Printer Friendly

Partially odious debts?



The despotic ruler of a poor nation borrows extensively from foreign creditors. He spends some of those funds on building statues of himself, others on buying arms for his brutal secret police, and he places the remainder in his personal bank accounts in Switzerland. The longer the despot stays in power, the poorer the nation becomes. Although the secret police are able to keep pro-democracy protests subdued by force for many years, eventually there is a popular revolt. The despot flees the scene with a few billion dollars of his ill-gotten gains. The populist regime that replaces the despot now has to pay the debts to the foreign creditors of the old regime. In effect, the populace suffers the costs of a despotic rule twice: first, when the despot was in power, depleting the resources of the nation and reducing the populace to penury; second, when the creditors whose funding sustained the despot collect payment for the debts of the despot.

The foregoing scenario will strike some as unfair. Others might be less sympathetic, invoking the need for the populace to bear collective responsibility for the contractual obligations incurred by the state. The legal rule here, a creature of public international law, is an unforgiving one that supports the notion of strict collective responsibility for state contractual obligations. Governments inherit the debts of prior governments, regardless of their differing characters, philosophies, or populist bona fides. (1) So the ANC-led government in South Africa inherited the debts of the predecessor apartheid regime. Corazon Acquino's populist government in the Philippines inherited the debts of the Marcos dictatorship. The Islamic Revolutionary government in Iran inherited the debts of the Shah. And so on and so forth. Unwilling to pierce the veil of the state entity and to separate the state from its government, sovereign-debt law views the state as strictly liable for debts accumulated by its prior despotic rulers.

This situation poses a dilemma for international law that private law has been tackling in a variety of settings: who among the innocent parties should bear the loss created by a culprit that can no longer be held accountable? Surely, neither the creditors nor the populace are the primary "wrongdoers"--they did not steal any money or commit any violation. Still, the greater blameworthiness of the ex-dictator's intentional actions should not blur the different shades of fault for the less-guilty parties. If their relative fault can be compared, it might provide a formula for dividing liability among them. (2) Thus, the purpose of this article is to borrow from a rich private-law tradition to explore the treatment of odious debt as a problem analogous to allocation of liability in private law.

The framework this paper adopts is economic. Drawing on the economic analysis of private law, it develops insights as to the structure of an optimal liability scheme. Under this approach, liability is imposed not on the basis of some intrinsic judgment as to the parties' relative blameworthiness, but rather in a forward-looking fashion, on parties who are best suited to take actions to prevent the loss. In addition, liability is imposed on a magnitude tailored to induce an optimal level of precautionary measures.

According to this approach, whether the people of a nation should bear responsibility for the debts of their despotic leaders, even when they have no control over those leaders and their borrowing decisions, cannot be resolved merely by invoking presumptions regarding collective responsibility and the legal entity of the state. From an economic perspective, a basic question is whether the people of a nation have cost-effective means to prevent the accumulation of insolvent odious debt. Are they better positioned than the creditors to take such harm-prevention measures? And if both sides--the people of the nation and the creditors--could have taken some steps to reduce the harm, how should liability be tailored to induce them to do so?

At first cut, a rule that accords creditors the unconditional presumptive right to recover seems undesirable. Such a rule does not distinguish between cases in which creditors share some of the blame and cases in which they are truly innocent. Instead, we examine a scheme under which creditors may bear some of the costs of insolvent odious debt. We argue and demonstrate that often creditors are better situated than the populace to protect against harm of insolvency. Creditors are also rational actors whose behavior can potentially change by an imposition of legal liability. By contrast, the imposition of liability on the populace for the ill deeds of a despot is often utterly useless in terms of providing prevention incentives; the very definition of the despot is one who has taken away power from the populace. In these situations, we argue, creditors should bear at least some of the liability. (3)

To illustrate, take the circumstance of the despotic dictator seeking to supplement his offshore bank account. Creditors will often be in a better position than the populace to monitor and control the misbehavior of this individual. If, for example, the despot asks a foreign bank to deposit funds borrowed in the name of the state into his personal Swiss bank account, the bank can simply refuse to issue the loan on the grounds that this looks like a case of the despot stealing from his people (and after all, the client borrowing the money is the people; the despot is only the agent). In cases in which the money is used to fund transactions, it might not be as straightforward to determine where the funds are diverted. But the point of our liability scheme is to induce the creditor to seek ways to determine how to prevent the potential harm. For example, creditors can make information regarding loans and a lack of accountability public, empowering domestic and international watchdog groups and according the populace or the international community some leverage to prevent the wasteful use of the debt. Effectively then, placing liability on creditors, while increasing the risk they face in the event that funds do disappear, can reduce despotic behavior and, paradoxically, increase the funds available for repayment ex post.

This comparative-fault calculus suggests that creditors should bear liability for the loss. But should their liability extend to the entire debt? Should they forfeit their entire right to recover the loan? The second component of the framework developed in this article focuses on an additional factor: the extent to which the populace benefited from the loan. The point here is, again, simple: Credit that was used in ways that did benefit the populace--in ways that would be regarded as legitimate if used by a democratic regime--should not be forfeited. A creditor's "fault" extends only to allowing funds to be exploited in a despotic fashion; it does not extend to legitimate uses of the funds. Thus, this framework provides enough flexibility to tailor the relief for the populace in accordance with the magnitude of the exploitation it suffered. The larger the fraction of the debt used to build hospitals and bridges, the more should be paid back. Even when the Saddams, the Abachas, the Mobutus borrow money, they tend to steal but a portion. It is only this portion that justifies placing a special prevention burden on creditors.

Thus, putting together the prevention and the relative-benefit factors leads to a liability scheme that would generally divide the loss between the creditors and the populace. The populace continues to be strictly liable for the national debt, but it now has a defense of contributory negligence against the creditors. If a successor government can show that creditors were at fault, the component of the debt that served ill purposes will be discharged. The more the populace benefited from the debt, the more they have to pay back.

Note that "creditor liability" is perhaps a misnomer, an overstatement. It is not that the debtor state should have an affirmative tort claim for "wrongful lending practices" or for "reckless financing" of the despotic regime. Nor should creditors be liable as accomplices for financing the crimes of the dictator. The sole issue at stake, in the analysis here, is the debt--and the normal context in which it arises is in a claim by the creditor to collect it. Thus, it is a defense accorded specifically to the populace that can be asserted only by its representative successor government. Some arguments raised in response to this proposal may be invoked to support a more ambitious scheme of holding creditors tortiously liable and subjecting them not only to the loss of the debt, but also to add-on damages. This is a more difficult case to make, with significantly narrower applicability. It is not the case advanced here.

Finally, there is another aspect to the comparative fault scheme that is worth noting at the outset. In the aftermath of the despotic regime, many countries are insolvent and unable to repay their loans. While there is no international bankruptcy regime that determines the priorities of the different creditors who line up to be paid, our analysis provides a compelling principle of priority. Among the different creditors, those who acted "negligently"--those who were able to better monitor the abuse of the funds--should have an inferior claim and be paid less than the good creditors recover. This scheme then accords justice not only between the debtor and the creditor, but also between different creditors. Creditors who follow responsible lending practices do not stand to lose under this proposal. On the contrary, by subordinating (or eliminating) the claims of the less responsible creditors, the remaining creditors would have a better chance to recover. (4)

This article proceeds as follows: Part II provides background on the problem. It sets out (a) how the state's being the debtor complicates the analysis, (b) the relevance of historical experience in attempting to address the problem, and (c) the relevance of the contractual nature of the contemporary problem. Part III is the core of the analysis, setting out the proposed partialliability scheme. Part IV works through practical factors that might complicate the operation of the partial-liability scheme. Part V concludes.



A. Complicating Factors when the State is the Debtor

The problem of agents, such as corporate managers, taking control from purported principals, such as shareholders, and seeking to abuse their position by engaging in self-dealing is a familiar one. Agency law, and by extension, corporate and securities law, seek to deal with this abuse problem by giving shareholders the right to sue if they can demonstrate to a court that some self-dealing, despotic CEO is seeking to fill his own pockets at the expense of his investors. This legal structure also regulates aspects relating to disclosure of the managers' conduct, the processes by which shareholders are allowed to challenge existing leadership, and the liability of actors that collaborate with, or facilitate, the managers' wrongdoing. With the misbehaving leader of a state, however, things get more complicated for several reasons.

First, with a despot in power, the state itself has no incentive to police the misbehavior of a leader. (5) The central problem is that the despot has taken control of the state. The people of the state, because they are a dispersed collective and fear the coercive powers of the state's police or military machinery, are unlikely to be able to do much. Enforcement, if it is going to happen, must occur from the outside. At the least, the solution to the collective-action problem among the populace has to occur from the outside.

Outside states, however, and the international community at large, do not have enough of an incentive to police domestic despotism. These states, even when their internal processes are working, have an incentive to maximize the welfare of their own inhabitants because those in power want to win the next election. History has shown that maximizing the welfare of a democratic state's inhabitants is often not aligned with policing odiousness elsewhere. Democratic regimes (taken as a proxy for non-odiousness) have supported many dictatorial states over the years in order to further their geopolitical interests. (6) For example, it may be in the interests of a nondespotic state seeking to obtain oil at a cheap price to make a side deal with a despotic leader elsewhere to support his despotism as long as he keeps up the supply of cheap oil. (7) On the flip side, it is also likely that a despotic state with no geopolitical importance or special natural resources will simply get ignored by the world community.

Moreover, because enforcement and policing are likely to occur only after the dictator has been overthrown, there is limited reach to the legal sanctions available. By that point, it is most likely that the bad guy is counting his ill-gotten gains in some undisclosed location, and disgorgement of his wealth is only partially possible. So, unlike the classic agent-expropriation problem, the main practical effect of the law here is to determine who among the secondary players should suffer the burden of the lost funds. Moreover, whereas in other areas agents can be deterred from stealing by the threat of ex post liability, and may even be discouraged from taking positions of power if such positions no longer enable them to steal, in the sovereign context it is quite unlikely that liability for odious debt would deter despotic leaders from assuming power.

Thus, traditional notions of agency and fiduciary law, which operate well in disciplining officers of organizations, do not provide the same adequate framework for state agents, particularly when these are despotic leaders who leave little room for the citizens to engage in monitoring. In the absence of a meaningful agency relationship between a state's leaders and its citizens, a legal scheme that holds the citizenry liable for the debts of its despotic leader is problematic. The presumption of strict governmental succession and of citizens' collective responsibility for loan repayment lacks an incentive foundation.

B. Historical Attempts to Tackle the Problem and Contemporary Context

The question of collective responsibility of a people for the misbehavior of their government, and specifically its debts, has been one of philosophic and legal inquiry through the ages. (8) Aristotle, Hume, and Jefferson are among those who have discussed it. (9) But the narrower question of a legal doctrine of odious debts had its heyday (that is, until now) roughly between 1890 and 1930. This period saw substantial amounts of lending by creditors in richer western countries to less-developed nations. Notably, these debtor sovereigns were not subject to formal sanctions in any court; the doctrine of sovereign immunity protected them. Sanctions for default and other debtor misbehavior were exclusively of the informal variety. And even on the informal-sanction front, there was no International Monetary Fund (IMF) or other multilateral body to perform intermediary functions. Among the informal or extra-legal sanctions of this era, the most well known is "gunboat diplomacy," which was essentially self-help, in that creditors might persuade their government to send in the troops to teach an offending debtor nation fiscal responsibility. (10) Despite the doctrine of sovereign immunity, the leading international-law scholars of the time suggested that there might be a narrow exception to the rule of strict governmental succession. When the creditors knowingly lent to a regime that was ruling without the consent of the governed and was not using the proceeds of those funds in the interests of its populace, debts might be suspect.

A Russian jurist of that period, Alexander Sack, characterized debts satisfying the three foregoing conditions as "odious" (he also characterized the regimes in question as "despotic"). When jurists such as Sack purported to be describing the state of international law at the time, they did not mean that these odious debts were uncollectible as a function of some legal proceeding. (11) Indeed, there was no meaningful legal proceeding to be had. Sack himself was proposing that there be an arbitral forum to resolve these matters. The characterization of these debts as suspect by scholars such as Charles Cheney Hyde (who was closely connected with the U.S. government) was in effect telling creditors that if they engaged in such problematic lending, their home governments might be less willing to apply informal sanctions (such as sending in the gunboats) when the time came to do so. (12)

For the remainder of the twentieth century, the topic of odious debts remained largely dormant, except for occasionally surfacing in particular situations, as when the Iranian Revolutionary Government sought to avoid payment on the Shah's debts and when the Chinese government was sued on unpaid bonds issued by the Imperial government in the early 1900s. And the topic arose in academic assertions that poor nations in Africa and elsewhere should not be liable for debts foisted on them by creditors such as the World Bank who were aware of the corrupt nature of the regimes doing the borrowing. These claims went nowhere.

The U.S. incursion into Iraq in 2001 changed everything. Iraq, under Saddam Hussein, managed to incur well over $100 billion in debt. And the U.S. government, the new power in Iraq, was not inclined to present that bill to its taxpayers (the Iraqi taxpayers were certainly not in any position to repay anything). The Bush Administration thus turned to a different strategy: making public assertions regarding the odiousness of Saddam's debt. Those assertions--which for many in the human-rights community signaled an avenue through which much of the debt of poor nations in Africa and elsewhere could be reduced--spurred new interest in the topic. Since then, Iraq's debt has largely been renegotiated; Iraq received one of the largest write-offs ever (eighty percent) in the Paris Club. (13) And with that resolution, the Bush Administration's interest in asserting the existence of a doctrine of odious debt waned. But not before the genie was out of the bottle: today, with or without the support of the U.S. government, a debate flourishes about what place the doctrine of odious debt (and query as to whether it could even be characterized as a doctrine) has in the modern context. (14) This renewed interest was further heightened when, in October 2006, Norway publicly disavowed the debts owed to it by five sovereigns on the grounds that those loans were made improperly by the Norwegian government. (15) Finally, as of this writing in September 2007, Ecuador is threatening to repudiate a subset of its prior debts on the grounds that these were illegitimate debts made to a prior military regime. (16)

The differences between the modern world of sovereign finance and the one that existed in the 1890 to 1930 period are relevant to the extent that a century-old doctrine is invoked in current debates. That prior world was one of informal or "soft" law. Assertions that a debt was odious were at best starting points in a bargaining among the involved states. In the current era, much is different. Ever since the early 1970s, sovereigns engaging in commercial activities have had their sovereign immunity stripped regarding transactions in the United States and the United Kingdom. In addition, sovereigns themselves, who do the overwhelming majority of their borrowing in these two jurisdictions, explicitly waive their immunity. Of course, obtaining a judgment against a sovereign is only one-half of the trick: one still needs to be able to enforce it.

Enforcement too, though, has become more feasible in recent years. In addition, multilateral institutions such as the IMF exercise oversight of the market and seek to solve coordination problems when they occur. Finally, developments in technology and globalization mean that creditors and creditor-support institutions (such as credit-rating agencies) are far more capable of monitoring sovereign behavior than they were a century ago, when creditors sometimes found themselves issuing loans to fictitious sovereigns. Thus, in the modern era, enforcement of debt is no longer restricted to soft mechanisms such as reputation and bargaining, but operates through formal channels in municipal courts. (17) And when these formal enforcement venues are available, creditors are able to gain full recovery. In effect, then, the existence of rigorous enforcement highlights a basic feature of current law: if the debt is recoverable (and not subject to the odious debt doctrine), it is recoverable in full. This is the zero-one, all-or-nothing nature of the recovery rule.

C. Novel Approaches to the Odious Debt Problem

Recognizing that under current rules creditors often get too much protection, new proposals have been put forth. The most prominent of these proposals favors an ex ante solution, in which sovereigns are in effect licensed ahead of time in terms of odiousness. If they insist, creditors have an opportunity to know in advance whether the debts that are being incurred are potentially suspect under the odiousness doctrine or not (and can either refrain from lending or charge a higher interest rate). (18) In a modification of the foregoing proposal, creditors who wish to lend to regimes that have been pre-designated as odious can do so, and their debts may still be valid if they perform "due diligence" by credibly determining that the proceeds of their loans are going toward non-odious purposes. (19)

This approach calls for a collective international body to make the odiousness determination. How should that body be designed? One area of importance involves the political versus legal nature of this body. Will it be a multilateral organization like the United Nations, which enjoys broad legitimacy but is highly political? Or should it be a court that can predict fairly well how other courts would decide the odiousness issue? If it is a professional legal tribunal (as opposed to a political body), need there be a new sovereign bankruptcy scheme, or will existing municipal courts suffice? Another aspect implicated by the ex ante approach is how centralized the decisionmaking body ought to be. Some proposals call for a single, centralized nonmarket decisionmaker (either a U.N.-type licensor or a specialized sovereign-bankruptcy court). Other proposals advocate a decentralized system wherein disputes conceivably could be resolved in a variety of jurisdictions, perhaps even the jurisdictions that the parties choose in their contracts.

A licensing system would provide creditors with much-needed predictability. With greater predictability, non-odious debt would be cheaper and odious debt would be costlier, both desirable effects. The problem with an ex ante licensing system is on a practical level: Can we expect an agreement by the majority of states over the creation of an international body that will be authorized to designate odiousness? Indeed, the U.N. Security Council already has the power to make such designations and chooses not to do so. Can this body have proper information about the odiousness of a particular debt or regime? If one believes that the primary problem in setting up such a legal scheme is the detection of wrongdoing, then it may be that an ex ante determination is not just difficult to make, but highly error prone. And finally, can such determinations be made without excessive political bias, especially when an odious regime is still in place and has the ability to make deals with other states with political power? (20)

Whether the ex ante licensing approach is viable is dubious. At any rate, it is not our purpose here to explore the different sides of that approach; others have done this in their comparisons of ex ante versus ex post solutions. (21) Still, the interest that such a novel solution has created is an indication of the need for a different regime than the existing one. Part III, following, sets out a framework that relies on ex post adjudication--a more traditional approach--but one that addresses the concerns that the ex ante scheme tackled.



The analysis in this section, the core of the paper, seeks to answer two questions. First, who should bear the liability for a lost debt--a debt whose monies have disappeared into despotic pockets--the creditor or the populace? Second, to the extent that the creditors should bear some liability, what should the magnitude of this burden be?

A. Comparative Fault

There is a paradox in blaming either the creditor or the populace for the lost debt, since the despotic dictator stole from them. Both the populace and the creditors are victims of the wrongdoing by the dictator, not the perpetrators of the harm. Why should the victims be blamed? Fault, under our analysis, must be understood, then, as a relative judgment. Surely, an overthrown dictator is the ultimate, blameworthy actor. But his disappearance makes this fact functionally irrelevant. What is left for the law to determine is the relative culpability of the disputing parties, the populace versus the creditors. Since the loss must fall somewhere, absent a third party willing to provide a subsidy, one of these two parties must end up bearing that loss. It is between these two parties that relative fault matters. Who is more at fault for letting the dictator misuse the money or run away with it?

The problem of allocating liability among the remaining solvent actors is familiar in the law. Does the innocent buyer of stolen artwork get to keep it after the thief or seller has disappeared with his money, or can the innocent original owner recover her painting? Does a person whose identity or checkbook was stolen have to pay for the purchases made by the thief, or do the banks or vendors bear this loss? Does a person who was the victim of defamatory materials that were wrongfully and anonymously posted on a website bear the loss, or should the operator of the website who allowed users to post content be held liable? In these situations, and many others, the purpose of the law is to identify the party who is comparatively more blameworthy and place liability on that party. It is usually the party who could have done more, ex ante, to prevent, detect, or insure against the loss.

This fundamental principle ought to apply in international disputes as well. (22) Who, then, is most responsible for the dictator's opportunities to steal, the populace or the creditors? In a variety of circumstances, the creditors are more at fault because they have greater means to prevent this loss. They are the cheaper loss-avoiders. External creditors who make loans to sovereigns are often able to exercise at least a minimal level of monitoring and control. This monitoring and control can occur in several ways.

Observation and Exit. Creditors are often able to observe a despot's intended use of borrowed funds as a byproduct of the credit transaction itself. For example, when the borrower insists on secrecy and asks for funds being borrowed supposedly on behalf of a sovereign to be deposited in a despot's personal offshore account, the creditor is likely able to observe the misbehavior at little or no cost. Control by the creditor can then be exercised, for example, by refusing to lend the funds or by mandating a different deposit venue. In this scenario, the creditors can both observe signs of misbehavior and protect against it at a low cost; it makes sense to require them to do so. Restricting creditors' ability to recover on their loans under such conditions would give creditors the incentive to exercise scrutiny and control; in fact, creditors are typically in the business of exercising precisely this type of scrutiny and control in the context of their non-sovereign lending.

Disclosure and Empowerment. Creditors can also exercise control via third parties--specifically, third parties who do not have access to information, but who might act (or threaten action) if provided with that information. The most relevant third party here is the populace of the sovereign debtor. Creditors could make public disclosures of the loans being made and of the purported reasons for the borrowing. The information in those disclosures would be demanded from the debtor by the creditors as a condition of making the loan. It is information the government might not otherwise be providing the populace. (23) When loans involve disbursements over an extended period, creditors can require periodic disclosures as to the results of the investments that were made. If, say, loans were taken out in order to build chicken farms, periodic public disclosure of the number of eggs produced could be required, so the public would be able to evaluate the efficacy of expenditures supposedly made in its interest. In theory, the despot, while having stolen the proceeds of the loans, could lie and say that X number of eggs were produced by the farms using the loans and that those eggs were distributed to the populace. But the public, because it has received no eggs, will recognize the lie. And the resulting resentment might become a focal point around which the populace coordinates to overthrow the despot--a risk that the despot will seek to avoid. Public disclosure--especially when it can effect coordination among the disaffected members of a populace--can be a cheap and effective means of empowerment. And from the perspective of the creditors, requiring disclosure would not only be inexpensive, but to the extent it produced constraint and better behavior on the part of the sovereign borrower, it might be value-enhancing.

As noted, some of the information disclosed by the dictator might be deceptive or fraudulent. If neither the creditors nor the populace can verify its accuracy, theft can occur despite this effort. It is important to remember that creditors are not strictly liable for the theft; they are only required to take sensible measures to try to prevent it. Disclosure may serve useful purposes at one time and be useless other times, but may be one of the sensible measures as long as it is cheap to require.

Governance. The third set of actions creditors can take falls into the governance category. As corporate-finance scholars observe, creditors engage not only in exit but also in voice. (24) And these rights to exercise voice come in the form of a combination of covenants and default events. These rights to exercise voice are important because the lending relationship is rarely going to be a zero-one (good-bad) proposition. In any sovereign-lending relationship, when money is about to be lent, the borrower will likely strive to show himself to be pure of heart and intent on investing the funds to serve the populace and increase economic growth. (25) Signs of trouble typically appear subsequently, when funds begin to move from the creditor to the debtor. Initial signs will not tell the full story; trouble may be brewing, but will likely not have emerged full-blown. Creditors can negotiate, ex ante, for control rights to kick in when these initial indications show themselves--indications that funds are perhaps being diverted to ill uses--for heightened involvement or control of the project. For example, debtors can be required to engage external expert monitors to evaluate whether specified milestones for the project are met. If those milestones are not met, either additional control rights for the creditor kick in or the debt gets accelerated. And, if one creditor accelerates, typically every other creditor will accelerate, as well. The result is that creditors have a rather big stick with which to reign in misbehavior when initial signs of that bad behavior show up. (26)

Demand Adequate Assurances. Even if not engraved in the initial debt contract, creditors can often be in a position to demand from the debtor adequate assurances that the funds will be used legitimately. Similar to the right of parties to a sale contract to seek adequate assurances of performance any time they have reasonable grounds for insecurity, (27) creditors can request their sovereign debtor to issue proof as to how funds are being used, and to make these statements public. Thus, to solidify a doctrine that would place some liability on creditors, it might be sensible for the official sector (that is, institutions such as the International Monetary Fund and the World Bank) to supplement private credit agreements with a generalized requirement that, perhaps as a condition of IMF or World Bank membership, grants creditors various rights to seek information and assurances in the course of the relationship, any time prior to the full repayment of the debt, as long as some grounds for insecurity exist.

These actions are only a partial list of the monitoring measures available to creditors. In addition, external creditors are likely better able than the populace to diversify the risk of a misbehaving despot. This, to be sure, suggests that the added liability be factored into the cost of credit, making it selectively more costly and less affordable to regimes that pose a greater risk of invoking the proposed defense ex post. This is not to say that such monitoring opportunities are, in fact, readily available. Effective measures will not always exist. But to the extent that it can be shown ex post that they existed but were not taken, that information should be considered by courts adjudicating the debt claim.

Measuring creditors' negligence, though, is only one side of the equation. Creditors' fault eventually has to be compared to the fault of the other party in the current litigation, the populace. Generally, however, the populace does not have effective ways to scrutinize or criticize the use of the funds. The dictator has taken away the ability of the populace to acquire information about the use of the money, and, more importantly, to do anything about it. The populace might suspect--in fact, it might know for sure--that its leader is stealing money borrowed from creditors. But absent some coordinating mechanism, there is nothing individual members of the populace can do.

Moreover, the problem of placing liability on the cheaper loss-avoider or insurer is not restricted to only two candidates, the populace versus the creditors. Other actors can either help the despot cause the harm or be instrumental in avoiding it. Despots do not usually act alone in enacting their oppression. They need support from actors among the populace who participate in odious transactions (for example, arms sellers whose weapons are then used to suppress a peaceful pro-democracy protest, or buyers of natural resources who pay bribes in return for discounted price); they need states to recognize their regime and to cooperate with it; they need international groups to fail to note the brutality of their regime; and so on. (28) If creditors can supposedly monitor and control some aspect of the behavior of the despotic regime, can other groups and entities not also engage in some monitoring, and ought they not be given the incentive to do so by the threat of liability?

Indeed, some who interact with the odious regime may be able to take better care and so avoid contributing to the harm inflicted by the regime. Creditors, in this sense, are no different from, say, weapons manufacturers, consultants, and the like. It may well be that a cause of action ought to lie against such contributory actors, shifting the cost of the harm done by their contribution or complacence to them and away from the populace that suffered it. To the extent that cheap prevention and precautionary actions can be taken, the same argument developed here would justify assigning liability to these actors. (29)

In addition, creditors' liability under this framework operates merely as a defense for a successor government against creditors' claims to recover the debt. For one, it is not an independent tort for which the populace can affirmatively recover. (30) It applies, at most, to the debt the creditors are seeking to recover. Moreover, as a defense to the strict liability of the borrower, creditor liability is based on negligence--that is, on the demonstrated failure of the creditors to take due care, The reason why creditors' liability ought to be based on negligence (and not be strict) has to so with the purpose of such liability: to induce care, not to shift the cost of the activity. In fact, in many situations, the required level of care is fairly small and not costly. If a small amount of care results in insulation from significant liability, it is likely that the creditor will exercise it. By contrast, if creditors were strictly liable, that is, regardless of how much care they took, they would stand to lose the entire loan, the risk on creditors would be much greater. The implications of such a rule would be far-reaching in terms of interest rates and the ability of dictatorships to borrow at all (the "activity level" effect, as it is known in the economics of tort law). This is not the place to weigh whether such constrictions on the activity of dictatorships are desirable. Arguments can be made that suffocating the credit lines of an odious state can have desirable effects (for example, increasing internal pressures to change the regime, reducing the incentive to capture control of the state) as well as undesirable effects (for example, increasing the distress of the populace). The point here is that strict liability--an absolute defense, regardless of creditors' efforts to monitor--would have an effect beyond inducing creditors' care. It would shift much of cost of the odious regime to creditors. It would probably have a crushing effect on lending activity to dictatorships. (31)

The argument in this section is simple: place liability on the least cost-avoider. If creditors are better preventers of the loss--if the government that succeeded the dictator can show that creditors failed to take due care and to monitor the funds that were stolen--creditors should bear the loss of their failure to take care.

B. Comparative Benefit

Not all borrowing by a despotic regime is bad. Although some of the borrowed funds end up stolen by the odious dictator, often significant fractions are channeled to desirable purposes. In fact, both the legitimate and illegitimate uses are often tied together in the same transaction. Money that is borrowed, to say, build bridges and schools, is partially used to that end and partially stolen by the corrupt dictator and its bureaucracy. It is rare for modem despots to steal all of the funds that an external lender might provide (even Saddam built hospitals). At worst, a fraction goes into the despot's pockets. That means that at least some goes to serve the oppressed public.

If the funds borrowed from a specific creditor were all channeled to a legitimate purpose, there would be no reason for leveling any type of liability on that creditor. (32) In other words, creditors' liability ought to be associated with the social harm from their acts. If there is no social harm if the money was used for purposes other than the dictator's self-enrichment--there is no ground for liability. There is no reason to induce creditors to engage in monitoring if, even in the absence of monitoring, the funds are used legitimately. It is only the fraction of the funds that were stolen that should potentially give rise to liability (and even then, only if the creditors were negligent).

The liability rule should give an incentive for creditors to ensure that their funds produce benefits. One way to do this is to reward creditors whose funds provide benefits by providing insulation from liability. The question is how much insulation is required. Under the Sackian odious debt doctrine, even a tiny benefit to the populace insulates the entire debt amount from the odious debt defense. (33) That provides nothing but the most minimal incentive for creditors. Under our framework, instead, creditors whose loans provide greater benefits should get greater amounts of the debt insulated from liability. This way, creditors have an incentive to demand credible information from the debtor demonstrating that the funds produced benefits. The greater the benefit demonstrated, the better the insulation from odious debt liability. In the end, creditors ought to be liable only for the fraction of the debt that did not serve legitimate benefits. Their liability extends in magnitude solely to the funds that were stolen or otherwise used in an odious fashion. Put differently, if the debtor can demonstrate positive uses for the funds--such as hospitals and schools--there seems to be little need to impose monitoring responsibilities on creditors.

The insulation from liability is akin to a security interest. To the extent the creditor can demonstrate that there was a benefit from its loans, a portion of its loan gets protection (security) against the odious debt defense. And, as with a piece of collateral, that reduces monitoring responsibilities for the creditor (they need only watch over the collateral); creditors who can ensure that there was a benefit derived from their loan reduce their need to monitor the general misbehavior of the debtor. Indeed, what should happen under such a regime is that external creditors will begin to make loans only for projects in which corruption and graft are minimal and in which the positive benefits from the use of the loan's proceeds (for example, the number of liters of milk produced from a cattle farm or number of children graduating from school), are readily visible. More fuzzy projects whose output benefits cannot be easily measured will become disfavored by external creditors.

The critical question is how to measure benefit. It is, in a sense, the flip-side of the problem of how to measure harm. Benefits from protecting a group of starving children from death are arguably incommensurable. But there is simultaneously the matter of those same funds having perpetuated the despot's reign; allowing him to starve even more children equates to commensurate harm.

Although measuring output (harm or benefit) may be difficult, measuring input may be less so. The problem here is akin to the familiar problem in labor economics of measuring a worker's output. In situations in which output is difficult to evaluate, employers will often reward employees according to input instead (lawyers, for example, are often paid on the basis of the number of hours they work). Applied here, if benefits are appraised in terms of how many dollars from a loan went into the production of benefit, the result is measurable. One no longer has to worry about how many starving children were saved or even whether the children were starving at all. The question is simply what fraction of the debt went to providing benefits. The downside here is that creditors have no incentive to ensure that their dollars go to the uses that produce the maximal benefits. The upside is that the creditors cannot argue that a small amount of their dollars providing a large benefit to some starving children justifies their having allowed the despot to steal a large fraction of the loan. Under the input-measurement scheme, the amount of dollars going to provide benefits would be the amount of liability insulation, no more.

Even if inputs, rather than outputs, are measured, it may be difficult to figure out which inputs gave rise to benefits, namely, how the money was used. If uncertainty prevails, it may be necessary to use presumptions, and we will further discuss the evidentiary issues below. The point here is more basic: if it is evident that some of the debt was used to benefit the populace, creditors should be able to recover this portion of the debt. Limiting the right to recovery would amount to unjust enrichment.

C. Partial Liability: Breaking the All-or-Nothing Structure of the Odious Debt Doctrine

The sum of the two elements above--liability for the failure to exercise due care and security when the provision of benefit is assured--produces a partial-liability scheme. Under this scheme, creditors are liable for portions of the debt that are obviously being used for the sole benefit of the despotic leader. The same is true with portions of the debt used to pay bribes and the like (the corrupt payments that the lender makes to induce the lending). Liability becomes more complicated when some portion of the debt is being used for less facially corrupt purposes, such as paying the salaries of the police (who then may do the work of putting down public demonstrations) or purchasing weapons for the military to ostensibly defend the nation. Here, liability will need to be apportioned based on the comparative fault of the various misbehaving parties involved. By contrast, under this partial-liability scheme, creditors are fully immune from liability for the portion of the debt that was used to benefit the populace.

This rule breaks the existing all-or-nothing structure of creditors' liability. Currently, the rule of strict governmental succession is one in which the debtor nation is always liable, despotic debts or not. That is, zero lender liability. Under the Sackian odious debt doctrine, if certain conditions are met, lenders must bear full liability. (34) Between these two polar solutions, there is no occasion for intermediate liability, no sharing rule, no balancing of the conflicting objectives. There is either full debtor liability or full creditor liability; there is no partial odiousness. (35) The problem with this structure is that it does not leave enough room to make comparative judgments of fault. Some creditors might know perfectly well that the loans go to despots and will be used in a bad way. Other creditors might only suspect but not know for sure. Still others suspect that only a fraction of the loan will be misdirected, with the other fraction used for good purposes. Moreover, creditors may have different degrees of suspicion, some more and others less certain about the bad use of the money. Finally, and perhaps most importantly, some creditors can do something about their suspicions and monitor the use of the funds. Should all these creditors be treated alike? Should the law ignore this information about the relative odiousness of the debt? Should debtors not get at least a partial break in some of these situations? The most basic feature of this comparative odious debt scheme is the division of liability: its departure from the rigid structure of all-or-nothing underlying existing doctrine.

The criterion for apportioning liability here is creditors' misbehavior; it is not that the debtor is very poor and cannot bear the burden of repayment. This latter "bankruptcy" justification for debt forgiveness may apply independently and in addition to the fault-based forgiveness scheme. In fact, to the extent that the debtor cannot pay back all its debt, it becomes all the more critical to apply a criterion for priority among creditors. One can thus view this proposal as a comparative-fault scheme, not only between a debtor and a given creditor, but also among the competing creditors. (36) Those creditors who are less at fault can recover more of their debt. This is precisely the spirit of bankruptcy law's "equitable subordination" doctrine, under which courts can change the priority in bankruptcy against those creditors whose conduct is inequitable. (37) Comparatively faulty creditors will recover less, under our scheme, not as a result of direct bankruptcy rules of priority, but rather by reducing the size of their legally recognized claim.

So while the poverty of the debtor does not directly indicate blameworthiness, it is nevertheless a factor that comes into play. First, it could operate to raise the level of due care on part of the creditors. That is, when lending to a desolate state captured by a despotic ruler, creditors have to be all the more careful and inquisitive. And second, anticipating potential insolvency, the creditors are all the more driven to monitor the use of the funds. In these dire situations, there are many ways for the funds to benefit the populace of the recipient state; creditors can more easily demand affirmative assurances that the money is used legitimately.



Proposing a partial-liability framework may sound plausible in theory. But where the rubber meets the road is in terms of implementability. An all-or-nothing scheme does have the advantage of being simpler. An intermediate-liability scheme adds new factors on which the legal consequences depend, making it more complex. This section addresses several complexity and implementation questions and responds to some principal objections.

A. The Availability of Substitute Pools from Which to Steal

Even if imposing liability on creditors would force them to monitor and would ensure that the borrowed funds are used for legitimate purposes, stealing by a despot might not diminish. Now, instead of stealing from borrowed money, he would steal from other resources, such as tax revenues or foreign grants. Put differently, the more a despot is driven to apply borrowed funds to essential uses that benefit the public, the more tax revenues and other forms of revenue are "freed" to be skimmed. Whereas debt would be monitored by creditors, tax and grant revenue is not. Thus, under this scenario, liability on creditors would not be effective in reducing the theft. In fact, it would operate only to increase interest rates and reduce borrowing.

This argument assumes that the different sources of revenue are perfect substitute resources for the purpose of theft by a despot. There are reasons why this assumption might be false. Consider, first, the substitution between credit and tax. Tax revenues tend to be relatively visible and therefore harder for a despot to steal without raising ire; external borrowing is relatively opaque and funds thus provided easy to pilfer. Tax revenues are collected internally by local tax agents. There will be processes in place to make sure that those revenues are accounted for, and that local agents do not steal the funds. By the time the funds get to the level of the despot, they would be difficult to steal without people's noticing. By contrast, funds provided by external borrowing are relatively easy to steal, for they come more directly to a despot than tax revenues. Thus, a dictator who stays in power by keeping the collective from rebelling against him must pick his opportunities to steal in a way that is less exposed, that hides from the public the true extent of his misbehavior.

Further, internationally borrowed funds arrive in foreign currency. Tax revenues, on the other hand, are collected in local money and have to be converted into a foreign currency before they can be meaningfully stolen by the despot (who is presumably planning to squirrel away his ill-gotten gains in some foreign jurisdiction). If the despot targeted tax revenues, he would face the additional hurdle of exchanging large sums of money, possibly exposing the corruption and likely wreaking havoc with the exchange rate and, in turn, bringing about additional distortions on the economy. By contrast, stealing a portion of the proceeds of an external loan, when the stealing occurs in the Caymans or in Switzerland, would attract less attention.

Finally, stealing debt is easier than stealing tax revenues because the populace cares less. When tax revenues are skimmed, it is the current generation that pays. When debt revenues are skimmed, it is the future generation (which will have to pay the debt) that bears the burden. So long as the current generation does not feel its pocket directly picked, it might not be as likely to revolt. Indeed, the clever despot will borrow on behalf of later generations and spend some of the proceeds of that borrowing on the current generation and steal the remainder. Thus, to the extent that public outcry is one mechanism that disciplines theft, it is more effective when the stealing targets tax money. (38)

What about stealing foreign grant money? The majority of borrowing by contemporary despots like Saddam Hussein and Sani Abacha was in the form of official credit. (39) Other sovereigns provide grants (albeit formally structured as loans) that are often designed to help the despot stay in power so he can continue serving the interests of the donor country. Would a dictator who is monitored by his commercial lenders target instead his country's foreign aid?

True, greedy despots might turn and skim off some of the foreign aid. If this foreign aid were in the form of credit, the successor government might want to have a similar defense against repayment as it has against commercial creditors. But since foreign aid is largely in the form of grants, the potential for increased stealing from this source is not as bothersome. It does not raise the repayment burden on the populace, nor the tax burden. It might come at the expense of other, more essential projects that were meant to be financed by the grant money and to benefit the populace. But here again, a donor state who cares that its support money benefits the oppressed populace can take affirmative steps to monitor the use of the grant. Indeed, stealing from foreign credit or grants is often more difficult than from external private credits because (a) foreign governments typically want something significant for their support (oil, for example, or suppression of Islamic or other political movements) and (b) foreign governments can monitor their funds quite well.

A fuller treatment of the substitution problem would recognize that a despot's choice of a source of theft is generally wider than taxes, grants, or external credit. Often, the country will have other sources of income such as export revenues (for example, diamond exports from Sierra Leone, oil from Venezuela, cocoa from Ghana, copper from the Congo) that are also susceptible to stealing. A despot that cannot borrow foreign money can instead sell its country's resources to foreign buyers for some side kickback. This suggests, perhaps, that a similar fault regime ought to apply vis-a-vis blameworthy buyers, to further shut down the opportunity to steal. But the basic point continues to hold: external sources of funds--money received from lenders and buyers in foreign currency on foreign land--are far more attractive (or available) to steal from than money raised through an internal bureaucracy from local sources and in local currency. The partial odiousness scheme targets and protects these vulnerable sources.

B. Distinguishing Between Legitimate and Illegitimate Uses of Funds

With any scheme imposing liability on creditors for a failure to monitor, a threshold question will be how creditors are to distinguish between legitimate and illegitimate uses of funds. Even some of the most despotic uses of funds (such as purchases of arms to assist the despot's military or the building of a giant statue of the despot on a horse) can have legitimate aspects (the military keeps the peace and these statues someday become tourist attractions). Are credit institutions capable of determining what uses of funds are legitimate? The partial-liability scheme then presents the concern of forcing them to be excessively cautious in their lending decisions--that is, lending only when the legitimacy of the project is certain.

Note, though, that this danger for creditors--this impediment to lending--is a different manifestation of the problem that creditors are limited in their abilities to police wrongdoing. This is why the optimal liability scheme needs to be calibrated to require them to take action only when reasonable. If the dictator is in the middle of a ruthless repression of an ethnic minority, in which, for example, he is illegally using poisonous gas, creditors should probably note that the letter of credit is being used to purchase additional poison canisters. But due care on the part of creditors will not enable them to control every transaction.

Note, too, that even if creditors are excessively strict in demanding that funds from their lending be used for purposes that are clearly legitimate (for example, only for chicken farms when the output is distributed equitably among the populace by internationally recognized NGOs), this is not necessarily bad. If the despot thinks that a high-risk project (that is, one with a high theft potential if financed externally) is in the interest of the populace, he still has access to tax revenues. But this option will be unappealing because the level of monitoring and control by his populace over the use of those revenues will be high compared to that for external debt. The populace will likely prefer the second scenario, in which the despot's spending, especially in terms of pet projects or general expenditures, is controlled. The optimal system then, from the point of view of the populace, will be one in which the despot is restricted from financing projects with a high theft potential with external debt. For those projects, he needs to go to tax revenues and justify the spending to his inevitably skeptical and risk-averse public. (40)

The question of what to do about the ambiguous cases is still unanswered. Working through the example of spending for military equipment should help do so. Given that a significant portion of odious lending likely occurs to fund military spending, that example is particularly apt for any discussion of how creditors would detect odious debts. (41) Military spending could arguably be for purposes of protecting against external threats or internal protests that destabilize the economy, but military expenditures could also be used to oppress peaceful internal dissent seeking to overthrow an evil despot. How should a creditor evaluate such a case, especially given the frequent possibility that the spending will be applied to both purposes? (42)

In light of this ambiguity of where to draw the line between legitimate and illegitimate uses, it might be tempting to say that creditors should be in the clear in such cases. After all, whether military expenditures are needed are matters of sovereignty, the kinds of matters that foreign creditors lack expertise in determining. (43) But perhaps the liability regime proposed here offers, once again, a way out of zero-one determinations. Better information is needed regarding the actual use of the military borrowing. When the potential use of loan proceeds is ambiguous and there is a high danger of despotic behavior, or when there are indications of past abuse of military spending, creditors should demand additional information. With more information, creditors could be more prudent in the release of funds. (44) The information might not always enable them to discern the legitimacy of the funds' use. But demanding that information would at least go a significant distance towards satisfying the due-care burden.

Perhaps there is room for more hardened presumptions regarding military lending. Because of the likelihood of corruption, the potential lack of public disclosure, and the possibility of the goods being used to harm the populace, most creditors know such credit to be suspect, and it should be made only with greater reluctance. To induce this reluctance, ex post scrutiny of such funding should be heightened. This way, creditors will not as easily accept the excuse that the nature of the goods and their quality cannot be disclosed for reasons of national security. Of course, creditors' concern about a court's subsequent questioning of the legitimacy of military spending is going to raise the cost of capital. This means that when it comes to military spending, those with despotic tendencies will be forced to look to tax revenues, where scrutiny will be more exacting and vague assertions of national security as a justification for secrecy will be less effective.

In the end, creditors may face some significant uncertainty regarding the odiousness of the debt they are accepting. But is the problem of uncertainty here more troubling than other types of uncertainty that potential injurers face when they have to choose levels of care? The phenomenon of uncertainty about the legal standards of due care appears in various other contexts, and has systematic implications for the behavior of tortfeasors and for the design of the legal standards. (45) In principle, it does not pose a unique challenge in the context of odious debt.

C. Monitoring Abilities of Bondholders Versus Banks

Determining whether the creditors knew or should have known of the regime's propensity for misbehavior often will be a difficult question. There will be the occasional easy cases in which the creditors directly deposit the loan proceeds into the personal Swiss bank account of a despot; there, one can safely say that the creditors knew or should have known that stealing was occurring. Those cases were likely easier to find in the days when sovereigns received most of their financing from large multinational banks. But a large portion of sovereign debt today is in the form of bonds. These are liquid instruments, and the individual holders are often atomistic and do not even know what bond they are holding in their portfolios at any given moment. Not only are these creditors unlikely to know much about the behavior of the governments to which they are lending, but the original creditors who did the lending are almost never going to be the ones eventually suing to get paid on their debt instruments. As a matter of implementation of the partial odiousness rule, it might be conjectured that the cost of determining what creditors knew or should have known are insurmountable. Assuming that any given sovereign is likely to have thousands of creditors scattered around the world at any given time, it seems impossible to make determinations about what these creditors knew at the time the loans were made. How then, can liability be placed on decentralized, uninformed investors?

The presence of an active and liquid bond market may ultimately provide more solutions than problems for the institution of a creditor-liability framework. It should be noted, though, that despotic borrowing is rarely in the form of bonds. (Saddam's $130-billion debt, for example, had no bond debt in it at all.) And although the question has not been studied, any bond issue--but especially SEC-registered offerings--requires due diligence by the underwriters, a fact that would likely result in reduced enthusiasm from despots for such debt.

More importantly, even if sovereign bonds are to be issued by a despotic regime, information about the liability risk would be reflected in the price of the security. As is well known, when small investors participate in many small-scale market trades--of any kind of security or tradable commodity--the individual investor need never obtain full information about the fundamental economic value of the traded asset. When efficient markets develop for such assets, the prices of the assets reflect all publicly available information. Sophisticated traders, brokers, and various information intermediaries obtain information about the value of assets, and their own buy-or-sell decisions provide the aggregate information to uninformed, unsophisticated individual traders.

This is particularly true for bond markets. If the market is active and recognizes that certain types of debts will be less likely to be repaid (for example, because the populace has not benefited from the debt in question or there was corruption involved in the creation of the debt), then the market will incorporate that information into the price. If information as to despotism is important for pricing--as the threat of liability will force it to be--then there will be an incentive for institutions to emerge that can acquire, assess, and sell this information. Other institutions might emerge to provide insurance for this particular type of risk--for example, swap contracts, in which the risk of odious debt liability is exchanged.

Creditors in an active market will not actually do any of the monitoring of debtors themselves. Rather, creditors will be able to purchase different types of bonds--costlier bonds that are certified to be free of odious debt liability or bonds that are bundled with liability insurance. In the same way that bondholders purchase municipal bonds of entities with which they have no familiarity, whereby they rely on market institutions such as underwriters and rating agencies to collect the necessary information and sell it to them, bondholders purchasing sovereign debt will rely on sophisticated intermediaries to ascertain the risk for odious debt liability. (46)

Nonmarket institutions are also likely to play a role in taking on monitoring tasks, especially if there is a need for ongoing monitoring of projects. To the extent multilateral development banks (MDBs) like the World Bank begin monitoring for corruption--and this is something that the World Bank has already begun to do in spades (47)--monitoring may be delegated to these organizations. This delegation would produce what one might call a multiplier effect vis-a-vis the World Bank's current anticorruption efforts. Most sovereign bonds already specify that it is an event of default under those contracts for the sovereign to lose either its membership of the IMF or its eligibility to draw from the IMF. (48) If the World Bank were to continue policing states for corruption levels, creditors could write in a similar event of default with respect to the Bank or other MDBs. (49)

If, under this proposed creditor-liability framework, private-lending institutions develop odiousness risk ratings for sovereign bonds and price these bonds accordingly, the scheme would begin to resemble the ex ante certification mechanism proposed by Kremer and Jayachandran. (50) As in their mechanism, in this framework the original lending decision is conditioned on the absence of odiousness (defined in governance terms). If the sovereign passes the test at the initial stage, the lender need have little fear of a subsequent legal challenge based on odious debt grounds. This scheme differs from the Kremer and Jayachandran idea in that no official international institution needs to determine which regimes are odious. Instead of a centralized political body making this decision and facing controversy and legitimacy concerns, here it is a decentralized assessment made by the market, manifested by whether the MDBs are prepared to lend.

D. Ex Post Verification and Other Problems in Adjudication

For a liability scheme to work, a court or other adjudicator has to have enough information to make correct decisions. Specifically, courts will need good information after the fact to determine (1) what creditors knew or should have known and (2) what benefits ensued from the debt. Creditors, it can easily be imagined, would likely deny in court that they had the ability or information to monitor the borrower, even when they did. And successor governments would perhaps argue that the debt was odious and unbeneficial to the populace, even when it was not.

That courts might make errors and fail to accurately determine liability and its magnitude is not necessarily an argument against the partial-liability framework proposed here. Any schema imposes errors, even if it does not require courts to affirmatively assess liability. The current scheme of zero odious debt liability is no exception. It makes the most extreme "error" of one type (the error of failure to inflict liability when doing so is socially desirable); at the same time, the current scheme makes no errors of a second type (the error of labeling a legitimate debt odious and relieving states from a socially desirable obligation to pay it back). A shift towards some liability on creditors would reduce the first type of error and increase the second type.

Is the ex post liability scheme more prone to error than other proposed mechanisms to resolve the problem of odious debt, such as the ex ante certification regime? Probably not. It is likely that information about the way funds were used is easier to dig out after the fact, after the odious regime was thrown out, than during its tenure. When the despot is in power, he has control of the state-police machinery and is presumably using it to further corruption. Outside investigations can be conducted, but those investigations are inevitably difficult if the state is seeking to hide information. (Just think of any one of the recent attempts by external investigators to determine the extent of some rogue state's Weapons of Mass Destruction or nuclear capabilities: not only are these investigations near impossible to conduct, but the results of the investigations can produce significant errors.) By contrast, once the despot has been removed and the new regime is seeking to determine how much corruption there was, there are often strong incentives for those with information--for example, members of the prior regime who want to demonstrate that they are distancing themselves from the prior regime--to come forward. The Oil-for-Food Program investigation led by Paul Volcker is a recent example of how such an investigation after regime change can yield enormous results.

Good information as to the benefit to the populace from the loans is also more likely to be available after the despot has been dethroned. While the despot is in power, he will seek to spin the evidence to demonstrate that the loan proceeds are being used to benefit the populace. And the creditors, at least the misbehaving ones, have every incentive to help the despot produce evidence suggesting a project that they financed was a legitimate, welfare-enhancing one. (51) After the despot is gone, information is available about whether the ostensible purpose for borrowing was false or not. Imagine an arms purchase--a favorite loan-driven purchase of despots who are seeking to siphon off money into their personal accounts. In theory, arms purchases could be a legitimate benefit, and it is hard to know in advance how they will be used by the despot. Ex post, concrete evidence is more likely to be furnished on whether these arms were used either as a front to obtain bribes or, worse, used to harm the populace.

It is not enough to argue though, as we do, that ex post information is more accurate than ex ante information. We have to show that ex post, it is possible to make determinations as to what creditors know, or could have known, ex ante. While it might be easy to verify ex post that funds were used to brutalize minorities, can courts verify also that the creditors know how the funds were going to be used? For some of the same reasons set out above, the answer should be in the affirmative. Once the despot has been overthrown, officials in the overthrown regime who are seeking to curry favor with the new regime have an incentive to reveal exactly what the creditors knew ex ante. Even if such information is not available, it is likely the generalized information regarding the practices of the prior regime--such as the fact that they took a customary five percent bribe on every loan transaction--will become available after the overthrow. This generalized information might cause a court analyzing the odiousness question to shift the burden of proof on to the creditors to demonstrate that they had somehow not had to pay the customary bribes.

Another problem has to do with incentives to litigate an odious debt defense. It might be feared that the successor government will have an incentive to tar the performance of the prior regime, bringing forth skewed evidence about the odious uses of the debt and leading courts to make erroneous findings. Contemporary reality, however, suggests that this should not be a big concern. Sovereigns are loathe to tell their creditors that they will not pay their debts, even when it is patently obvious that at least some of those creditors not only helped prop up an illegitimate regime, but also knew that their funds were being used to oppress the populace. South Africa, Iraq, Argentina, and the Philippines provide vivid modern examples of instances in which the new regime pays debts of a prior regime that is widely agreed to have been illegitimate. If anything, sovereigns seem unduly risk-averse in asserting that the debts of prior regimes are illegitimate. (52)

Perhaps this predicted reluctance to overassert an odious debt defense would be because new rulers in many previously odious dictatorships are not as clean as one would hope; they want to be able to borrow and steal money too! Getting on the moral high horse to refuse illegitimate loans of the prior regime would reduce the likelihood of getting new improper loans. Further, even a pure-hearted, democratic regime replacing a kleptocrat might choose to forego the odious debt defense. The new regime is inevitably seeking to keep itself in power and needs external financing to please the current populace (even if additional burdens are placed on later generations). Creditors might condition the availability of new funds on a promise for continued repayment of the old, and odious, debt. As is the case in business bankruptcy, whereby creditors of the ongoing concern often condition new debt on some form of de facto priority with respect to old debt, the case of "sovereign bankruptcy" might exhibit similar problems, leading to underassertion of the odious debt defense. (53)

Whether an odious debt defense would be over- or underasserted is, therefore, hard to predict. The two effects pull in different directions. If the dominant effect is the underassertion of the defense, potential solutions include, as Paul Carrington suggests, granting private rights of action to individual citizens. (54) As a practical matter though, it is premature to deal with solutions to second-order problems that might never arise. Only once an odious debt defense is in place will it be possible to tell what type of fine tuning is needed to adjust its proper scope.

E. Cost of Credit

The partial odious debt framework is likely to make sovereign debt more difficult to collect. This effect might hamper some creditors more severely (those that issue bad credit), and others less significantly (those that issue legitimate funding). But so long as it is not fully ascertainable to creditors ex ante what the status of their debt will be, there is now an added risk of debtor immunity. And even if courts make sound ex post determinations in this area, it is possible that more nations will try to assert an odious debt defense, making repayment delayed and more costly to enforce. These effects, a critic might point out, will not escape the attention of potential creditors. It is likely that creditors will offset this added risk by increasing the cost of loans.

It is not altogether clear that this conjectured effect on the cost of loans is bad. If credit becomes unaffordable to some dictators, let it be so. The key, of course, lies in the quality of information that creditors have when they price the loans. The better the prediction that creditors can make regarding the likelihood of a successful odiousness defense, the more likely it is that the added price of credit will reflect a true fundamental problem with the loan. If, say, creditors can perfectly differentiate good from bad borrowing, they will know that only the bad borrowing carries the new, heightened risk of nonrepayment. Only this kind of debt will be subject to an increased price.

Moreover, in this environment in which the odiousness risk is reflected in the cost of capital, good ex ante incentives might emerge. First, borrowers will have a stronger incentive, at the margin, to substitute some bad borrowing for good borrowing to enjoy the price discount. That is, they will turn to projects that more clearly service their populace. Second, borrowers will have a stronger incentive to assure creditors that the debt is good and thus likely to be repaid, rendering the creditors' task of monitoring the uses of the funds more tractable (and also making the assumption that creditors are able to price-differentiate the risk more plausibly). Third, the ability of bad creditors to collude with despots to reduce payments to the good creditors so as to increase payments to the bad ones will diminish. (55) Despots may no longer make credible commitments to repay bad loans--those might be discharged by a subsequent government or perhaps even by private suits (to the extent private plaintiffs are allowed to derail the repayment of bad credits). Thus, for a variety of reasons, it is plausible that the added cost faced by creditors under the proposed scheme will be borne disproportionately by those that more often engage in bad lending.

F. Circumvention by Creditors

As with any legal scheme that imposes penalties, there is a risk that the wrongdoers (here, the creditors) will take actions to circumvent the scheme and therefore avoid the risk of penalty. To the extent circumvention is easy and inexpensive, the legal scheme will fail. For it to work, the cost of compliance must be low--so low that creditors should not have an incentive to circumvent it. Several circumvention techniques are hypothetically possible, but they can be addressed.

Opting Out. If one jurisdiction (say, either New York or the United Kingdom, the two that dominate cross-border lending) adopts a version of the partial-liability framework, creditors will simply move to a jurisdiction that is more lenient. Surely, it is in the interest of both creditors and odious borrowers to shortchange any future successor government and deprive it of the repayment immunity that the liability scheme might afford it. But this concern proves too much. Under this logic, there is no room for policy-setters to debate any element of the optimal odious debt regime. No matter what the rule is, creditors and borrowers will flee to the most lenient jurisdiction.

One way to avoid this race to the bottom, and to render odious debt rules effective against choice-of-law circumvention, is to have them adopted as part of an international scheme that simultaneously binds all jurisdictions. The sovereign-bankruptcy scheme favored by some, including Joseph Stiglitz, would fit such a mold. (56) But the problem with this type of international scheme is that, at least based on the markets' recent and resounding rejection of the IMF's proposed Sovereign Debt Restructuring Mechanism, it is unlikely to come to fruition any time soon. As a practical matter, to the extent odious debt conceptions are adopted, these adoptions will likely come via municipal courts.

Moreover, this opting-out concern overlooks another important aspect of lending markets--network effects. Research and recent events indicate that the markets for sovereign bonds are characterized by significant network effects. (57) The bonds, whether under New York or U.K. law, use standard contractual provisions and are traded with the assumption that they are in fact standard. For any individual issuer to switch to the law of a different "shelter" state would likely be prohibitively expensive. The switch could realistically only occur if important market players coordinated. And coordination on a global market of this type is expensive and complicated. Surely, legitimate borrowers and their creditors will have no reason to participate in the coordination and to switch to new states with new laws--they have no reason to fear odious debt liability. In addition, it is likely that these private actors will need the assistance of the official sector to help them coordinate. And the official sector, especially if it is backing an odious debt initiative, is hardly likely to support efforts to shift the governing law of contracts to, say, Mongolia.

The harder question has to do with the possibility of competition between New York and the United Kingdom; the two already-established jurisdictions whose laws govern standard sovereign bond documents. Take a government such as that in China or Russia today. It might fear that a future regime would seek to repudiate some or all the debts that the current regime incurred. China has issued debt in both the New York and U.K. markets. If New York courts began showing sympathy to odious debt defenses, China could easily switch its borrowing to the London market (indeed, Russia today issues exclusively under U.K. law). But would China wish to do this? To take such an action would signal not only to its creditors (current and future), but also to its own populace, that it is concerned about the legitimacy of the uses of borrowed funds. Nor is it at all clear that China's creditors would prefer a legal scheme in which loans by good creditors were treated the same as those by bad ones. Most creditors likely prefer a scheme in which they know that their fellow creditors are helping ensure that the debtor is not being run by a kleptocrat. (58) Finally, for political reasons, it is hard to imagine either the United Kingdom or New York seeking to be the regime that attracts the borrowing business of that subset of sovereigns that signals itself to be oppressing its populace.

Creative Refinancing. Another potential mode of circumvention is "credit laundering." (59) The idea is that creditors lending for an odious purpose would then immediately lend additional funds so that the odious debtor could repay the first (and most problematic) debt. The second debt then would have been incurred only to repay a prior debt. This chain can be extended (a third loan to pay the second, and so on) until the connection between the last link in the loan chain and the original odious purpose is so vague and indefinite that it is unlikely to be detected by courts.

Could the foregoing be done? The IMF has done it in the past, with the IMF's goals usually being to protect against the possibility that the debtor might default on its IMF obligations. But the continual extension and refinancing of the first loan is hardly going to be cheap. Lenders do eventually want to get repaid, and they would worry about lending when the borrower might have a defense against repayment. There is no guarantee that the sovereign's records will not reveal that these refinancings lead back to an original odious loan. (60) True, a subsequent lender may be oblivious to the original odious purposes of the debt--especially if this lender is the unsophisticated bond investor. But, again, market forces are likely to handle this problem through information intermediaries. Finally, even if some such circumvention worked in shielding creditors from liability, it would raise the cost of borrowing for despotic dictators, reducing their opportunity to steal.

Ex Post Circumvention. Creditors might also bypass an odious debt doctrine by making side deals with the successor government. Creditors could offer the new government fresh loans, with the promise to turn a blind eye to any new stealing, in exchange for the promise that the new government will not assert any odious debt defenses. To make the deal sweeter, creditors could also offer the new government additional loans to help make payments on the old odious loans. In other words, to assert an odious debt liability claim, the new agent of the populace must be immune to bribes. The political and bureaucratic culture that enabled the previous despot to claim power might tolerate subsequent-albeit more minor forms of corruption.

This is possible, but it is also expensive for the creditors. Creditors engaging in such circumvention actions are not only paying bribes, but are making additional loans that are just as, or even more, susceptible to odious debt defenses. As with a creative refinancing problem, the essential characterization of the debt as odious survives the multitude of iterations of subsequent refinancing. Once again, the point is not that the solution proposed makes it impossible for creditors to engage in such circumvention, but that such actions are likely more expensive than the alternative of simply taking the due level of care that our comparative-fault framework requires.

G. Shifting Uses by the Debtor

Another circumvention problem deals, like the substitute-pools problem, with the fungibility of money. The despot can promise to use the borrowed funds to build chicken farms, only to redirect the money later to his Swiss Bank account. Or, the despot can actually build the chicken farms, and soon after privatize them through sales to foreign firms, pocketing some of the cash revenues. Creditors, therefore, could claim to have lent funds for chicken farms while fully anticipating that the debtor would turn around and steal some of those funds.

If the creditors took steps to prevent this occurrence, they indeed should be shielded from liability. Recall: to be shielded, it is not required that the creditors warrant the use of the funds, only that they take due care. One measure that creditors can often take to prevent circumvention of the designed use by debtors is to make the purposes publicly known. This publicity would not guarantee that the money would reach its purposes, but it would make it more difficult for the despot to redirect this money. For the despot to stay in power, it is key to prevent the populace from recognizing harsh injustice, especially such that might lead them to coordinate and rebel. Keeping the populace uninformed about the true levels of despotism is crucial in preventing coordination. The despot who wants to steal, therefore, is going to prefer to steal from funds about which his populace is relatively uninformed. But if creditors, as part of their due-care efforts, have to ensure that the lending and its purposes and amounts are made public, and if they also have to ensure that the funds are at least routed through official channels, they will have done a significant amount to increase the costs of stealing the funds for the despot.



In another version of the scenario sketched at the outset of this article, a corrupt and despotic regime has just been overthrown by the citizens of a poor country. A big bank that loaned money to the despot (loans governed by New York law) is seeking repayment from the new democratic regime in federal district court in New York. The overthrown regime can produce evidence that a bribe (say, five percent of the loan amount) was paid by the lender to the despot to induce the loan. The temptation here is to say that the loan should not be collectible. After all, the creditor (typically rich) misbehaved. And the debtor (innocent and powerless at the time the loan was made) is poor. Along these lines, in a recent case decided by a tribunal of the International Center for Settlement of Investment Disputes, a Canadian businessman who had paid a $2-million "cash donation" to former Kenyan president Daniel Arap Moi to obtain a contract was denied any recovery on his claim. (61) Contemporary proponents of the odious debt doctrine applauded the complete denial of recovery as consistent with how a modem odious debt doctrine should apply. (62)

The problem with the foregoing example is that the perspective fails to consider ex ante effects. What if the remaining ninety-five percent of the proceeds had been used to fund a children's hospital? To penalize lenders to the entire extent of the loan, even when there is a significant benefit to the populace, gives the lenders little incentive to ensure that at least a portion of the loan goes to beneficial purposes. So, the proceeds of loans will either be stolen completely or they will not be made at all because of the heightened risks to lenders. To the extent the citizens of this poor nation need funds immediately to buy basic humanitarian goods, a doctrine that results in a complete denial of access to funding might not be optimal.

Against this full creditor-liability consequence, the liability framework developed in this article designates a more moderate sanction: forfeiting part of the debt. In this light, the scheme is more pro-creditor, assuring creditors greater recovery than the more extreme liability rule would afford. But in other contexts, courts continue to enforce full payment of sovereign debt. In many of these cases the proposed liability rule will have an anticreditor consequence, limiting the amount they can recover.

Whether the proposed liability scheme identifies the ideal way to arbitrate the comparative fault and benefit of the creditors and the populace might not be answered unless tested. It may well be that additional or alternative factors should be accounted for in dividing the liability. Still, the contribution of this article goes beyond the operative criteria it proposes. It lies in the recognition that an odious debt doctrine need not make all-or-nothing determinations. Some creditors can do more, and others can do less to monitor or reduce the theft of the funds. In the same way that bad creditor behavior can vary along a continuum, so too should liability.

(1.) See ERNST FEILCHENFELD, PUBLIC DEBTS AND STATE SUCCESSION (1931); M.H. Hoeflich, Through a Glass Darkly: Reflections Upon the History of International Public Debt in Connection With State Succession, 1982 U. ILL. L. REV. 39 (1982). For a contemporary overview, see Lee C. Buchheit et al., The Dilemma of Odious Debts, 56 DUKE L.J. 1201, 1208-24 (2007).

(2.) Although we have not found analysis of this comparative-fault perspective among contemporary commentators on the odious debt subject, it appears to have been considered over a century ago in the context of the repudiation of debts by the various American states. See WILLIAM A. SCOT]?, THE REPUDIATION OF STATE DEBTS (1902). The state debts that Scott was analyzing had often been incurred by corrupt legislatures and governors, sometimes in violation of the state's own laws. The question then was how to allocate liability between the "State and the innocent bondholders" once the corrupt legislature or governor was gone. Id. at 201. And part of the answer to the question was framed in negligence terms--that is, whether the bondholders had inadequately investigated the legal legitimacy of the bond issuance in question. Id. at 202.

(3.) Although not always articulated in these terms, the insight that creditors are often in a better position to prevent the harm in question appears to drive many of the calls for increased creditor liability for odious debts. See, e.g., Kunibert Rafter, Risks of Lending and Liability of Lenders, 21 ETHICS AND INT'L AFFAIRS 85 (2007) (making a similar point about the inefficiency of a system that does not require creditors in the sovereign-lending context to exercise due care); JOSEPH HANLON, OPEN UNIVERSITY, DEFINING ILLEGITIMATE DEBT AND LINKING ITS CANCELLATION TO ECONOMIC JUSTICE 11 (2002), available at DefiningIllegtimateDebt.pdf (discussing creditor responsibility on account of its negligence in the making of many of these loans).

(4.) On the relationship between odious debt liability and priority, see also Adam Feibelman, Contract, Priority and Odious Debt, 85 N.C.L. REV. 727 (2007) (suggesting that the issue of odious debt liability is best understood as a matter of inter-creditor priority); Buchheit et al, supra note 1, (suggesting the application of a version of Veil Piercing doctrine to the odious debt context and arguing that implicit in Veil Piercing is a notion of priorities based on relative levels of innocence of the various investors).

(5.) See generally Paul D. Carrington, Law and Transnational Corruption: The Need for Lincoln's Law Abroad, 70 LAW & CONTEMP. PROBS. 109 (Autumn 2007).

(6.) Among the articles in this symposium that emphasize the foregoing point are Albert H. Choi & Eric A. Posner, A Critique of the Odious Debt Doctrine, 70 LAW &. CONTEMP. PROBS. 33 (Summer 2007); Anna Gelpern, Odious, Not Debt, 70 LAW & CONTEMP. PROBS. 81 (Summer 2007); Tom Ginsburg & Thomas S. Ulen, Odious Debt, Odious Credit, Economic Development, and Democratization, 70 LAW. & CONTEMP. PROBS. 115 (Summer 2007).

(7.) The recent decision by the U.K. government to shut down its corruption probe of an arms deal between a U.K. company and the Saudi government in response to the threat by the Saudi government to scuttle a potential arms deal illustrates the general point being made in the text. See Saudi Defense Deal Probe Ditched, BBC NEWS, Dec. 15, 2006, available at business/6180945.stm (last visited Sept. 5, 2007), in which Lord Goldsmith, the Attorney General, justifies the dropping of the probe as a matter of balancing domestic public interests (presumably the arms contracts that would produce domestic jobs) versus the rule of law.

(8.) The intellectual history of this topic has been described in detail by a number of articles (including many in this symposium) and we will not reinvent the wheel by including citations for every historical fact. Interested readers who want citations for the factual assertions made in the subsection can look to Buchheit et al., supra note 1; Tai-Heng Cheng, Renegotiating the Odious Debt Doctrine, 70 LAW & CONTEMP. PROBS. 7 (Summer 2007); James Feinerman, Odious Debts, Old and New: The Legal Intellectual History of an Idea, 70 LAW & CONTEMP. PROBS. 193 (Autumn 2007); ASHFAQ KHALFAN, JEFF KING & BRYAN THOMAS, CTR. FOR INT'L SUSTAINABLE DEV. LAW, ADVANCING THE ODIOUS DEBT DOCTRINE (2003), Debt_Doctrine.pdf.

(9.) See Buchheit et al., supra note 1.

(10.) See Kris James Mitchener & Marc Weidenmier, SUPERSANCTIONS AND SOVEREIGN DEBT REPAYMENT (2005), (describing the importance of "gunboat diplomacy" as a sanctioning mechanism in the 1870-1913 period).

(11.) Although contemporary writers in the odious debt literature have not plumbed Sack's own motives, they are perhaps worth investigating given his contemporary revolutionary status among at least some of those who are urging the adoption of a Sackian doctrine. Examining Sack's motives might cast suspicion on the value of the version of the doctrine he proposed. Sack was a Russian emigre to France at the time that he articulated his odious debt doctrine; he had fled the communist regime, having been closely tied to the Tsarist regime. In France at the time, there was a financial crisis because the communist regime that took over Russia was refusing to pay the Tsar's debts. A large portion of the lending to the Tsar had come from the bond purchases of the French public. And the French public was incensed at the refusal of the communist regime to pay the Tsar's debts. Viewed in this light, one might ask whether Sack's articulation of a highly narrow odious debt doctrine (Sack's writings confirm that he was no friend of the communist overthrow of the Tsar) was done with the goal of ensuring that none of the Tsarist bonds held by Sack's French hosts could be deemed uncollectible under an odious debt doctrine. Cf. Gunter Frankenberg & Rolf Knieper, Legal Problems of the Overindebtedness of Developing Countries: The Current Relevance of the Doctrine of Odious Debts, 12 INT'L J. SOC. L. 415, 427 (1984) (noting that the French commentators of the 1920s, notably Jeze (1921) and Sack (1927), were at the extreme end of the spectrum in terms of their eagerness to protect creditor claims).

(12.) See Charley Cheney Hyde, The Negotiation of External Loans with Foreign Governments, 16 AM. J. INT'L L. 523, 531 (1922).

(13.) See Andy Metzger, A Mission Accomplished, LEGAL TIMES, Apr. 10, 2006, at 1.

(14.) For a fuller discussion of the odious debt doctrine and its modern reincarnation in the context of Iraq, see Jai Damle, The Odious Debt Doctrine After Iraq, 70 LAW & CONTEMP. PROBS. 139 (Autumn 2007). See also Anna Gelpern, What Iraq and Argentina Might Learn from Each Other, 6 CHI. J. INT'L L. 391 (2005).

(15.) See Kunibert Rafter, Odious, Illegitimate, Illegal, or Legal Debts--What Difference Does it Make for International Chapter 9 Arbitration?, 70 LAW & CONTEMP. PROBS. 221 (Autumn 2007) (discussing the Norwegian initiative).

(16.) See Lester Pimentel & Jeb Blount, Ecuador Calls Foreign Debt "Illegitimate,'" May Default on 60%, Jan. 19, 2007, BLOOMBERG NEWS (BLOOMBERG.COM), (last visited Sept. 5, 2007).

(17.) See Buchheit et al., supra note 1, at 55-56.

(18.) Seema Jayachandran & Michael Kremer, Odious Debt, 96 AM. ECON. REV. 82, 90-91 (2006); see also Patrick Bolton & David Skeel, Odious Debt or Odious Regimes?, 70 LAW & CONTEMP. PROBS. 83 (Autumn 2007) (proposing a workable definition of what an odious regime might be).

(19.) Jonathan Shafter, The Due Diligence Model: A New Approach to the Problem of Odious Debts, 21 ETHICS & INT'L AFF. 49, 50-51(2007).

(20.) The numerous institutional barriers to the setting up of an international regime to govern and police odious debts are discussed in Paul Stephan, The Institutionalist Implications of an Odious Debt Doctrine, 70 LAW & CONTEMP. PROBS. 213 (Summer 2007).

(21.) See Bolton & Skeel, supra note 18, at 98-99; see also Stephania Bonilla, A LAW AND ECONOMICS ANALYSIS OF ODIOUS DEBTS 22-29 (2006), abstract_id=946111.

(22.) For another proposal to base international trade liability on the "buyer in good faith" doctrine, see Leif Wenar, Property Rights and the Resource Curse (Apr. 5, 2007) (unpublished manuscript, on file with authors).

(23.) This appears to be the basic idea behind the Publish What You Pay (PWYP) initiative. Under the PWYP, a group of thirty NGOs have requested companies doing business in resource-rich developing countries to disclose publicly what they are paying for resources such as oil, gas, and minerals--the idea being that the sovereigns will be less likely to misbehave if they are more accountable. For details, see Welcome to the Publish What You Pay Website, (last visited Mar. 27, 2007). Along these same is the Extractive Industries Transparency Initiative (EITI) that aims to attack the "resource curse" by improving transparency and accountability. An EITI report is a publicly accessible external audit of extractive resource sales (for example sales of oil, copper, and diamonds) that the government in question implements so that the citizens can see how much the government has generated in income from the sale of the state's natural resources. For details, see EITI: About EITI, (last visited Sept. 5, 2007).

(24.) See, e.g., George G. Triantis & Ronald J. Daniels, The Role of Debt in Interactive Corporate Governance, 83 CAL. L. REV. 4, 1073, 1079, 1085-88 (1995).

(25.) That is, unless the lender is explicitly colluding with the despot to enable the stealing from the populace. But this is the extreme case in which the creditor's fault in the matter is readily apparent.

(26.) Actions that creditors could take in terms of contract-based governance are discussed in Feibelman, supra note 4. Feibelman differs from us in that he is optimistic that creditors are likely to, on their own or with the assistance of the official sector, make these contractual reforms to police odious debt even in the absence of a legal scheme that pushes them to coordinate such a reform. Id.

(27.) U.C.C. [section] 2-609 (2007).

(28.) See David Gray, Devilry, Complicity, and Greed: Transitional Justice and Odious Debt, 70 LAW & CONTEMP. PROBS. 137 (Summer 2007) (emphasizing that despots rarely, if ever, act in isolation; usually, at least a portion of the local populace and the international community supports them).

(29.) For such a proposal, see Wenar, supra note 22 (suggesting that loss should lie on buyers who paid the dictators for the right to extract natural resources).

(30.) In this sense, the liability regime proposed here differs from proposals to place liability on creditors and credit card issuers in domestic markets for what some commentators view as "reckless lending." See, e.g., John A.E. Pottow, Private Liability for Reckless Consumer Lending, 2007 U. ILL. L. REV. 405 (2007).

(31.) On the "crushing" effect of strict liability, see STEVEN SHAVELL, ECONOMIC ANALYSIS OF ACCIDENT LAW 108 (1987).

(32.) The possibility that this borrowing made it possible for the dictator to steal from other sources, such as tax revenues, is discussed below. See infra Part IV.A.

(33.) See Buchheit et al., supra note 1, at 16.

(34.) See id.

(35.) The modern literature on odious debt follows this all-or-nothing approach, with some taking the position that there should be no odious debt liability ever for creditors and others proposing full forgiveness under a certain set of conditions (or at least, burden shifting to the creditors to have to prove non-odiousness). More specifically, even if some of these commentators implicitly envision a partial-odiousness scheme, we are not aware of any discussions of it. For examples of contemporary commentators calling for the implementation of something akin to the Sackian all-or-nothing scheme, see PATRICIA ADAMS, ODIOUS DEBTS: LOOSE LENDING, CORRUPTION AND THE THIRD WORLD'S ENVIRONMENTAL LEGACY (1991); NOREENA HERTZ, THE DEBT THREAT: HOW DEBT IS DESTROYING THE THIRD WORLD 129-41 (2004). This all-or-nothing feature is also characteristic of agency law, which some modem commentators have proposed be used (at least by analogy) in the contemporary odious debt context. See Deborah DeMott, Agency by Analogy: A Comment on Odious Debt, 70 LAW & CONTEMP. PROBS. 157 (Autumn 2007) (suggesting that the all-or-nothing characteristic of agency law makes it an unattractive tool in many contexts).

(36.) This intercreditor perspective is discussed also in Feibelman, supra note 4.

(37.) 11 U.S.C. [section] 510(c) (1994).

(38.) The subject of why governments are reluctant to tax and instead prefer to borrow has, as one might expect, been much discussed in the tax literature. See, e.g., Neil Buchanan, Social Security, Generational Justice, and Long-Term Deficits, 58 TAX L.REV. 3, 275 (2005); Daniel Shaviro, Reckless Disregard: The Bush Administration's Policy of Cutting Taxes in the Face of an Enormous Fiscal Gap, 45 B.C. L. REV. 5, 1285 (2003-2004).

(39.) With Iraq, private credits made up approximately $20 billion of the $120 billion in unpaid debts that the Saddam regime had accumulated at the time of his overthrow. See Gelpern, supra note 6, at 89.

(40.) The populace may be relatively powerless under the despot, but the despot is always going to be worried that something will happen to solve the collective-action problem and produce an uprising. More sophisticated expositions of the basic idea that what despots fear is being overthrown look to the types of groups that are most likely to lead an uprising (for example, the military or some political party) and the question of how the despot either co-opts or suppresses them to stay in power. See, e.g., BUENO DE MESQUITA, ET AL., THE LOGIC OF POLITICAL SURVIVAL (2003).

(41.) In response to our draft, a leading expert in sovereign lending estimated for us that approximately eighty percent of what would qualify as odious lending was likely in the form of funding for military spending. E-mail from Lee C. Buchheit (Dec. 25, 2006) (on file with authors). There is, of course, a long history of scholarship pointing to the involvement of large banks in the funding of wars around the globe. The classic work on this topic is H.C. ENGELBRECHT & F.C. HANIGEN, MERCHANTS OF DEATH: A STUDY OF THE INTERNATIONAL ARMAMENT INDUSTRY (1934). See also T. Hunt Tooley, Merchants of Death Revisited: Armaments, Bankers and the First World War, 19 J. LIBERTARIAN STUD. 1, 37 (2005).

(42.) Military spending is potentially problematic for at least three reasons. First, the items are typically big-ticket matters (jet planes, warships, tanks) that have no clearly defined market price. The lack of a visible market price means that outside observers will find it hard to tell, for example, whether a warship really cost $800 million or just $600 million. Second, the government has the ready excuse of "national security" for keeping matters secret from both the creditors and the public. So, if the planes are second-rate and keep crashing into the ocean, the public might never know that the $80 million plane was really an $8 million bucket of bolts. Vis-a-vis the hypothetical bank issuing, for example a letter of credit, it is likely that the invoice for arms will not specify "poison gas," but will rather say something vague along the lines of "special equipment." Third, this is an item that can always be used to harm the populace. Chicken farms or children's hospitals, by contrast, do not present similar dangers.

(43.) For example, in the scheme set out by Buchheit, Gulati, and Thompson, lenders for military spending would probably find themselves immune because of the lack of clear evidence of wrongdoing. See Buchheit et al., supra note 1.

(44.) Eighty-five years ago, Charles Cheney Hyde made precisely this point in his admonition to lenders, explaining to lenders that allowing ambiguity in the use of proceeds was an invitation to later repudiation by debtors. See Hyde, supra note 12, at 531-32. Hyde was not saying that lenders should bear such a burden, rather that the climate of the time was such that a failure to exercise care might result in repudiation. He then goes on to lecture creditors on the prudent measures they should take when arranging the loans. Id. at 539-41.

(45.) Richard Craswell & John E. Calfee, Deterrence and Uncertain Legal Standards, 2 J.L. ECON. & ORG. 279 (1986); STEVEN SHAVELL, ECONOMIC ANALYSIS OF ACCIDENT LAW 79-83 (1987).

(46.) Intermediaries who are in the business of ascertaining compliance with contract standards exist in similar settings. For example, firms operating in today's global economy often have to outsource portions of their production into countries where traditional enforcement mechanisms such as courts are underdeveloped. These firms have to use alternative mechanisms to ensure that they receive high quality products and that their suppliers comply with matters such as labor practices and pollution. To

provide the necessary ongoing monitoring, an entire "assurance" industry has developed that is a combination of private assurance providers and NGOs. See Margaret Blair et al., Assurance Services as a Substitute for Law in Global Commerce 4-5 n. 12 (Vanderbilt Univ. Law School, Law and Econ. Working Paper No. 07-06, 2007),, (observing that one of the earliest uses of "assurance" services may have been in a sovereign context, where boards of state accountants verified state revenues and expenditures in ancient Athens 500 to 300 B.C. (citing George J. Costouros, Auditing the Athenian State of the Golden Age (500-300 B.C.), ACCT. HISTORIAN'S J. 41, 41-50 (Spring 1978)). What creditors would be looking for in our regime are the services of similar assurance firms--firms who would monitor the use of the money.

(47.) See Damle, supra note 14, at 154 (describing the recent efforts by the World Bank to police state corruption by restricting funds in circumstances in which corruption levels are deemed high).

(48.) See George Triantis & G. Mitu Gulati, Nineteenth Annual Corporate Law Symposium: Debt as a Lever of Control: The Economics of Creditor Control: Sovereign Versus Corporate Debt, 75 U. CIN. L REV. 977 (2007).

(49.) For the foregoing to happen effectively, though, the World Bank would have to work at engineering at least four actions: first, provide reasonably clear guidance on its policies regarding lending to governance-challenged countries; second, promise to be more transparent for its reasons for disapproval of a borrowing request; third, promote action by private-sector lenders (perhaps with the assistance of trade associations) to include in debt contracts (a) a condition precedent to drawdown that the country not have been refused a drawing under an MDB facility based on governance problems and possibly (b) an event of default triggered by an inability to draw from the MDBs for, say, a period of nine months as a result of governance issues; and fourth, the event-of-default point would be strengthened significantly if the Bank were to itself have a similar event of default in its own debt contracts. If the coverage were widespread enough in the market, sovereigns would have a huge incentive to clean up corruption. They could no longer play off the official and private lenders against each other, which is what happens now.

(50.) Kremer & Jayachandran, supra note 18, at 90.

(51.) The Oil-for-Food program illustrates the point. Misbehaving creditors cooperated with the rogue regime to fool others into believing that the resources Iraq was receiving were going only to further vital humanitarian goals. Now, in retrospect, we see that only a small fraction of those funds that were supposed to go for food and health actually went to those purposes. We are grateful to Jeffrey Meyer, one of the investigators of the Oil-for-Food program, for this point. For additional details, see JEFFREY A. MEYER & MARK G. CALIFANO, GOOD INTENTIONS CORRUPTED: THE OIL-FOR-FOOD SCANDAL AND THE THREAT TO THE U.N. (2006).

(52.) There is an argument that the current regimes in South Africa and Iraq could affirmatively sue some of the creditors and corporations that engaged in corrupt dealings that helped prop up the prior regimes in those countries. And yet, in neither of these cases have the current regimes been willing to bring these actions. And this is true even where there was concrete evidence produced by a neutral, external body that could have been the basis for suits against a number of foreign corporations--as in the case of Iraq as a result of the Oil-for-Food program investigations. See Mark Turner, et al., Iraq's Oil-for-Food Scandal Perpetrators Go Unpunished, FIN. TIMES, Dec. 8, 2006, at 6 available at 6300.

(53.) For an explication of this problem, see Carrington, supra note 5.

(54.) Id.

(55.) The idea that some bad creditors engage in actions that potentially reduce economic growth and the likelihood of the good creditors being repaid is one that has been in the news recently. In the United Kingdom, in a case involving the Republic of Zambia, a vulture fund allegedly bribed officials in the government to refuse a restructuring offer that would have reduced the nation's debt dramatically and instead allow the sale of the debt to a vulture fund that planned to sue Zambia for the full amount. See Alan Beattie, 'Vulture Fund' in Zambia Case Gain, FIN. TIMES, Feb. 16, 2007, at 6, available at e2340.html. Also illustrative is the ongoing litigation in New York involving the Republic of Congo, where a different vulture fund is alleging that the Congolese government has been colluding with certain other bad creditors to corruptly siphon out assets in a manner that has restricted payment to the good creditors. See John Donnelly, In Oil-Rich Nation, Charges of Skimming, BOSTON GLOBE, Nov. 25, 2005, at A1.

(56.) See Joseph Stiglitz, Odious Rulers, Odious Debts, ATLANTIC MONTHLY, Nov. 2003, at 39, 42, available at

(57.) See Stephen J. Choi & G. Mitu Gulati, Innovation in Boilerplate Contracts: An Empirical Examination of Sovereign Bonds, 53 EMORY. L.J. 929, 930 (2004); Anna Gelpern & Mitu Gulati, Public Symbol in Private Contract, WASH. U. L. REV. (forthcoming 2007), available at (manuscript at 18-25).

(58.) See Feibelman, supra note 4.

(59.) See Rafter, supra note 15 (citing Hanlon on this point). In the discussion that follows, we do not use the term laundering because to us that connotes illegal conduct under current domestic U.S. and other laws. The conduct being described, though, is not clearly illegal under current domestic laws.

(60.) Ecuador, in its current threat to repudiate certain of its debts, has targeted loans made to the military rulers from the 1970s. Those loans have subsequently been refinanced using new loans. Ecuador is asserting that all of the refinancings are tainted by the original bad purpose for the loans and, therefore, will not get repaid. For our purposes, the point is that creative refinancing does not appear to have worked with Ecuador--at least, not yet. See Pimentel & Blount, supra note 16.

(61.), News Articles--Kenya, odiousdebts/index.cfm?DSP=content&ContentID=16728 (last visited Feb. 15, 2007) (full opinion on file with authors).

(62.) See, News Articles--Kenya, (last visited Feb. 15, 2007). Note that the failure of the businessman to successfully argue a restitution claim may have in part been a function of the failure to adequately argue it. Our purpose in using the example, though, is to make a conceptual point.

Copyright [c] 2007 by Omri Ben-Shahar and Mitu Gulati.

This Article is also available at

OMRI BEN-SHAHAR, Omri Ben-Shahar is the Kirkland & Ellis Professor of Law and Economics at the University of Michigan.

MITU GULATI, Mitu Gulati is Professor of Law at Duke University. The authors acknowledge comments from Lee Buchheit, Stephen Choi, Adam Feibelman, Anna Gelpern, Caroline Gentile, Jeff King, Kimberly Krawiec, Sabine Michalowski, Eric Posner, John Pottow, Susan Rose-Ackerman, Kunibert Rafter, Robert Scott, Jonathan Shafter, Steve Shavell, Paul Stephan, and Tom Ulen. Thanks also to Joan Magat, Meredith Levy, and Jai Damle for editorial suggestions.
COPYRIGHT 2007 Duke University, School of Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2007 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Ben-Shahar, Omri; Gulati, G. Mitu
Publication:Law and Contemporary Problems
Date:Sep 22, 2007
Previous Article:Odious debt wears two faces: systemic illegitimacy, problems, and opportunities in traditional odious debt conceptions in globalized economic regimes.
Next Article:Odious debts or odious regimes?

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters