As several prior posts on this blog have discussed (see here, here, and here), corruption has emerged as a significant and controversial issue in international investor-state arbitration proceedings, with a number high-profile cases in which the tribunal refuses to provide relief on the grounds that the underlying contract was procured through corruption. In these cases, corruption allegations usually surface at the initiative of one of the parties. For example, this summer, Djibouti filed an arbitration against Dubai-owned port operator DP World, seeking annulment of a port concession because DP World allegedly formed its contract with Djibouti for the operation of Africa’s largest container terminal through corrupt means. However, in rare instances, corruption can enter the picture without either party raising the issue during the proceedings. In these cases, the arbitral tribunal considers the issue of corruption sua sponte, even when neither party alleges corruption by the other.

Perhaps the most prominent example of this is the tribunal’s decision in Metal-Tech v. Uzbekistan. In Metal-Tech, the ICSID tribunal, in its words, “required explanations” from the parties for suspicious facts that “emerged in the course of the arbitration”–in particular the fact that Metal-Tech had paid exorbitant, seemingly unjustifiable sums for consulting services to an Uzbeki government official and individuals with close ties to Uzbeki leadership. The ICSID tribunal then essentially placed the burden of disproving corruption in light of this circumstantial evidence on Metal-Tech, which could not come up with enough evidence to overcome the tribunal’s presumption. The ICSID tribunal held it did not have jurisdiction and dismissed Metal-Tech’s claim.

On the surface, sua sponte efforts by tribunals to address corruption may seem like a positive step in the anticorruption fight. Indeed, it might seem irresponsible for the tribunal to stick its head in the sand given such facially suspicious facts. As Michael Hwang and Kevin Lim assert in a recent paper endorsing this sua sponte practice, “Tribunals must remain vigilant and alert to the possibility of corrupt dealings being hidden by one or both parties, otherwise they may become unwitting accessories to heinous acts.” But in fact, the approach adopted by the tribunal in Metal-Tech, might do more harm than good. Indeed, by engaging in sua sponte considerations of corruption, arbitral tribunals might unwittingly perpetuate corruption under several different scenarios:

• First, tribunals might allow corrupt states to reap benefits from wrongdoing by dismissing claims against states where investors paid bribes to corrupt state officials. This irony was raised by Sam in his post on World Duty Free v. Kenya and holds even truer here. Uzbekistan was essentially rewarded with immunity for the corrupt acts of some Uzbeki officials, without even having to shoulder the burden of raising hypocritical arguments and bringing its own wrongdoing to light. Although the tribunal did require each party to bear its own costs because of Uzbekistan’s contribution to the wrongdoing, this seems like a small price to pay for immunity from Metal-Tech’s claims.
• Second, and relatedly, tribunals’ independent considerations of corruption might reward countries whose corruption and bribery laws are vague or arbitrary either on paper or in practice. Russia comes to mind as an example I’ve discussed on the blog before. An investor might enter into a transaction that, at the time, the investor viewed in good faith as legitimate, in the face of a murky divide between legal and illegal action. Later, a tribunal could take a differing view of the transaction and dismiss a claim by the investor. Assuming that open-ended legal standards in this arena are undesirable, such an outcome is neither fair to the investor nor promotes the development of bright line rules in the host country.
• Third, sua sponte consideration of corruption might displace the development of more sustainable and robust doctrinal mechanisms for addressing corruption in international arbitration. The UNCITRAL Model Law and the New York Convention allow domestic courts to set aside an arbitral award if the tribunal considered a dispute outside of the scope of the “submission to arbitration.” Thus, if a tribunal’s sua sponte consideration of corruption results in the issuance of an award against a corrupt state, that state’s courts may set the award aside on the grounds that corruption fell outside of the scope of issues submitted to the tribunal for consideration. Although Hwang and Lim note that domestic courts rarely set aside awards on these grounds, this practice could become more common if tribunals begin to punish states rather than investors once fishy facts are uncovered in the course of the arbitration. Rather than applaud tribunals’ ingenuity in the absence of corruption-related doctrines in international law, the focus should shift to developing corruption doctrines that can be used as tools by the parties themselves and incorporated into tribunal awards without fear that these awards could be set aside later.
• Fourth, tribunals might unfairly shift the burden onto parties to disprove corruption, and in the process sacrifice useful fact-finding that exposes and memorializes corruption, a useful function of arbitration awards that I discussed in a previous post. Corruption and bribery often go undocumented. By forcing parties to come up with affirmative evidence disproving corruption when parties were not prepared to do so at the outset of the arbitration, and dismissing claims when parties inevitably come up short, tribunals do a disservice to the parties and to the fact-finding process. Facts about corruption will not be exposed in the most comprehensive, useful way possible.

In sum, we should hesitate to praise tribunals for initiating corruption inquiries, even when it’s clear that corruption is percolating under the surface of the relevant transaction or investment. Even if their heart is in the right place, tribunals may inadvertently entrench corruption through such sua sponte considerations.