Today’s guest post is from Roxie Larin, a lawyer who previously served as Senior Legal Counsel for HSBC Holdings and is now an independent researcher and consultant on corruption, compliance, and white collar crime issues.
The U.S. Foreign Corrupt Practices Act (FCPA) is a powerful tool that the U.S. government has wielded to combat overseas bribery—not just bribery committed by U.S. citizens or firms, but also bribery committed by foreign nationals outside of U.S. territory. (The FCPA also applies to any individual, including a non-U.S. person or firm, who participates in an FCPA violation while in the United States, but this territorial jurisdiction is standard and noncontroversial.) The FCPA, unlike many other U.S. statutes, does not require a nexus of the alleged crime to the United States so long as certain other criteria are satisfied. For one thing, the statute applies to companies, including foreign companies, that issue securities in the U.S. In addition, the FCPA covers non-U.S. individuals or companies that act as an employee, officer, director, or agent of an entity that is itself covered by the FCPA (either a U.S. domestic concern or a foreign issuer of U.S. securities), even if all of the relevant conduct takes place outside U.S. territory.
In pursuing FCPA cases against non-U.S. entities for FCPA violations committed wholly outside U.S. territory, the agencies that enforce the FCPA—the Department of Justice (DOJ) and Securities and Exchange Commission (SEC)—have pushed the boundaries of this latter jurisdictional provision. They have done so in part by stretching to its limits (and perhaps beyond) what it means to act as an “agent” of a U.S. firm or issuer. (The FCPA provisions covering foreign “officers” and “employees” of issuers and domestic concerns are more straightforward, but also more rarely invoked. It’s rare for the government to have evidence implicating a corporate officer, and the employee designation doesn’t help unless the government is either able to dispense with notions of corporate separateness, given that foreign nationals are typically employed by a company organized under the laws of their local jurisdiction.) Until recently, the government’s expansive agency-based theories of extraterritorial jurisdiction had neither been tested nor fully articulated beyond a few generic paragraphs in the government’s FCPA Resource Guide. In many cases, foreign companies affiliated with an issuer or domestic concern have settled with the U.S. government before trial, presumably conceding jurisdiction on the theory that the foreign company acted as an agent of the issuer or domestic concern. (This concession may be in part because a guilty plea by a foreign affiliate is often a condition for leniency towards the U.S. company.) Hence, the government has not had to prove its jurisdiction over these foreign defendants.
But there was bound to be a reckoning over the U.S. government’s untested theories of extraterritorial FCPA jurisdiction, and the SEC and DOJ’s expansive theories are increasingly being tested in court cases brought against individuals who, sensibly, are more prone to litigating their freedom than companies are their capital. And it turns out that the U.S. government’s expansive conception of “agency” may be difficult to sustain in cases where the foreign national defendant—the supposed “agent” of the U.S. firm or issuer—is a low- or mid-level employee of a foreign affiliate, and even more difficult to sustain so where the domestic concern is only an affiliate and not the parent company.
Under traditional principles of agency law, to prove an agency relationship the government would have to show, first, that the U.S. company and the foreign national agreed to an agency relationship, and, second, that the U.S. company exercised “interim control” over the foreign national’s performance and duties. The concept of “interim control” means, at the very least, that the U.S. company initially stated what the foreign national should do, and had the right to give instructions or directions once the relationship was established.
The key case on this issue, which is still working its way through the U.S. court system, is that of Lawrence Hoskins. Mr. Hoskins, a British citizen and employee of French company, Alstom SA, allegedly authorized improper payments to two consultants retained by Alstom’s U.S. subsidiary for the purpose of paying bribes to Indonesian officials. The DOJ charged him with criminal violations of the FCPA’s anti-bribery provisions. Although the government had charged Mr. Hoskins with conspiring to violate the FCPA, as well as aiding and abetting FCPA violations by Alstom, an appeals court rejected these theories of liability in cases where the foreign national could not be charged directly with a substantive FCPA violation. But the appeals court left open the possibility that Mr. Hoskins might still be liable for acting as an “agent” of Alstom U.S. (which, as a domestic concern, would be subject to the FCPA).
On this point, the trial court instructed the jury that it could find that an agency relationship existed if it found “(1) a manifestation by [Alstom U.S.] that [Mr. Hoskins] will act for [Alstom U.S.]; (2) acceptance by [Mr. Hoskins] of the undertaking; and (3) an understanding between the parties that [Alstom U.S.] will be in control of the undertaking.” The jury convicted, but in a surprise ruling, the trial judge took the remarkable step of overturning the conviction on the FCPA charges. According to the judge, although Alstom U.S. controlled the project and gave Mr. Hoskins instructions, there was no evidence that Alstom U.S. controlled how Mr. Hoskins performed his tasks. As a result, the court ruled that there was insufficient evidence to support a finding that Alstom U.S. had an “interim right of control” over Mr. Hoskins, one of the requisite elements of an agency relationship. The government has appealed, and the implications of Hoskins have yet to play out.
Hoskins is the first case in which the government’s expansive “agency” theory for FCPA liability is being tested in court, but others are in the pipeline. This past April, the SEC filed civil FCPA charges against Asante Berko, a former employee of Goldman Sachs’ U.K. subsidiary. Mr. Berko, a dual Ghanian and U.S. citizen, is accused of facilitating the payment of bribes by a Turkish bank client to Ghanian officials—conduct that took place entirely outside U.S. territory. The theory of FCPA jurisdiction over Mr. Berko—as in Hopkins —is that Mr. Berko was acting as an “agent and/or employee” of Goldman Sachs U.S (a U.S. firm clearly subject to the FCPA). It’s unclear why the SEC didn’t just charge Mr. Berko directly as a “domestic concern,” since, as noted above, he has U.S. citizenship. Perhaps the SEC may have calculated that it would be more expedient to involve Goldman U.S. (which was not charged or otherwise named in the complaint) in order to put pressure on Goldman to cooperate with the SEC’s investigation. In any event, given the way the SEC chose to frame its complaint, the case will turn on the existence (or not) of an agency relationship, either between Goldman U.K and Goldman U.S., or between Mr. Berko and Goldman U.S.
On its face, the government’s claim looks tenuous. On the allegation that Goldman U.K. (Mr. Berko’s employer) was an agent of Goldman U.S., the SEC has not yet plead facts sufficient to overcome the bedrock principle that a parent company is not legally responsible for the acts of its subsidiary. As for the SEC’s alternative theory that Mr. Berko was himself an agent of Goldman U.S., the SEC has not offered much evidence that Goldman U.S. had an interim right of control over Mr. Berko’s conduct in relation to the project at issue. The SEC’s view appears to be that a memorandum completed in part by Mr. Berko about the project and provided to Goldman U.S. is sufficient to make Mr. Berko an agent of Goldman U.S. Yet such memos are a typical part of the compliance process for getting approval from a bank risk committee at the outset of a proposed project. The SEC has not alleged facts sufficient to establish an understanding between Mr. Berko and Goldman U.S. that the latter would be in control of Mr. Berko’s actions on the project, bypassing the Mr. Berko’s usual reporting line to his employer, Goldman U.K. This is essentially the same problem that the trial court identified in Hoskins.
Of course, we still don’t know how these cases will come out. It’s possible that the government will ultimately prevail on its expansive agency theories. But it’s more likely that the government will lose. What will the consequences of such losses be? In particular, what will the ramifications be for FCPA cases brought against foreign corporations—especially the foreign subsidiaries of U.S. firms or issuers—from whom U.S. authorities have routinely extracted settlements on the basis of unproven theories of extraterritorial jurisdiction? It’s not clear how many of these firms will have the stomach to contest FCPA charges in court, even if judicial rulings involving individual defendants substantially undermine the DOJ and SEC’s preferred “agency” theory for holding foreign affiliates liable for FCPA violations committed outside U.S. territory. But it’s possible that at least some of these defendants may be emboldened to challenge the government’s basis for statutory jurisdiction in court, or at least to insist at the negotiating table that such jurisdiction is improper.
Moreover, the courts’ greater scrutiny of the agency theory may have perverse effects that undermine good corporate governance. The problem runs as follows: Under the standard agency law principles of the sort that the Hoskins trial judge’s ruling emphasized, the less involved the U.S. issuer or company is in directing or overseeing the foreign affiliate engaged in the alleged bribery scheme, the harder it will be for the government to establish jurisdiction over the foreign affiliate. This may give multinational companies based in (or issuing securities in) the United States an incentive to avoid anything that would give the appearance of “interim control” over the foreign affiliate’s dealings with government officials (for example, bidding for public contracts), including decentralization of governance structures and strict adherence to corporate formalities and separateness between parents and subsidiaries. After all, attempts by the U.S. parent to exercise greater control over foreign subsidiaries—for example, by closely monitoring dealings with foreign governments and insisting on certain compliance practices—may be more likely to be viewed by courts (and by the DOJ and SEC) as establishing an agency relationship, which could render both the U.S. parent and the foreign subsidiary subject to FCPA liability if someone at the foreign subsidiary—someone with no other U.S. connection—pays a bribe to a foreign official. But a truly effective anti-bribery compliance program needs uniformity in application across an organization and the involvement of the parent in setting a good corporate culture from the top-down—exactly the opposite of what the focus on an agency relationship would encourage.
This is not to say that courts are wrong to scrutinize more rigorously the government’s theories of extraterritorial jurisdiction, or that it is wrong for the U.S. enforcement agencies to bring hard cases that push the limits of statutory jurisdiction. The problem lies with the FCPA itself, or more accurately with the unreasonable expectation that the FCPA can function as a global anticorruption regime, one that empowers U.S. enforcers to go after cross-border bribery wherever it takes place. The FCPA is a crucial tool, but it has multiple weaknesses and limitations. These shortcomings have been tolerated because the FCPA has been effective at combating global corruption. But if the government starts to lose FCPA cases against foreign companies and nationals because of the agency issues described above—or, even worse, if foreign actors seek to exploit these jurisdictional gaps by implementing less centralized and less effective compliance controls—then perhaps it’s time to start discussing other approaches.