).

Alerted by their Brazilian counterparts, US and Swiss authorities launched their own investigations, which resulted in unprecedented degrees of prosecutorial cooperation, even formalised in the Declaración de Brasilia sobre Cooperación Jurídica Internacional contra la Corrupción (informally renamed “the Brasilia Agreement”) whereby ‘prosecutors agreed to create joint investigative teams and to strengthen international cooperation to coordinate the investigations in Brazil and the other countries where Odebrecht engaged in bribery’ (OECD 2017, p. 145). This is also to be framed against enhanced efforts on the part of the US administration towards strengthening anticorruption enforcement within the cooperation framework established through the OECD Convention on Combatting Bribery of Foreign Public Officials in International Business Transactions (Davis 2018, pp. 1266–67). Notably, at the same time, Brazil also granted a leniency agreement to another corporation, J&F Investimentos SA, accused of corrupting ‘a thousand politicians over the course of a decade’ (Mendelsohn 2017, p. vii), including former Brazil president Dilma Rousseff (Sobrosa Cordero et al. 2019, p. 24); this combination nominated 2017 and 2018 as the “two golden years” for anticorruption prosecutions in the country, ‘with a memorable level of public scandal’ (Taffarello 2019, p. 32).

Meanwhile, ‘a class-action lawsuit was filed in a New York court alleging that Graña y Montero[, Peru’s most prominent builder and frequent project-partner with Odebrecht, had] violated the Securities Exchange Act by making false or misleading statements and failing to disclose it was aware of Odebrecht’s bribes’ (Dube et al. 2017).6 This notwithstanding, investigations on the wider Odebrecht scandal in Peru were reportedly halted for months due to a number of suspects’ suicides (including that of former president Alan Gabriel Ludwig García Pérez, who shot himself on 17 April 2019) and other suspected deaths, poisonings, and serious injuries, as well as coordinated death threats to witnesses and whistleblowers; in Colombia, ‘one of three arbitrators working to resolve issues with a contract between Odebrecht and Colombia’s [Agencia Nacional de Infraestructura] died of a heart attack’ (Asmann 2019). When Odebrecht—after pleading guilty—attempted to sell (both domestically and abroad) its share participations in projects such as the “Southern Peruvian Gas Pipeline” (GSP) Consortium, emergency decrees by the Peruvian government impeded these transactions, leading to simple breaches of contracts on the part of Odebrecht rather than to the application of fines calculated on the transactions’ amounts, which would have actually benefitted Peru (Moncada Alcantara 2018; Barclay 2019, p. 15). Johnson (2020, p. 65) rightly observed that ‘barring [an entity] from participating in public works contracts, could lead to bankruptcy of that entity […]. If the public works contractor is a large firm, as was Odebrecht, its bankruptcy could lead to significant unemployment in the region’. Indeed, Odebrecht’s share-price dropped drastically, and the company rapidly went bankrupt, directly impacting 100,000 employees according to a conservative estimation (Taylor 2016). This counts as just part of the broader issues faced by both prosecutorial agencies and corporations in Latin American jurisdictions once they entered into collaboration agreements as a follow-up to corporate settlements and their executives’ plea deals. Taffarello (2019, pp. 33–34) recalls how in Brazil, for instance,

the multi-agency system to fight corruption which was established by the Brazilian Constitution had proved its importance to prevent powerful political or economic actors from capturing the agencies of control. However, in the wake of some very large-scale plea deals and settlements, wh[ose] extent and complexity had probably never been foreseen and whose statements of facts touched the powers and jurisdictions of multiple authorities, it proved to be a hard test for them to coordinate and cooperate with each other […]. In this regard, two important documents were released: the Agency’s guidelines for leniency agreements, published in September 2017; and the joint guidelines on plea agreements, […] published in May 2018. […] Irrespective of the challenges and perplexities still at stake, it is fair to assert that enforcement policies relying on plea deals and settlements are here to stay.7

In the aftermath of the prosecutorial findings recalled above, a dark reality surfaced: over two decades, numerous national elections in Latin America had been influenced or wholly orchestrated through Odebrecht’s bribes, especially in Peru (Rodríguez-Olivari 2020, pp. 166–67).8 Thousands of politicians across the region, including several heads of state, were thus removed from office, sentenced, and incarcerated. To exemplify this, Ecuador’s vice president Jorge David Glas Espinel was convicted, Peru’s former president Ollanta Moisés Humala Tasso spent nine months in pretrial detention, and another former president of Peru, Alejandro Celestino Toledo Manrique, was arrested in California and will be extradited in the near future (Madureira 2023). Moreover, Peru’s then-president Pedro Pablo Kuczynski was nearly impeached (and he then resigned), former Peruvian first lady Nadine Heredia Alarcón de Humala is facing jail (Chavkin 2020), and in Brazil, Odebrecht admitted to having corrupted the governors of Paraná, Goiás, Minas Gerais, and Rio de Janeiro, as well as former president Michel Temer, among others (Sobrosa Cordero et al. 2019, p. 24; Johnson 2020, p. 51; Chávez Bravo and Tovar Mena 2020). In Panama, former presidents Ricardo Alberto Martinelli Berrocal and Juan Carlos Varela Rodríguez have been indicted for corruption (Goodwin 2020), as have two of Martinelli’s sons, who are awaiting extradition from Guatemala (TeleSUR 2021). Martinelli has served his prison sentence in the US, and is now permanently forbidden entry therein (Medina 2023); he will soon face corruption charges in Panama, as well (al-Jazeera 2023).

However, the most notable case—also because of the geopolitical weight of the country concerned—is possibly that of Brazil’s former president Lula, who allegedly favoured both Petrobras and Odebrecht and was thus charged with passive corruption as well as money laundering. One should indeed write “allegedly” because the matter is exceedingly controversial, meaning that although President Lula’s conviction was confirmed up to Brazil’s Supreme Federal Court (Moyer 2019, p. 7), later leaked Telegram conversations between the judge and the lead prosecutor in his case (Sérgio Fernando Moro and Deltan Martinazzo Dallagnol, respectively) provide credible room for assuming that at least parts of those charges—just like the entire, hypocritically sudden anticorruption “awakening” of the country—were orchestrated by Lula’s political opponents in order for him to be stripped of his political rights, including that to stand as candidate in the federal elections (Greenwald and Pougy 2019; Fogel 2019; Londoño and Casado 2019; Fishman et al. 2019). Lula even petitioned the UN Human Rights Committee to restore his political rights, but the Committee’s (cautious yet favourable) decision was dismissed as “non-binding” international law by Brazil’s Superior Electoral Court (Gurmendi Dunkelberg 2018). His charges have since been cleared, and having won another round of elections, at the time of writing he holds office as Brazil’s president again.

These vicissitudes will not surprise anyone who is familiar with Latin American “judicial politics”, which alternatively destabilises left-wing and right-wing parties alike; as recounted by Dezalay and Garth (2002, p. 222) two decades ago already, prosecutorial activism across several jurisdictions within this region, but especially in Brazil, has long proven to represent an effective springboard from second-tier (legal career, including the traditional judiciary) to first-tier (nation-wide politics) elitism. In the case scrutinised here, even those who subscribe to the view that Lula’s wrongs still stood evident despite the aforementioned suspicious prosecutorial activism, acknowledge the ‘conflicts of interest between law enforcers’ duties to the investigation and their own private economic and political benefit’ (Rose-Ackerman and de Mattos Pimenta 2020, p. 199).9

4. Disapplying the Clean-Hands Doctrine Where Nobody’s Hands Are Actually Clean

Alongside its collaboration with the investigatory authorities, Odebrecht also tried to protect its investments in Latin America by means of “forum shopping” through its foreign subsidiaries, including from a tax haven (Luxembourg) by reliance on the Convenio sobre Promoción y Protección Recíproca de Inversiones entre el Gobierno de la República del Perú y la Unión Económica Belgo-Luxemburguesa for challenging the freezing of the aforementioned GSP project.10 In this and similar cases, while Latin American jurisdictions will centre their defence around the “clean-hands” doctrine,11 thus submitting that Odebrecht is not initiating these arbitration proceedings and/or would not be participating therein ‘with clean hands’ (CIAR Global 2020b), Odebrecht could advance the argument that host states are systemically corrupted to such an extent that bribery is somewhat “normalised”. Hence, as a minimum, Peru would be ordered to recover part of the arbitration’s costs (López 2020), but most saliently, terminating contracts obtained or performed through “encouraged” or “unavoided” corruption would amount to reacting asymmetrically—through estoppel (Bedoya Denegri 2020, p. 6; Davis 2019, p. 133)—and might be treated as indirect expropriation.12 Bulovsky (2019, p. 132) sides with this approach:

[a]llowing a host [S]tate to invoke its own malfeasance to avoid liability disproportionately punishes investors and unfairly advantages corrupt [S]tates. Corruption is not a unilateral act—a host [S]tate has either requested or accepted a bribe paid by an investor. […] The investor is merely one party to the corruption—“it takes two to tango”.13

Phrased differently: In these situations, from a systemic viewpoint, no one’s hands are truly clean—which is why I define this corruption as “embedded”, and such embeddedness exhibits legal implications. The Peruvian president dismissed Odebrecht’s USD 1.2 billion lawsuit as baseless, confidently claiming that the arbitration tribunal will be sensitive to fighting corruption, as this is now high on states’ agenda (Andina 2020), and yet, it is quite a curious coincidence that Peru experienced a sudden epiphany regarding corruption within its jurisdiction and decided to finally address it vigorously at exactly this time. Collusive corruption and “corruption by necessity”—so to write—are no justification for these acts, neither ethically nor legally; if the two who “tango” externalise their mutual capture onto societal third parties, including the few honest taxpayers, this is not to be accepted or incentivised by legal doctrines. Rather, in keeping with the dancing metaphor, it is just about how to “fairly” and “contextually” partition the choreography movements (in legal jargon, “apportion liabilities”) between those two parties who indeed “tango”—corporations and the state.

Regrettably, paying bribes, in certain Latin American jurisdictions, may also represent a last-resort compromise for business survival amid public extortion and systemic opacity (Wahrman 2020, p. 37); for instance, with regards to the quid pro quo surrounding the construction of the first line of a metro system in Lima, Odebrecht executives ‘stated that, had Odebrecht refused to pay the larger bribes, [Peru’s Vice Minister of Transport and Communications] would have allocated the project to a different bidder—which suggests that the bidders competed in bribes’ (Campos et al. 2021, pp. 182–83). Furthermore, Odebrecht corrupted thousands of officials for hundreds of projects, but each bribe was relatively small; it ‘won only a small [economic] advantage over other competitors by paying bribes[; indeed, it] increased its market share dramatically while its overall profits remained flat’ (ibid., p. 185). Bribes rarely needed to hit high sums, which was also due to Odebrecht’s perfect mimesis of the surrounding corruption-permeated politico-business environment, where it easily formed “corruption cartels” with supposed competitors (Morales and Morales 2019, p. 8), thus agreeing ex ante on the quantum of bribes to be proposed to state authorities in order to certainly win at least a fraction of those bribe-based competitions. One can also record countless instances of apparently immaculate “Western” companies whose Latin American branches inevitably fell into corruption, one of the most recent examples thereof being Siemens in Argentina and, again, Brazil (US DOJ 2018; Knobloch 2013; ACFCS 2014), although it did not spare other corruption-prone jurisdictions in other continents (Venard 2018). Equally true, however, is that multinationals—including Odebrecht—whose original geographical market is corruption-prone tend to “replicate their embeddedness” and financially exploit it by investing in foreign markets which are thought to be (actually or potentially) more, as opposed to less, corrupt (Wahrman 2020, pp. 38–44). This means that although Odebrecht-like multinationals subscribe to the mentioned embeddedness in their “jurisdiction of origin” also out of “necessity”, they later acquiesce to it without even trying to unembed themselves via market diversification; they pursue the reverse, if anything. And they use transnational corporate law and dispute-settlement mechanisms to perpetuate and replicate conditions of embeddedness from their original location’s corruption.

The present article, however, now sidelines investment-arbitration matters to focus instead on a related “sub-scandal” which falls entirely within the perimeter of the international commercial arbitration community.

5. Diligite Iustitiam qui Iudicatis Terram: Odebrecht’s Arbitration Scandal in Peru

Amusingly enough, in tempore non suspecto, it was the same legal counsel of Odebrecht, by means of a scholarly article, who cited the ‘risk of corruption’ between the factors Brazilian businesses should have considered before taking a dispute to arbitration rather than submitting it before domestic courts (Sampaio Valverde 2006, p. 515). Arbitration is indeed intimately connected to corruption in multiple ways, although most of the available literature analyses one side of the story only: the ability of arbitrators, duty bearing thereupon, as well as the feasibility and convenience of identifying and denouncing contracts obtained or performed through corruption by the parties to arbitral proceedings; recently, these questions showcased their relevance, e.g., in Colombia during arbitrations linked to the Odebrecht scandal (Arias and Lamo 2020). Nonetheless, far less frequently explored are the institutional frameworks and professional circumstances which may make arbitrators themselves prone to corruption when immersed in certain sociopolitical environments shaped by regulatory capture and endemic market distortions caused by politically driven, endorsed, or at least negligently overlooked corporate misconducts. Hence, Sampaio appeared prophetically right; the likeliness of facing corrupted arbitrators is one of the factors Latin American businesses are advised to consider before entering into arbitration agreements, and all throughout potential proceedings as well. As was remarked by Debuchy and Kamath (2022), the availability, efficiency, and transparent functioning of arbitral institutions, experts, and procedures is of the essence in all those contexts where construction-aimed public–private partnerships revolving around key state infrastructure are at stake, such as, indeed, the case across Latin American jurisdictions.

The Odebrecht saga contributes to suggesting that the mentioned risk of having disputes submitted to corrupted (or “borderline-corrupted”, so to write) arbitrators is not just theoretically sound—meaning that the political and economic conditions in certain societies might more easily expose arbitrators to corruption—but also practically verifiable and, at least in this case, verified. Before dismissing such a case as “the exception that proves the rule”, “one single instance”, or “an isolated incident” that occurred by chance, one should perhaps examine the environmental factors which contributed to its manifestation, and wonder whether they could find replication elsewhere in similar contexts; in other words, one should assess their socioeconomic and cultural embeddedness.

The case I am referring to is the “scandal within the scandal” of corrupted arbitration practices in Peru,14 through which Odebrecht overcharged the state by securing lucrative price increases for the procurement calls it had initially won upon lower expected cost. Funnily—but not especially so—Sampaio himself submitted a guilty plea once Odebrecht’s plots had been discovered (Schoenberg et al. 2016), which means that while he was drafting the aforementioned journal article, the corporation he was acting as counsel for was already active in corrupting countless individuals around the globe, including those commercial arbitrators he was writing about and warning against. ‘It would be a pity if those who promote investment in Peru stopped taking advantage of the strength and throughput of the private sector that is currently backing the country, delaying projects that need to be implemented immediately’, Jorge H. S. Barata, Odebrecht’s director-superintendent in Peru (Cole 2017), said (Oxford Business Group 2012) exactly around the time Odebrecht itself was delaying all projects by inflating its own profits through several rounds of arbitration. These arbitration-related events in Peru have been well-summarised by Ríos Pizarro (2019) and also briefly reported in a subsequent arbitration whereby a US investor, a potential competitor of Odebrecht in Peru, lamented15 the latter’s scarce success in fighting corruption and ensuring that the construction tenders were won on the basis of economic-engineering merit rather than political connections. Nonetheless, I will briefly retrace the story’s timeline and add a few elements which might help the reader contextualise both the facts and their possible long-term implications.

In February 2017, following Odebrecht executives’ confessions as outlined supra, the Group’s treasurer testified that a Peruvian arbitrator, Jorge Horacio Cánepa Torre, had received USD 3 million from Odebrecht via his bank account offshored with Banca Privada d’Andorra S.A., in order for Odebrecht to obtain surcharges to the original contracts price through seventeen arbitral proceedings against Peru. In September 2017, Peru’s prosecutor thus concluded that Cánepa had been corrupted by Odebrecht to pay governmental officials and other arbitrators on its behalf (El Comercio 2019; Campos et al. 2021, pp. 180–82), including paying public administrators from the Ministry of Transport and Communications not to appeal (technically “denounce”) the award before Peruvian domestic courts (Panorama 2019). While apparently a newcomer to corruption charges, Cánepa—a politician himself, for the Popular Christian Party—had already been involved in public scandals over the previous decades, the most serious of which had led him to flee Peru and seek asylum in the US. In the guilty plea inaugurating his new role as a “protected witness” in the Odebrecht saga, Cánepa disclosed the names of thirteen arbitrators which he would have allegedly bribed, providing only wire transfers to support his claim. Accusations were thus formalised against three more—renowned—arbitrators (Fernando Cantuarias Salaverry, Franz Kundmüller Caminiti, and Mario Castillo) on the circumstantial grounds of having determined their fees based on the complexity of the case, having set the Tribunals’ fees without complying with the fees chart of the Lima Chamber of Commerce (“LCC”), and having held a case-management conference with the parties to define the procedural rules and appoint the chairperson. Two months later, after the prosecutor had accused the arbitrators of specific passive bribery, aggravated collusion, aggravated illicit association, and money laundering, the relevant Criminal Court issued a pretrial detention order against all involved arbitrators, fearing they would have otherwise fled the country by pointing to professional conferences overseas and similar commitments abroad as a cover-up.

Still, in November 2019, the relevant Court of Appeals reversed the detention order for Cantuarias, Kundmuller, Castillo, and other five arbitrators, upon acknowledging that the first-instance court had misread their conduct and misconstrued the crimes based on mostly circumstantial evidence; however, it still imposed several degrees of restriction to the arbitrators’ freedom to travel (Franco et al. 2019). Put differently, the Court of Appeals found that the prosecutor and the Criminal Court had demonstrated an extremely poor understanding of standard arbitration practices that, taken in isolation, should have elicited no suspicion per se. The problem, once again, revolves around embeddedness. While being ordinary and accepted as normal, for the time being, within the arbitration community, such practices are not necessarily the smartest ways to dispel suspicion of corrupted or “borderline” behaviour, nor to prevent it. Within certain contexts where corruption is a collegial and widespread endeavour, commonly innocent practices might well raise suspicion if the prosecution or generalist judges do not share the arbitral professional milieu. This is indeed what occurred in Peru as well. To begin with, the Prosecutor and the Criminal Court had disapplied the difference between ad hoc and institutional arbitration; while fee charts are binding on the second, they are irrelevant for the first, so that any difference between the fees set by Cantuarias, Kundmuller, and Castillo on the one hand, and by the LCC chart on the other hand, could not, per se, justify judicial inferences of corruption. Along similar lines, adjusting arbitration fees in accordance with the increased economic momentum of a case is an unfortunate, and yet widespread and commonly accepted, practice in the arbitral profession. Not only is this accepted, but it is even sanctioned by law: Peru’s Arbitration Law explicitly mentions the amount of a dispute—together with its “complexity”, “effort needed”, etc.—as a factor for arbitrators to consider when setting their fees.16 Of course, this cannot mean that fees which keep increasing by each hearing—all hearings pertaining to the same company and its subsidiaries—are or should be considered normal; however, they cannot be taken at face value to charge an arbitration with corruption, either—bearing in mind inter alia that fees are to be known by the parties in advance of the proceedings rather than once the amount of an award has been fixed (which is also a good attempt to fight preventive bribery). Constantly or unchallengedly readjusted fees might reasonably be taken as corroborative evidence of bribery networks under certain circumstances, but would never be appraised as stand-alone proofs thereof.

The question, of course, is not only whether these arbitrators accepted bribes first-hand, but if they have or ought to have noticed (and reported to relevant authorities) suspect moves in these reiterated requests for tender-value recalculation, or in their colleagues’ behaviours and/or threads of appointments. Indeed, ‘Odebrecht había prevalecido en 35 arbitrajes y el Estado peruano en apenas siete. Pero lo más importante era que gracias a esas 35 victorias, Odebrecht ganó 254 millones 656 mil 753 dólares. El Estado, a su turno ganó, por sus siete “victorias”, cero dólares y cero centavos. Era más que razonable asumir corrupción’ (Gorriti Ellenbogen and Mella 2019), and Cánepa arguably could not have carried it all out alone; what is more, he was only directly involved in nineteen out of those thirty-five arbitrations. Beyond this, the prosecutorial strategy was also built on case-management conferences which are, again, standard practices—although I accept that they should more wisely be limited to appointment-related agendas of availability and qualification, and take place in institutional settings, with registered participants. As of today, there is no formal procedure to make sure these meetings are not excessively deformalized and eventually abused (also, potentially, for corruption purposes). As a side note, the “international standing” of those arbitrators made potential conferences abroad credible as professional commitments, and if anything, they should have counted as reassuring factors rather than being regarded as “escape routes”; in theory, an argument can be advanced that the more transcontinental an arbitrator’s practice and affiliation, the less probable their embeddedness in a well-engineered local system of corrupted capture. This is because elite arbitrators are transnationalising their adherence and informal “referentiability” to common standards of conduct and de facto oversight, in line with the upper circles of most supranationally operating legal professions (Moreira 2022a, pp. 66–68). That is not the case here: although the Odebrecht affair as a whole resonated globally, its sub-scandal in Peru can in fact be deemed relatively local.

The global arbitration community’s reactions to the pretrial order mentioned above further testify to the invisible watershed between those arbitrators who are considered “insiders” owing to their international reputations, and those who are identified as “outsiders” by reason of the local reach of their practice (Moreira 2022b). While the latter’s embeddedness in a structured but externally unknown relational system makes their choices more suspicious to external eyes, the former’s integrity goes not simply unquestioned, but even forcefully restated by the other members of the international circle. Perhaps unsurprisingly, when all arbitrators were incarcerated, the president of the International Chamber of Commerce dispatched two letters to Peru’s Ministry of Justice, “attesting” to the uncompromising morale of “its” arbitrators (i.e., the ones who held informal affiliations to the global arbitration hub) and pledging their release. The International Bar Association transmitted a letter, as well, and further concerns were voiced, e.g., by the Spanish Club of Arbitration as well as by a famed London-based arbitration scholar, Catherine A. Rogers (Simpson 2021, p. 11).17

6. Corruptissima Re Publica Plurimae Leges: Peru’s Rushed, Naive Anticorruption Legislation

Does Odebrecht’s Peru sub-scandal come as a surprise? Not particularly. Both the normalisation of corruption and the regulatory overreaction—though not infrequently a superficial one—thereto exhibit a long history rooted in Peru’s uncertain relationship with democratic emancipation since the time of colonial reforms in the mid-XVIII century. This is not the appropriate forum to discuss how Peru’s corruption links to and diverges from the wider Latin American issue with crony malversation; Quiroz (2008) is one of the authorities in articulating Peruvian peculiarities against their regional setting. What matters here is to emphasise that “embedding” corruption to then rush into disingenuous regulatory responses thereto is not new in Peru; Odebrecht’s Peruvian sub-saga merely offers one more illustration of this mechanism, but it provides a chance to advise on how to refrain from such severe (and yet in fact trivial and decontextualised) responses when it comes to alternative-dispute-settlement institutions—and domestic commercial arbitration more specifically.

In fairness, the partly unjust treatment of these arbitrators in Peru represents a dark page for the country’s justice system, but at the same time, one cannot disregard the impression that certain “normal”, “common”, or at least “accepted” behaviours practiced by arbitrators do not stand as the most functioning barriers against corruption, which would be all the more important within the Latin American—and, more specifically, Peruvian—context. What is plausibly innocent elsewhere should be preventatively avoided where large-scale corruption embeddedness manifests itself. Consequently, it is not just a matter of “educating the judiciary/prosecution” about how arbitral proceedings are usually organised and managed, but of reforming those “borderline”, old-fashioned practices in the arbitration field which are poorly suited to fighting bribery, especially in corruption-prone environments, and can as such be more frequently “misunderstood” despite their legalistic lawfulness. Otherwise put, one cannot confine themselves to argue that the domestic backlash against arbitrators in Peru could ensue from the very same populist overreaction to these arbitrators’ potential misconducts/crimes—as, e.g., Mizner (2020) did, by asserting that ‘[i]n an era of populism and democratic change, a system which can be perceived to challenge state authority needs to be alert and responsive to the threat to honest arbitrators’. In fact, “populism” is exacerbated by arbitral customs which run on the edge between legitimate and corrupted practices, showcasing an “edgy attitude” which does not help defuse the tension between populist/no-global movements’ bias against “business-friendly” dispute management and the undeniable importance of out-of-court alternative-dispute-resolution settlements. It is not a matter of ‘popular upheavals against our liberal economies and […] globalization’ (Mourre 2020, p. 1) to be approached condescendingly and simplistically, but rather an issue whose deeply rooted phenomena of systematic capture are to be taken more seriously as well as prevented more consistently (law-wise, as well).

To be sure, the right balance should be kept between tribunals’ autonomy on the one hand and governmental interference with those same tribunals through overregulation on the other; this is a delicate balancing exercise which Peru itself had to struggle with in the immediate aftermath of the Odebrecht scandal. I argue that Peru, legislating in a rush and misappreciating the intricated, sophisticated threads of its own embeddedness, failed to perform such an exercise.

In fact, when it comes to these complex choices, the state has to ensure that corruption is prevented, and simultaneously that corporations performing business in the country keep choosing its domestic arbitral institutions rather than drafting contractual arbitration choice-of-forum clauses that only include foreign fora of higher reputations for independence and freedom. As for Peru, mandatory arbitration for public-procurement disputes, introduced well before the Odebrecht scandal to overcome the procedural length and “inflexibility” of the traditional court circuit, has precipitated other problems related to the costs, transparency, and overall credibility of such a compulsory arbitration system. Furthermore, compulsory arbitration powers a vicious circle of mistrust and structural non-reform in other areas of state justice: ‘How can an individual trust the Judiciary if the State itself, backed by the Executive and Legislative branches, does not? How does the citizen [feel and behave] when he/she must seek justice through the Judiciary, knowing not only that the system is unbearably lethargic, but also recognizably corrupt?’ (Pestana Macedo 2020, p. 10).

Most recently, the country tried to address corruption comprehensively and resolutely, with mixed results; for instance, while according to Law No. 27,588 ‘directors of government-owned companies or government representatives in directories and advisers, officials or servers […] are prohibited from […] intervening as attorneys, advisers, sponsors, experts or arbitrators of individuals in processes that have pending with the same division of the government in which they provide their services, while they hold the position or fulfil the conferred assignment’ (Chávez Bravo and Tovar Mena 202018), this does not preclude preallotted revolving doors over time, nor does it speak to the embeddedness of all these actors within a context whose endemic corruption cannot be simply eradicated legislatively.

Multiple criticisms have been mounting since January 2020 after the Peruvian government bypassed Congress debates to issue its Emergency Decree on Arbitration;19 they posit that the executive went a few steps too far in regulating the domestic arbitration market (Jones 2020) and that although conceived for proceedings to which the state is a party (as it clearly appears from the preamble), the decree provides highly disputable “cures” applicable to state-participating international arbitrations as well (Forno 2020). Amid other measures, it sets a claimed sum beyond which ad hoc arbitration is ruled out, although ‘the question arises as to how should it be reflected in the arbitration clause, since the claim amount is not known at the time the clause is stipulated’ (Mereminskaya and Inostroza 2020), so that this uncertainty could contribute to pushing Peruvian law to the periphery of the laws contractually chosen by corporations operating in Peru for arbitrating their disputes. Furthermore, the Emergency Decree stipulates overbroad—and possibly arbitrary (which might lead to unintended effects)—incompatibilities for the appointment of arbitrators, and institutes a roster of arbitrators who are permitted to arbitrate disputes Peru is a party to, while leaving the issue of arbitration fees unaddressed; indeed, although high-amount disputes will be handled through institutional arbitration only, the LCC still permits several readjustments to arbitrators’ remuneration as cases grow in complexity, and there are still no rules to debar recurrent appointments,20 nor to ensure that over-informal case-management conferences are not exploited as bribery backdoors. The short-sightedness and incompleteness of these rules warrant a question: could it be that a conclusion formulated with regards to Latin American politicians removed from office for corruption—i.e., that ‘whether the formal rules for removal are followed is not dispositive that institutions are working as they should’ (Helmke 2020, p. 101)—is also of relevance for the arbitration community and the procedures which should govern arbitral appointments and proceedings? Finally, these rules are straightforwardly designed on the emotional wave of a corruption scandal, which makes it anything but certain that they will be converted into ordinary laws—and, in the positive, how.

The outspoken rationale underpinning this decree is that state authorities, representatives, and policymakers can be trusted (and thus act as “guarantors” for the healthy administration of arbitral proceedings) while individual arbitrators themselves might not; nonetheless, a more accurate narrative would defer to the evidence that the main controversy at stake lies with the mutual capture between state and non-state entities, rather than with contrasting a few “bad apples” coming from “the outside”. Also, arbitrators’ secretaries and even arbitral institutions as a whole are not automatically exempted from corrupted practices, either, as verified, e.g., in Russia in 1993 vis-à-vis the Technostroyexport v. International Development & Trade Services, Inc. case21 (Hernández Bernal 2019); nevertheless, it shall be noted that Russian arbitral institutions, just like Chinese or Cuban ones, are interfered with by the state more pervasively than their Western or even most Latin American counterparts would be. Furthermore, Odebrecht was charged in Switzerland for failing to exercise due diligence—i.e., to prevent bribery (Bühr and Henzelin 2017, pp. 161, 164), so that one might wonder what a similar provision would entail if enacted in Peru in terms of participating to commercial arbitral proceedings; should corporations take extra measures to ensure their representatives are not going to bribe arbitrators for the sake of being awarded higher profits? And could these rules be observed by corporations competing exactly on bribes (as remarked supra) within a deeply captured economy?

My last note on the facts one should be aware of with regards to the Peruvian arbitration scandal invests the line of commercial arbitrations challenging Peru for the entire series of procurement tenders assigned to Odebrecht’s competitors for reasons which have turned out to be (partly or wholly) unrelated to Odebrecht’s genuine economic and engineering competitiveness. Other cases arose from Peruvian authorities’ petitions to foreign courts to annul (technically “vacate”, i.e., set aside) arbitral awards rendered in favour to Odebrecht-affiliated corporations which had challenged those same authorities’ revocation of contracts obtained by paying bribes to Peruvian officials.22

7. Are Commercial Arbitrators Liable for Their Professional Misconducts?

The just-recounted “saga within the saga” probably represents the most intricated scandal involving commercial arbitrators in recent decades, extending from Peru to Colombia (El Heraldo 2017), whose government ‘started an investigation against three arbitrators because they resigned in a process also related to the Brazilian company’ (de la Jara 2017), and even to Libya. When Odebrecht sued the North African country before the International Chamber of Commerce (“ICC”) Arbitration Court, in two cases, for the suspension of construction projects and workers’ evacuation during the 2011 war,23 the tribunal dismissed Odebrecht’s claims by upholding Libya’s force majeure, but suspicion of previous arbitrators’ corruption was raised by Libyan authorities, as well—without the benefit of hindsight (Zaptia 2018, 2019).24

And yet, arbitration scandals revolving around Odebrecht are far from being the sole ones (either in Latin America, or in other regions); in Peru itself, they ‘followed a[nother] scandal in 2015 when it emerged that arbitrators in an ad hoc dispute between the regional government of Arequipa and medical company Oncoserv had received bribes to favour the company’ (Global Arbitration Review 2017). Similar misconducts, which were, however, not tantamount to corruption as it is strictly understood, had been ascertained in cases such as República de Colombia v. Cauca Company, Puma v. Estudio 2000, or Goller v. Liberty/Guangying v. Eurasia (Hernández Bernal 2019), but also most recently in Panama owing to suspicious documentary nondisclosure (CIAR Global 2020a). Thus, alongside corruption charges, the Odebrecht-related Peruvian scandal provides us with a chance to wonder, more broadly, whether arbitrators should be granted immunity from liability as far as their professional activities are concerned (on the model of administrative and ordinary judges across several jurisdictions), and in the positive, what might fall within the scope of said immunity. Could it encompass the civil aspect only?

On the criminal side, in 2017–2020 alone, Kenyan, Spanish, and Qatari courts—among others—sentenced arbitrators to prison terms for fraud, negligence, breach of trust, et similia, begging the question of whether this trend is set to crystallise in the near future, and what the consequences of such normalisation would be. Arbitration institutions’ rules do exist, but the exceptions thereof leave significant room for discretion to domestic judges as well as legislators (Lee 2021). In Italy, for instance, the question of whether an arbitrator can be prosecuted criminally for corruption is still unsettled, meaning that courts have answered to it unevenly, depending on the circumstances (Benedettelli 2020, pp. §4.141–§4.143). Other European jurisdictions, such as Austria and Switzerland, have opted otherwise, with the introduction of specific provisions targeting corrupt arbitrators (Santoro and Babiy 2013, pp. §5.02–§5.03).

In the coming section, I draw a number of generalisable policy suggestions from the lessons one may want to learn from the Odebrecht saga, as well as from all the theoretical insights proposed throughout the previous sections. The role of arbitrators is inspected more closely, with reference to the legal significance of their embeddedness in highly corrupted environments, and the consequences thereof for the performance of their duties, their professional liability, and their societal function as “alternative (?) dispute settlers” more broadly. What collocation are arbitrations assigned within Latin America’s “legal elitism”? Can arbitrators be trusted? Should they be? And, in the negative, what is the “embedded arbitrator” there for?

8. The Limits and Potential of the “Embedded Arbitrator”

The professional ethics of arbitrators remains a terrain for contestation, in particular when it comes to international arbitrators exploiting diverging (or unequally demanding) ethical expectations and requirements under the domestic arbitral rules of different jurisdictions (Benson 2009, p. 83), thus operating arbitral schemes motivated by what I would call “ethical shopping”. An entire stream of literature has flourished lately (see, e.g., most recently, Low 2021) to peruse arbitrators’ ability (as opposed to mere willingness) to fight corruption, also with reference to international conventions (Yan 2021). These scholars’ main concern lies with whether, as a matter of public policy, arbitrators have the capability for and should be entrusted with implementing contrasting actions against alleged corruption they might come to be aware of in the course of the arbitral proceedings (Haugeneder 2021). I reckon they would have the capability to do so, but this was not the focus of the present work; the second part of those scholars’ concern, instead, is tied to the legal–criminological question posed by this article: can the embedded arbitrator be “dis-embedded” from corruption-prone jurisdictions permeated by and ruled through regulatory capture? Two fundamental methodological problems arise here.

First stands the hurdle of defining what a “corruption-prone” jurisdiction exactly is; of course, the distinction between fully corrupted and corruption-free jurisdictions—where these two poles are just fictional, as it is always a matter of shades and degree—is merely an abstract one that can be intuitively grasped but not theorised (all the less so given that beyond surveys based on subjective feelings, we have no statistically significant comparative data on corruption). Hence, such a distinction is not theorised here, either; the policy suggestions formulated hereinafter are simply grounded in an if-premise (if a jurisdiction’s institutions are likely more prone to systemic corruption than not, …). It seems legitimate to write, as was done supra, that there are strong indicators to give one some confidence to hypothesise that, e.g., Latin American institutions face a systemic-corruption problem; this cannot mean, however, that the jurisdictions where such institutions are located can be sharply categorised as “corrupted” altogether.

As for the second methodological problem, it pertains to regulatory capture. Indeed, one may argue (e.g., Vecellio Segate 2022, pp. 393–405) that if a jurisdiction is truly regulatorily captured, its laws will be as well, as the outcome of a captured legislative process itself, so that regulating any assumed-as-captured process under such a regime would not remove the consequences of regulatory capture in the first place, including those regarding the conduct of supposedly captured arbitrators. This objection does feature some merit; however, one should concede that while identifying capture in downstream actions (such as, indeed, over the course of arbitral proceedings), ex post, is both harder and riskier, rectifying captured laws might be to an extent more feasible, at least to external observers. This also means that “community control” might prove tighter and both legislators and arbitrators might feel restrained from drafting overcaptured laws or exploiting the latter’s loopholes. In this sense, regulating captured procedures ex ante is not a panacea, but would probably prove more desirable than its alternative, which is not regulating them, thus leaving capture to go unchecked and reign unhindered. Furthermore, it is easier to audit legislative processes than arbitral proceedings, which are subject to confidentiality rules as well as expectations, and are often preferred by the parties exactly due to the degree of privacy offered thereby.

Resultantly, I assume one can refer to Peru-like contexts without even trying to scientifically define what could fall within such likeness’ scope, and I also build the following considerations on the premise that “uncapturing” arbitrators through regulation in those jurisdictions is to an extent possible and meaningful. While it would not fully prevent arbitrators from falling captured by the system, it would improve the situation, especially by unmasking systemic capture ex ante and debunking arbitrators’ own capture before their reference professional community, eventually increasing the risk inherent to arbitrators’ moral hazard. In other words, the regulatory action suggested hereinafter would seek to disrupt the ‘referential moral obligations’ (Katzarova 2015, p. 60) that seemingly keep arbitrators from realising the gravity of their embeddedness against theoretical professional standards of dis-embedded—and thus corruption-free—morality. Similarly to what other authors (e.g., Pertiwi 2020) have proposed in other contexts to help explain the ineffectiveness of rationalist, “positivistic” prosecution of embedded corruption, it can prompt regulatory action which accounts for networked, structured, captured, and professionally entrenched malversation as opposed to corruption which can be eradicated by law as a mere series of instances—however serious.

Having clarified the above, how should arbitration be regulated in Peru-like contexts? I do not wish to attempt one single prescription to fit all scenarios; rather, I briefly illustrate three potential solutions, which might be applied context-dependently with some margin of appreciation on the part of policymakers, depending on the regional sociocultural features, on the structure of the relevant markets, on the profile of the relevant arbitration communities, and on the perceived severity of the problem such jurisdictions self-assess to have with corruption (both generally and—actually or potentially—in arbitration).

The most radical option is that of forbidding the submission of public-procurement contracts before the scrutiny of arbitration, preferring administrative courts instead (despite expected delays in the administration of justice). The rationale is that while corruption in corporation-to-corporation arbitrations is problematic but confined to the private sector (unless such corporations are wholly state-owned), thus raising public issues only indirectly, state-to-corporation arbitrations certainly involve taxpayers’ money and cannot be entrusted to arbitration if the latter is prone to corruption. Self-evidently, if the relevant jurisdiction is deemed captured, its administrative courts will be just as corruption-prone; however, they are on average subjected to higher checks and balances compared with arbitrators, including enhanced whistleblowing, transparency, publication requirements, institutionalised and untailored-to-the-case appointments, and compulsory internal and external auditing procedures. In this case, efficiency concerns would be radically sidelined to favour decreased risks of capture, corruption, and, thus, mishandling of taxpayers’ money. Because this solution is the most extreme, warranting almost an overhaul of the way local commercial arbitration works, it should be adopted to address only the gravest situations (and yet, possibly before they concretise). It should be considered, however, that because regulatory and political instability could prompt the parties to resort to foreign laws within their choice-of-law contractual arrangements (Payet and Patrón 2020, p. 82), administrative courts would be required to apply such foreign laws despite being far less familiar with those laws compared with arbitrators.

An intermediate, slightly less demanding option is to submit public-procurement contracts to institutional arbitration only, thus eliding ad hoc arbitrations from the equation and entrusting arbitral institutions with oversight over the arbitrators’ conduct. Apart from its condescendence, this solution appears problematic market-wise, in that the relatively free market of arbitration would be intruded into and distorted through institutionalisation. Indeed, while the appointment of ad hoc arbitrators can be explained satisfactorily through market-allocation models, institutional arbitrators are not infrequently appointed in compliance with institutionalised logics of registration, rotation, and anyhow “listing”, which lower the degree of competition within the arbitration market; this might eventually prove counterproductive, and offset certain anticorruption benefits stemming from this choice. As outlined previously, one aspect that the Emergency Decree in Peru failed to address is the recurrent appointment of arbitrators across multiple proceedings involving the same corporation (including its subsidiaries); besides practical inconveniences related to the identification of all such networks of subsidiaries and thus to the belonging of several companies to the same corporate group internationally, one side effect of hyper-regulation would possibly be the creation of arbitration cartels so as to circumvent institutionalised appointments. This is probably the reason that, despite multiple related reappointments being discouraged by several—exceedingly soft and vague—arbitration guidelines (Giraldo-Carrillo 2011), not rarely disregarded and primarily addressed to international investment arbitrators (Newcombe 2011), no binding provision has ever been formalised in this respect. Nevertheless, in certain jurisdictions, the (certain) benefits of hard laws on repeated appointments might outweigh their (potential) undesired effects. Again, this variable is to be assessed in a case-by-case fashion.

The third, least radical compromise between efficiency strategies and anticorruption policing is that of introducing a number of minor yet binding procedures to further arbitrators’ professionalism and reduce a few circumstantial chances to pay or receive bribes. For instance, holding case-management conferences in “controlled environments” might slightly reduce the chance to have unregistered lobbyists, governmental consultants, and tax advisors informally participating. Even if remote case-management conferences normalise in the aftermath of the current pandemic and its promotion of “smart working”, ensuring that online organisational meetings can, by law, only be participated in by invited and listed professionals would remain of the essence.

Transparency-wise, whichever the option that best adapts to a given scenario, the publication of awards touching on public interests seems advisable as well—though this would require extensive articulation of what interests are to be deemed “public”. Peru’s arbitration rules that are currently in force25 do provide for arbitral awards ‘where the State is a party’ to be published for at least one year,26 but this appears to capture just a minor fraction of the entire spectrum of awards where substantial public interests are at stake in Peru.

It seems relevant to remark that civil liability for arbitrators would not readily replace any of these proposals, mainly due to three reasons. First, practically and conceptually, civil liability centres on the arbitrator as a “detached” and somewhat “abstract” individual agent rather than as an embedded representative of (and “mediator” between) the two poles of a mutually captured system. Second, the balance between arbitrators’ subjection to carefully calibrated degrees of liability for their faults, on the one side, and those arbitrators’ professional freedom and independence, on the other side, would be difficult to codify and enforce. Third, civil liability does not sufficiently redress taxpayers for the wasteful misallocation of their money when corruption for public-procurement contracts is at stake, so this liability regime falls outside the scope of the problem this paper has focused on. Indeed, regarding public-procurement and repricing-aimed bribes, criminal liability, when applicable, already punishes arbitrators for the damage they cause to society by accepting bribes; if we understand the public damage, i.e., the financial harm to all taxpayers, as ‘the difference between the [corrupt] contract price and the fair market price’ (Davis 2019, p. 61), even ignoring investigatory and enforcement costs out of simplification, civil liability would not significantly redress taxpayers monetarily. Moreover, as the present work has strived to argue from a socio-criminological standpoint, it would be theoretically meaningless to redress either the state or the corporation for bribes accepted by arbitrators, given that public officials are most often involved in these types of transactions, either proactively or out of negligence towards preventing them.

9. Concluding Notes

This paper endeavours to be one of the first-ever works to conceptualise corruption in arbitration by focusing not on bribes arbitrators should have denounced (e.g., MacKinnon et al. 2022), or on the unenforceability of arbitral awards rendered by corrupt arbitral tribunals, but on bribes arbitrators themselves should not have accepted or encouraged in the first place. Its claimed merit is that of conceptualising the behaviour of the “embedded arbitrators”, as well as outlining plausible regulatory responses thereto. In other words, its novelty resides in addressing the individual arbitrator as a professional who, just like any other, might become embedded within large-scale and state-participating corruption systems, thus proving not only unable to identify—let alone denounce—corrupt practices it might witness or suspect, but also a proactive agent of corruption itself.

In truth, globally speaking, corruption has not necessarily emerged as a main concern in the field of arbitration generally, and particularly in international and/or institutional arbitral proceedings, but it may become so in ad hoc domestic arbitrations involving public contracts (i.e., arbitrating on taxpayers’ money allocation) and engaging local (subsidiaries of multinational) corporations, performed in certain jurisdictions where corruption is (or is perceived to be) an endemic social plague. From a criminological perspective, politically active arbitrators are expectedly more likely to take bribes where repeated appointments vis-à-vis the same (group of) corporations are allowed (or even incentivised) under the misleading flag of “arbitral consistency”.

The multibillion-USD and cross-jurisdictional Odebrecht scandal, with particular reference to its “arbitration sub-scandal” in Peru, has definitely demonstrated that arbitrators, ordinarily assumed as trustworthy in literature, are not to be deemed completely immune to corruption in regulatorily captured jurisdictions where bribery-aided ties between state institutions and corporations are normalised or at least widespread.

In order to potentially contain the problem and prevent corruption phenomena from occurring on such a wide scale, corruption-prone jurisdictions might wish to consider a portfolio of regulatory options to improve professionalism, transparency, accountability, and, somehow, “diversity” on the part of arbitrators. The three solutions sketched here represent some of those options, but no panacea; in particular, it seems wise to consider the counterargument that arbitration as a free market may embody a few deviant behaviours, but it is best self-immunised against corruption, due indeed to long-term reputational competition. This counterargument is meaningful and should never be discarded. For this reason as well, my rudimentary proposals warrant further elaboration in scholarship, possibly calling for regionalised “variations on a theme” and a broad range of insiders’ views, from the extended legal community and beyond.