It is widely agreed that foreign bribery is capable of causing harm to a range of different victims, including the governments whose officials are bribed (the so-called “demand-side countries”), and the citizens of those countries. Yet traditionally, when supply-side countries (those with jurisdiction over the firms that paid bribes abroad) reach settlement agreements with corporate defendants in these cases, the fines and penalties collected—which can sometimes run into the tens or even hundreds of millions of dollars—go to the supply-side government treasuries, a fact that has attracted considerable discussion and criticism.

In recent years, we’ve started to see some changes in the approach taken by supply-side governments on this issue, with the United Kingdom being particularly active. On several notable occasions, the UK’s Serious Fraud Office (SFO) has included in its settlement agreements with corporate defendants specific provisions to remediate the victims of foreign bribery. Importantly, such remediation (not just in the UK case, but more generally) can take two forms, which are often unhelpfully conflated:

• In some cases, the resolution of a bribery case may include compensation to identifiable victims, if it can be shown that the victims suffered a direct loss, the value of which can be reasonably estimated. The victim might be a foreign government itself. For example, the 2015 deferred prosecution agreement negotiated between the SFO and Standard Bank included a payment to the Tanzanian Government, because in that case an agent of Standard Bank had used money to which the Tanzanian government was entitled in order to pay an illegal bribe. The payment to the Tanzanian government in the settlement agreement was compensation for this loss.
• In many cases, though, the harm done by foreign corruption is more diffuse, the victims are difficult to identify individually, and the monetary value of the harm inflicted is impossible to calculate. Nonetheless, even though traditional victim compensation is not possible in these cases, it is still possible, and often desirable, for a portion of the fines and penalties collected from the responsible corporation to be directed toward improving the lives and livelihoods of the population victimized by the misconduct—perhaps by making a payment to the government of the demand-side country, possibly earmarked for a specific purpose, or perhaps by donating money to charities, or by purchasing assets that benefit the public, or even by making payments directly to citizens. Though these sorts of payments are also sometimes described as “victim compensation,” I prefer the term reparations, which makes clear that these payments are not “compensation” in the traditional, narrower sense, but rather payments intended for the benefit of a general populace or society at large. An example of this sort of reparations payment can be found in another case involving the SFO and Tanzania, this one the SFO’s 2010 settlement agreement with BAE Systems for illegal commissions that the company had paid to an intermediary in connection with the sale of an aircraft radar system to the Tanzanian government. (Technically, BAE admitted and was penalized for an accounting offense—failing to keep accurate records of the payments—rather than the underlying bribery.) The settlement required BAE systems to pay approximately £30 million for the purpose of buying educational materials in Tanzania. There is no evidence to suggest that BAE System’s misconduct in connection with the radar system sale caused any damage, let alone £30 million worth of damage, to Tanzania’s education system. So this payment was not “victim compensation” in the narrow sense, but rather an effort to offset some of the damage BAE’s wrongful conduct had done at a more general, societal level.

The legal mechanisms for determining compensation awards, though imperfect, are relatively straightforward. Determining an award of reparations is much more complicated, because (almost by definition) it will not be clear exactly who suffered due to the act of foreign bribery, nor how much loss was suffered, nor how that loss should be recouped. (While the United Kingdom does have “compensation principles” in place which are intended to provide a guiding framework for remedial awards in foreign bribery cases, these principles are phrased at too high a level of abstraction to be much use.) One question that will need to be addressed, and the one I want to focus on here, is whether there must be some kind of nexus between the harm caused by a particular act of bribery and the proposed reparations. Of course, as I have explained, reparations are distinct from compensation, and will not require a showing of a quantifiable harm to an identifiable victim. But does the reparations payment need to have any strong connection—in sector, location, or amount—with the harm plausibly caused by the defendant’s act of bribery?

The BAE Systems settlement discussed above is a nice illustration of a case where there is no obvious nexus between the wrongdoing and the reparations payment. There is no direct connection between BAE’s misconduct (related, again, to a sale of radar equipment) and the investment of £30 million toward improving the station of the Tanzania’s school-aged youth. One might argue that BAE’s payment must be understood in the broader context of the company’s participation in and propagation of a long-running and systemic culture of bribery. But even stipulating that this is so, it is unclear why the appropriate response to BAE’s corrupt conduct would be to invest specifically in the Tanzanian education system. The SFO did not attempt to show that Tanzanian schoolchildren and teachers were the victims of BAE’s bribery, beyond making the general assertion that bribery harms all citizens (in some remote way). Nor was any clear methodology applied in arriving at the sum of £30 million. Why that amount? Why not more? Or less?

This lack of methodology highlights one of the disadvantages of pursuing reparations without insisting upon any nexus between the harm caused by the defendant’s corrupt conduct and the specific form and amount of the reparations paid. If one assumes that remediation payments should be consistent and predictable, and that those with the power to award reparations should in some way be accountable, then the lack of a guiding framework surrounding reparations is a significant concern.

That said, insisting upon a “nexus” between an act of bribery and a remedial response, and demanding some clear formula for calculating reparations amounts, might be asking too much. Imposing such requirements could drastically reduce the number of cases in which reparations are paid. Of course, making a £30 million donation to the school-aged children and teachers of Tanzania may not have a clear, direct connection to the harm that BAE Systems unlawful conduct did to Tanzania’s citizens, but that does not mean that this donation was undesirable as a matter of policy. Enforcement authorities like the SFO, if unconstrained in the sorts of reparations that it can demand in its settlements with foreign bribery defendants, have more freedom to rely on the expertise of the aid agencies like the UK’s Department for International Development (DFID) to invest the proceeds of foreign bribery settlements in projects which will do the most good—and that entail the least risk that the reparations monies will themselves be misappropriated (another common concern in this context). For example, in the Chad Oil case, the SFO transferred £4.4 million recovered from the defendant to DFID, which then invested that money in humanitarian anti-poverty programs.

My tentative view is that, prior to inserting a term providing for reparations into a settlement agreement, enforcement agencies should ask whether it is appropriate to insist upon a nexus requirement. If the answer to this question is no, then it may be appropriate to do as the UK government did in the Chad Oil case, and simply invest the money into public infrastructure and efforts to tackle poverty. In deciding whether it is appropriate to insist upon a nexus, enforcement agencies should consider whether such a nexus exists. This could be done by attempting to discern whether an identifiable portion of society suffered an ascertainable loss. In other words, enforcement authorities can and should direct reparations payments in ways that remediate harms that have an identifiable nexus with the defendant’s unlawful conduct when it is possible to do so. (If, for example, a pharmaceutical firm has paid bribes to government hospital administrators, reparations should be directed to the health sector, rather than education or infrastructure.) But if it’s not possible to establish that the bribe payments were closely associated with any particular type of social harm, then the enforcement agencies should have the flexibility to negotiate whatever sort of reparations payments would do the most good in the affected society. Applying this approach to the BAE Systems case, insisting on a nexus requirement would likely have been inappropriate, as there was no connection whatsoever between the illegal activity (the accounting offense) and harm suffered. Insisting on a nexus would have simply resulted in their being no reparations. I do note however that the case may have been different had the settlement agreement concerned an actual corruption-related offense. If it had been, then one could plausibly argue that the harm had been principally to aviation institutions in Tanzania, or infrastructure more broadly, by increasing the government’s procurement costs.

Yes, this approach might be ad hoc and unpredictable, but ad hoc and unpredictable reparations are better than nothing. And we shouldn’t lose sight of the fact that settlement agreements are the product of a discretionary negotiation process; most enforcement agencies are generally under no requirement to pursue remediation at all, let alone deliver reparations for a specific way. Still, this does not prevent discussion as to what might be best practice, especially when tens of millions of dollars and the interests of the citizens of developing nations are at stake.