Share This
June 2002 Cover Art
Search Finance & Development

Advanced Search
About F&D


Back Issues

Write Us

Copyright Information

Free Email Notification

Receive emails when we post new items of interest to you.

Subscribe or Modify your profile

Finance & Development
A quarterly magazine of the IMF
June 2002, Volume 39, Number 2

Odious Debt
Michael Kremer and Seema Jayachandran

Many developing countries are carrying debt incurred by rulers who borrowed without the people's consent and used the funds either to repress the people or for personal gain. A new approach is warranted to prevent dictators from running up debts, looting their countries, and passing on their debts to the population.

Under the law in many countries, individuals do not have to repay if others fraudulently borrow in their name, and corporations are not liable for contracts that their chief executive officers or other agents enter into without the authority to bind the corporations. The legal doctrine of odious debt makes an analogous argument that sovereign debt incurred without the consent of the people and not benefiting the people is odious and should not be transferable to a successor government, especially if creditors are aware of these facts in advance.

Box 1: Possible candidates for the odious label

The doctrine of odious debt originated in 1898 after the Spanish-American War. During peace negotiations, the United States argued that neither it nor Cuba should be held responsible for debt the colonial rulers had incurred without the consent of the Cuban people and not used for their benefit. Although Spain never accepted the validity of this argument, the United States implicitly prevailed, and Spain took responsibility for the Cuban debt under the Paris peace treaty. Soon after, legal scholars elaborated a similar doctrine. Since then, numerous regimes have evinced odiousness. Some potential recent examples include

  • Anastasio Somoza (Nicaragua)
    Reportedly looted $100-500 million
  • Ferdinand Marcos (Philippines)
    Amassed a $10 billion fortune
  • Jean-Claude Duvalier (Haiti)
    Successors claim he absconded with $900 million
  • Apartheid government (South Africa)
    Widely condemned by the international community
    Spent heavily on police and military to repress the African majority
  • Mobutu Sese Seko (Congo/former Za�re)
    Thought to have expropriated $4 billion to personal accounts
    Enriched cronies
  • Sani Abacha (Nigeria)
    Reportedly held $2 billion in Swiss bank accounts in 1999
  • Franjo Tudjman (Croatia)
    Looted unknown amount
    Suppressed the media and reportedly was behind violent attacks on his political opponents

However, this doctrine has gained little momentum within the international legal community, although many countries could qualify (see Box 1). For example, through the 1980s, South Africa's apartheid regime borrowed from private banks, devoting a large percentage of its budget to finance the military and police and otherwise repress the African majority. The South African people now bear the debts of their repressors. Despite appeals—from the Archbishop of Cape Town and South Africa's Truth and Reconciliation Commission—to have the odious apartheid-era debt written off, the postapartheid government has accepted responsibility for the debt. It seems to fear that defaulting would hurt its chances of attracting foreign investment and wants to be seen as playing by the rules of capitalism. South Africa is not poor enough to qualify for debt relief under the Heavily Indebted Poor Countries Initiative.

Similarly, Anastasio Somoza was reported to have looted $100-500 million from Nicaragua by the time he was overthrown in 1979. Sandinista leader Daniel Ortega told the United Nations General Assembly that his government would repudiate Somoza's debt, but reconsidered when his country's allies in Cuba advised him that doing so would unwisely alienate Nicaragua from Western capitalist countries.

Some countries have attempted to confiscate and restitute funds that an ex-ruler salted away abroad, but with mixed results. For example, Nigeria recently recouped money from Sani Abacha's family, but the Philippines has little to show for its protracted campaign to repatriate Ferdinand Marcos's fortune. Moreover, any money that has been squandered is gone forever.

What can be done to eliminate odious debt? In a recent study, we argued for the creation of an independent institution that could assess whether regimes are legitimate and declare any sovereign debt subsequently incurred by illegitimate ones odious and thus not the obligation of successor governments. If structured correctly, such an institution could restrict dictators' ability to loot, limit the debt burden of poor countries, reduce risk for banks, and lower interest rates for legitimate governments that borrow. This policy can be viewed as a form of economic sanction that no one would have an incentive to evade.

A new approach to odious debt

As it stands now, countries repay debt even if it is odious because, if they failed to do so, their assets abroad could be seized and their reputations would be tarnished, making it more difficult for them to borrow again or attract foreign investment. However, if there were an institution that assessed, and announced, whether regimes were odious, this could create a new equilibrium (that is, market outcome) in which countries' reputations would not be hurt by refusal to repay illegitimate debts, just as individuals' credit ratings are not hurt by refusal to pay debts that others fraudulently incur in their name. For example, if the world's leading powers, international organizations, and financial institutions declared a regime odious and announced that they would consider successor governments justified in repudiating any new loans the odious regime incurred, a private bank—even an unscrupulous one—would think twice before lending to the regime. This argument draws upon a well-known result in game theory that repeated games have many possible outcomes and that simply making some information public can create a new—and, in this case, better—one.

There is no guarantee, however, that everyone would coordinate on this new equilibrium. We propose two mechanisms to ensure that lending to odious regimes is eliminated. First, laws in creditor countries could be changed to disallow seizure of a country's assets for nonrepayment of odious debt. That is, odious debt contracts could be made legally unenforceable. Second, foreign aid to successor regimes could be made contingent on nonrepayment of odious debt. In other words, donors could refuse to aid a country that, in effect, was handing the funds over to banks with illegitimate claims. If foreign aid is valuable enough, successor governments will be compelled to repudiate odious loans, and banks to refrain from originating them. (Interestingly, the same reasoning suggests a potential way to solve the moral hazard problem associated with foreign aid (see Box 2).)

Box 2: A solution for moral hazard problems?

A similar policy approach might be useful for a very different purpose: limiting international bailouts of legitimate governments that borrow to finance economically disastrous policies. If the population of a country chooses such a government, some would argue that it is their prerogative, and it would be a breach of sovereignty to block the government's ability to borrow. However, many contend that the international financial institutions should not have to subsidize wasteful spending and that they sometimes do so in the form of international aid packages to countries whose economies have collapsed. This is the familiar moral hazard argument: the expectation of bailouts from international financial institutions encourages commercial banks and bondholders to make loans that governments cannot reasonably repay on their own.

The international financial institutions could discourage this type of opportunistic lending by private creditors in the following way. Unlike for cases of odious debt, a panel of independent jurists is not needed, and the institutions themselves could assess governments on purely economic grounds. They could announce, first, that a government is pursuing a policy that, in their view, is likely to prevent it from fulfilling its debt obligations. They could state that, once the country begins following sound policies again, they might provide aid, but not to help repay debt issued after the announcement. In particular, an international financial institution could, before granting assistance, stipulate that countries not simultaneously repay any loans taken after its announcement. This condition would eliminate any incentive for commercial banks to lend solely in anticipation of loans from an international financial institution. In this way, the institutions would avoid encouraging private lending to the country motivated by anticipation of a bailout. Unlike in cases of odious debt, loans would not be considered illegitimate and unenforceable. If creditors thought foreign aid would be unnecessary, they would continue to lend.

With this approach, the international financial institutions would be able to continue to give aid packages to countries that followed good policies but suffered bad luck; however, they would not bail out creditors who had opportunistically lent to countries that were following risky policies.

Better than trade sanctions

When the international community wants to take a position against a government that suppresses democracy and human rights—but does not want to resort to war—it sometimes imposes trade sanctions. Limiting an odious regime's ability to borrow can be considered a more attractive type of economic sanction. Like other sanctions, the threat of limits on borrowing could create incentives for regimes to reform. Governments wanting to retain the ability to borrow might loot less. Would-be dictators might even be discouraged from seeking power if sovereign borrowing were not one of the spoils of office.

Limiting borrowing, moreover, avoids two key shortcomings of trade sanctions. First, third parties have incentives to evade most trade sanctions. Curtailing odious debt, in contrast, is a self-enforcing sanction. The difference is that, as long as a few creditors and investors are willing to lend to and invest in a country that has refused to pay odious debt, banks know that if they disregard the sanction and issue odious debt, successor regimes will not have incentives to repay it. A bank might issue a short-term loan that it expects the odious regime to repay before it leaves office, but this amounts to a personal loan to the odious regime and does not hurt the people. If the odious regime repays it, there is no debt for the people to inherit; if the odious regime fails to repay it, the bank bears the downside since successor governments would not be responsible for it.

Another problem with trade sanctions is that they often harm the people they are intended to help. For example, if firms in a country are prevented from selling their products abroad, the loss of revenue may cause them to fire workers or lower wages. In contrast, curtailing dictators' ability to borrow, loot, and saddle the people with large debts would hurt illegitimate regimes but help their populations. The burden of repaying the debts would almost certainly outweigh any short-run benefit the population would obtain from proceeds of the loan that trickled down to them. (If a regime loots only a small amount and most of the proceeds flow to the people, the regime probably should not be considered odious.)

More countries engage in foreign trade than in sovereign borrowing, so limits on borrowing could be applied only in certain cases. Nonetheless, in these cases it could have a significant impact. For example, Franjo Tudjman of Croatia was arguably an odious ruler, having suppressed the media, instigated violence against political opponents, and looted public funds. In 1997, at the behest of the United States, Germany, and Britain—which were concerned about the "unsatisfactory state of democracy in Croatia"—the IMF cut off aid earmarked for Croatia. Still, commercial banks lent an additional $2 billion to the Croatian government between the IMF decision and Tudjman's death in December 1999. If the proposed institution had existed, creditors might not have granted Tudjman the subsequent loans, and the Croatian people would not bear the debt today. Such potential applications suggest that limits on borrowing should be part of the toolkit of policies available to the international community.

Inspiring truthfulness

But how does the international community ensure that an institution charged with assessing the legitimacy of debt would do so truthfully? An institution whose primary concern is the welfare of developing country populations might be tempted to declare legitimate debt odious so that a country would not have to repay it. However, if creditors anticipate being unable to collect on even legitimate loans, they will be wary of lending at all, and the debt market will shut down. This danger is a main reason the doctrine of odious debt has gained little support within the legal community.

To overcome this risk, the institution could be empowered to rule only on future loans to a government and not on existing debt. Then creditors will not face the uncertainty that loans they issue will be declared odious later. Moreover, the institution will be more likely to be truthful. Even if the institution is more concerned with the welfare of debtors than of creditors, it would have incentives to judge a regime honestly because honesty benefits the population. If the institution were to falsely call a legitimate government odious, it would deprive a country of profitable investments financed by loans. If it falsely called an odious government legitimate, the government would be able to borrow and then misuse the funds.

Restricting an institution to ruling on the legitimacy of loans before they are incurred would also limit the potential for favoritism toward creditors. If the institution favored creditors and ruled on existing debt, it might fail to declare some debts odious. But if it ruled only on future loans, even a small amount of concern for truthfulness or for the welfare of people in borrowing countries would prevent the institution from calling an odious government legitimate. This is because, before a loan is issued, the expected profits of a loan are very small for banks, as they have many alternative uses for their capital. Decisions on outstanding debt, in contrast, would be a "zero-sum game" (in which the gains to the winners equal the losses to the losers) between debtors and creditors, so a biased institution could help whichever party it favored. Because false rulings about future debt hurt the populations of borrowing countries and cannot substantially help creditors, an institution empowered only to block future lending would be unlikely to make biased judgments in order to help debtors or creditors.

There remains a possibility that an institution that rules on future debt may be biased for or against certain governments. If the major powers regard a country as an important trade partner or a strategic ally, the institution might fail to brand the government odious regardless of its misdeeds. Because such regimes with powerful friends can borrow at present, biased decisions in their favor would simply maintain the status quo. If, instead, the institution disliked a government for foreign policy reasons, even though the government had the consent of the people or spent for their benefit, the institution might falsely term it odious, thus cutting it off from lending. If this happened, citizens of the country would be worse off than under the status quo. The institution could be designed under a "do no harm" principle. Requiring unanimity or a two-thirds vote to declare a regime odious could safeguard against the possibility that a country would falsely be branded odious because of the biases of a few members of the institution.

How would the proposed institution affect an odious regime that inherited legitimate debt from the previous government? Even under the status quo, an odious regime would most likely prefer not to repay its creditors but keep the repayment money for itself. Extracting these resources from the regime would be difficult; the best one could do is prevent it from procuring more resources. Thus, the international community may as well allow an odious regime to roll over existing debt and interest payments (that is, defer repayment by renegotiating with bankers) while blocking it from borrowing further. Otherwise, an odious regime could default outright on legitimate loans, which would disrupt the international financial system.

Picking the judges

Another important question is which institution might judge odiousness. One option is a new international judicial body that hears cases brought against particular regimes. It could be composed of professional jurists representing several countries, similar to the International Court of Justice in the Hague or the proposed International Criminal Court. With independent jurists assessing odiousness, bilateral lenders and the international financial institutions might also be induced to lend only to legitimate regimes. The international financial institutions would not need to consider politics because, if a successor government were expected to repudiate the illegitimate debt, loans to an odious regime would be imprudent strictly on the grounds of financial risk.

If the major powers demanded a veto, another option would be for decisions to be made by the United Nations Security Council, where permanent members—China, France, Russia, the United States, and the United Kingdom—have veto power. A specific case might trigger the adoption of this policy; the next time the Security Council wishes to sanction a government, its approach could include blocking it from borrowing.

It may also be possible for major creditor countries to implement this system using domestic institutions only. Suppose that the United States changed its laws to prevent seizure of a foreign government's assets when it repudiated odious debt, that U.S. courts declared a regime odious, and that the United States announced it would oppose IMF or World Bank aid packages to a successor regime that repaid illegitimate debt. Suppose the European Union took similar steps. Then banks even in third countries would probably be reluctant to lend to that regime.

Another option would be for a nongovernmental organization (NGO) to identify which regimes are odious. If a panel of prominent, well-respected individuals including former heads of state of debtor countries (for example, Nelson Mandela), international lawyers, representatives of Transparency International, and human rights scholars promulgated a list of odious regimes, creditors might be reluctant to lend to governments on the list. Banks would face opprobrium in making such loans, and, in any case, successor regimes would be likely to refuse to repay. An interesting hybrid approach would be to combine announcements by an NGO with enforcement by domestic government institutions. For example, when U.S. courts hear cases brought by creditors seeking to seize assets of a successor government as repayment for debts incurred by a putatively odious predecessor government, the courts could take into consideration whether the predecessor regime had been on the NGO list when the loan was made.

In short, the international community or even a few major countries, possibly in concert with nongovernmental agencies, could create a new norm under which a country would not be responsible for odious debt. Creditors therefore would not issue odious debt in the first place. This new approach would be in line with the growing recognition in international law that some uses of power by government officials might be illegitimate or criminal, the prosecution of Slobodan Milosevic for war crimes being just one example of this trend.

The policy we have laid out would help legitimate creditors and debtors. Creditors would benefit from knowing the "rules of the game" in advance. Currently, there is a movement to nullify some debt on the grounds of odiousness, but it is hard for creditors to anticipate which loans will be considered odious in the future. If odiousness were declared in advance, banks would avoid lending to odious regimes in the first place and no longer face the risk of large losses if a successful campaign nullified their outstanding loans. Less uncertainty has the added benefit that interest rates for legitimate borrowers would be lower. But most important, dictators would no longer be able to borrow, use the proceeds for illegitimate purposes, and then saddle the people with their debts.

This article is based on a recent study by the authors, "Odious Debt,"presented at the IMF's February 2002 conference on macroeconomics and poverty. See

Michael Kremer is Professor of Economics at Harvard University and Senior Fellow at the Brookings Institution.

Seema Jayachandran is a graduate student in the Department of Economics at Harvard University.