VoxEU Column Development Financial Markets Global economy International Finance

Debt is not always odious: A theory of repression and growth traps

While many theories of international borrowing emphasize its advantages, it has proven difficult to empirically establish a correlation between a developing country’s use of foreign financing and good outcomes such as stronger growth. This column proposes a theoretical framework that reconciles the above puzzle. It establishes that a developing country’s propensity to save is essential in determining whether the government’s ability to borrow in international markets is welfare improving for its citizens or not. Hence, debt is not always 'odious' and alternative policies such as debt ceilings may prove more useful, especially in the midst of the current pandemic.

Many theories of international borrowing emphasize better risk sharing and the ability to draw on international savings to finance domestic growth as the two main advantages of international borrowing for a country (see for example Kletzer and Wright 2000). The first is achieved for example in case of an economic or natural calamity, when a country can then borrow to smooth consumption. Yet, it has proven difficult to empirically establish a positive correlation between a developing country’s use of foreign financing and good outcomes such as stronger economic growth (see Aizenman et al. 2004, Prasad et al 2006, and Gourinchas and Jeanne 2013). What are possible explanations for the divergence between theory and evidence?

One weakness of many existing models is the assumption that governments maximize the utility of their citizenry in the long run. However, in reality governments in many developing countries are often myopic, wasteful, and even rapacious. Whether poverty reduces the quality of government or the poor quality of government entrenches poverty is unclear.

A second weakness follows from the above assumption:  in such models, once the government is assumed to maximize the welfare of its citizenry, the optimal strategy is to default on its foreign debt (Bulow and Rogoff 1989a, Bulow and Rogoff 1989b, and Tomz 2012). To explain the existence of sovereign debt, researchers then have to appeal to a variety of mechanisms that enforce sovereign repayment such as a government’s concern for its reputation or punishment strategies by other countries. Unfortunately, there is little empirical evidence for these mechanisms (see Eichengreen 1987 and Arellano 2008  for examples).

Myopic and self-interested (“odious”) governments

In Acharya et al. 2020, we avoid these weaknesses by assuming that a country’s government is myopic (it is in power for only one period) and self-interested, hence all of its spending does little for the welfare of its citizenry. We then ask whether access to sovereign borrowing, taking into account the need for international investors to be confident that the borrowing will be repaid, may be welfare-improving for the country’s citizens.

We further assume that the government cannot selectively default on foreign investors only as it cannot distinguish between foreign and domestic holders. Hence, in case of a default domestic investors also experience significant losses.1 We assume that default costs are rising in the size of sovereign bonds held by domestic investors. The government in our model thus does not default on external debt for two reasons. First, it will incur the deadweight cost immediately. Second, its short horizon implies that it does not trade off the deadweight cost of default against the present value of the outstanding debt, but only against the net debt repayments it has to make during its period in power. This implies that a sizeable amount of debt issuance can be supported with modest deadweight costs.

The government raises funds from taxing the domestic real sector (e.g. manufacturing), and from borrowing (both domestically and internationally). The government’s ability to borrow alters the tax it will impose on the real sector. The higher the tax it imposes, the lower the amount that the private sector allocates to real investment, leaving more of its endowment to consumption and financial savings in government bonds. Hence, when a government has access to external sovereign borrowing, its taxation is driven by two sets of opposing concerns. A higher tax rate will curb real investment, resulting in lower future revenues to be taxed, lower surplus available to a future government to repay debt, thus reducing the future government’s ability to pay. But it will also raise the domestic private sector’s financial savings in government debt, increasing the willingness to pay of a future government. Depending on the constraint that limits the myopic, self-interested government’s ability to issue debt today, it may behave better or worse than a government that did not have the ability to borrow.

Why poor country governments with low savings behave worse

A literature on 'odious' debt takes the view that the ability to borrow internationally essentially gives the odious government more resources to waste or steal, with the repayment eventually being extracted by international lenders from the citizens (see Buchheit 2006 and Jayachandran and Kremer 2006). Therefore, some commentators advocate for declaring debt issued by such governments 'odious', and recommend limiting the enforcement of such debt in international courts. Other commentators (for example Choi and Posner 2007) have argued that the value of such proposals should depend on whether a country’s citizens will be better off when it no longer has the ability to borrow or not. The answer, our model suggests, is not straightforward.

For a country that starts out with low endowments (a developing country) but a high propensity to save among the citizenry, the government may have little need to channel more resources into financial savings, and it will lower tax rates relative to autarky in our model. In such a case, the government’s ability to issue debt tends to be beneficial for the citizen in the long run because the need to convince debt holders to repay limits the government’s rapacity and enhances steady-state consumption relative to autarky, i.e. there is a ‘growth boost’.

Conversely, for a country with low starting endowment and a low propensity to save among the citizenry, the government may set higher-than-autarky tax rates. This could push the country into a lower consumption ‘growth trap’, precisely because each rapacious government represses in order to enhance its debt issuance, leaving the next period government again with a low-endowment economy that is heavily indebted. In such a case, the repression gets entrenched ad infinitum. For the citizens of such countries, sovereign debt is truly odious.

To summarize, whether access to international borrowing improves or hurts citizen’s welfare in a developing country depends on its specific situation, not just on the odious nature of its government.  Debt is not always odious, even if the government is.

A potential resolution of the 'allocation puzzle' in international finance

A related question that our model may provide an answer to is the puzzlingly weak or even negative correlation between growth and use of foreign borrowing in developing countries (see Aizenman et al. 2004, Prasad et al. 2006, and Gourinchas and Jeanne 2013). Does foreign borrowing indeed damage growth prospects?

Our model suggests that this is not always the case.  Specifically, suppose the differential reliance on foreign borrowing across countries arises due to cross-country differences in the citizen’s propensity to save. Our model implies that the government of a country with a high domestic propensity to save will be more growth-friendly in its policies. Under certain conditions, this country will rely less on foreign borrowing than a similar country with a low domestic propensity to save. The citizens of the latter will experience more repressive policies from their government as it tries to boost its own borrowing.

In sum, the absence of a positive correlation between foreign borrowing and economic growth for developing countries documented in the literature may stem from the endogenous selection of which countries rely more on foreign borrowing, rather than some direct adverse effect of foreign borrowing on country growth and development.

Debt relief and debt ceiling

In our model, debt relief will do little for a country’s citizens, even when the country is in a growth trap. The government will simply use the expanded space to borrow, and spend it quickly. It will soon be back to pre-relief levels of debt – this was a common concern with the debt relief measures in developing countries in the late 1990s and early 2000s. Indeed, in a number of cases, debt relief has had little long-term beneficial effect as governments continue to misspend.2

In contrast to the ineffectiveness of debt relief our model predicts that debt ceilings that also bind future governments can be desirable. In particular, countries could decide to limit their own ability to borrow through a constitutional debt ceiling. Alternatively, well-intentioned international lenders could set informal debt limits for countries – though this requires a collective agreement as the country is capable of servicing more debt at the ceiling. We examine the consequences of such debt limits, and discuss why they might be more effective in developing countries than in rich countries, even if the quality of their governments is similar.

Dealing with sovereign defaults from large shocks

At the time of writing, most of the world still faces the Covid-19 pandemic, which leads to an arguably large endowment shock both in developing and developed countries. The pandemic clearly reduces production, taxes, future endowments, and the government’s ability to service debt, possibly pushing some developing countries into growth traps. Furthermore, the nature of the shock requires large government expenditure on healthcare as well as potential fiscal transfers to boost household endowments.  

Our model suggests that an efficient mechanism to help developing economies recover from such a shock could be ‘targeted relief’ -- a combination of (i) debt relief to avoid the default costs which can be a significant shock to government resources; and, (ii) continued access to debt markets, with the utilization of proceeds from debt issuance monitored (perhaps by a multilateral agency) for specific deployment toward containing the pandemic and its economic fallout.

More generally, a model of sovereign debt

Our model cannot only be applied to developing countries, however. Developed countries are simply countries with higher endowments and financial markets with higher costs of default. Our paper thus offers a tractable model of sovereign debt even in developed countries. As debt levels mount across the industrial world, it may prove relevant in a wider set of scenarios than the ones we have focused on.


Acharya, V V,  R G Rajan and J Shim (2020), “When is Debt Odious? A Theory of Repression and Growth Traps”, Working Paper, New York University School of Business.

Aizenman, J, B Pinto and A Radziwill (2004) “Sources for financing domestic capital – Is foreign saving a viable option for developing countries?”, National Bureau of Economic Research Working Paper 10624.

Arellano, A (2008), “Default risk and income fluctuations in emerging economies”, American Economic Review 98(3): 690–712.

Buchheit, L C, G M Gulati and R B Thompson (2006), “The dilemma of odious debts”, Duke Law Journal 56: 1201.

Bulow, J and K Rogoff (1989a), “A constant recontracting model of sovereign debt”, Journal of Political Economy 97(1): 155–178.

Bulow, J and K Rogoff (1989b), “Sovereign debt: Is to forgive to forget?”, American Economic Review 79(1): 43.

Choi, A H and E A Posner (2007), “A critique of the odious debt doctrine”, Law and Contemporary Problems 70(3): 33–51.

Eichengreen, B (1987), “Til debt do us part: the US capital market and foreign lending, 1920-1955”, NBER Working Paper.

Gourinchas, P-O and O Jeanne (2013), “Capital flows to developing countries: The allocation puzzle”, Review of Economic Studies 80(4): 1484–1515.

Jayachandran, S and M Kremer (2006), “Odious debt”, American Economic Review 96(1): 82–92.

Kletzer, K M and B D Wright (2000), “Sovereign debt as intertemporal barter”, American Economic Review 90(3): 621–639.

Prasad, E, R G Rajan, A Subramanian et al (2006), “Patterns of international capital flows and their implications for economic development”, in Proceedings of the 2006 Jackson Hole Symposium, Federal Reserve Bank of Kansas City, 119–158.

Tomz, M (2012), Reputation and international cooperation: sovereign debt across three centuries, Princeton, NJ: Princeton University Press.


1 For instance, if these are banks, the government will have to bail them out to have any reasonable economic output.

2 See here for a list of countries and the risk of debt distress prepared by the World Bank.

1,575 Reads