VoxEU Column EU policies Europe's nations and regions Global crisis

External versus domestic debt in the euro crisis

As EU leaders muddle through the Eurozone crisis, the debate about its root causes continues. The debate is important if we are to understand how to prevent future crises. This column argues that the focus on total public debt is misleading – external debt is the key to the turmoil in European economies.

Does external debt matter in a monetary union? The case of Portugal illustrates the importance of foreign debt. The risk premium on Portuguese public debt rose continuously until the country was forced to turn to the European Financial Stability Facility (EFSF) for emergency financing. But its numbers on the fiscal side are no worse than those of France. Markets are worried about Portugal for a different reason – its high external debt – specifically, that of its private sector (banks and enterprises).

The fact that Belgium doesn’t have a crisis despite its poor fiscal position serves to illustrate the importance of external debt from the other side. Belgium’s debt-to-GDP ratio is well above that of Portugal (around 100%) and its political system is hopelessly divided. It has been over a year since the last election and the country still doesn’t have a new government. But despite these negatives, it faces a risk premium of only around 100 basis points over German debt. The reason, I assert, is that Belgium is a net creditor towards the rest of the world to the tune of 50% of its GDP, much more than even Germany.

Simple evidence

A standard story in the Eurozone crisis is that public debt is the key (Pagano 2010), and of course there is no denying that the fiscal side matters – indeed, the old joke is that the “IMF” in “IMF rescue packages” stands for “It’s Mostly Fiscal”. But the Eurozone experience seems to suggest that public debt alone is not sufficient as an explanation. Figure 1 shows a simple scatter plot of the risk premium on longer-term government bonds against the current balance (average over the last three years before the crisis). The strong, non-linear relationship between the spread and the current-account balance is obvious. The non-linearity of the relationship is natural given the positive feedback loop between higher debt and a higher risk premium which in turn makes debt service more expensive.

Figure 1.

A similarly strong relationship between risk premia and various indicators of foreign debt has been found in a number of other studies.1 The IMF (2010) finds, for example, that both the current account and cross-border bank liabilities are as important as predictors of CDS spreads as the fiscal deficit. Similarly Barrios et al. (2009) find an important role for the current account in determining risk premia.

Why the link between external imbalance and sovereign debt risk?

What is the reason for these empirical findings? To understand the link, consider first why the debts of most advanced nations outside the Eurozone are typically considered risk-free. Countries that have their own currency and issue government debt in that currency never have to outright default. If push comes to shove, they can always use the printing press to pay off the debt. Fear of this outcome most definitely affects interest rates, but the spread is best viewed as an inflation premium, not a default premium.

In a monetary union, the usual assumption that public debt is riskless is not valid. Countries like Greece don’t have access the ultimate option of printing money. In this sense, the public debt of Eurozone countries resembles that of emerging markets (Corsetti 2010).

The crux of the importance of external debt lies in the fact that even Eurozone nations retain full sovereignty over the taxation of their citizens. The logic is somewhat subtle and best explained by an extreme example that makes the point extremely clear. Suppose a nation’s entire debt is held by one man and the nation faces a debt crisis. If this bond holder is a resident of the nation, the government could impose a tax on him equal to, say, 50% of the value of his government bond holdings. Using this new tax revenue, the government could pay down its debt by 50%. Of course this would be an outrageous expropriation and make it harder to issue debt in the future, but it would not be a default.

By contrast, suppose the sole bond holder were a foreign citizen living abroad. In this case, the government could no longer freely tax the individual. Governments do not have a free hand in taxing non-citizens; they are bound by existing treaties and international norms.

The baseline point is that as long as Eurozone members retain full taxing powers, they can always service their domestic debts, even without access to the printing press. For example, governments could reduce the value of public debt held by residents by some form of lump-sum tax, such as a wealth tax. The government could just pass a law that forces every holder of a government bond to pay a tax equivalent to 50% of the face value of the bond.2 The value of public debt would thus be halved, much in the same way as it would be if the government ordered the central bank to double the money supply, which would presumably lead to a doubling of prices.

This is why, I believe, it is foreign debt that constitutes the underlying problem for the solvency of a sovereign, even in the Eurozone.3

Considering complexities

Things of course get more complicated when there are large gross, cross-border asset positions. It is thus possible that for a country without any net foreign debt, a large part of public debt is held by foreign residents (who want to diversify their holdings) but residents also have equally large foreign assets. But, even in this case, at least in principle, the government can still service its debt by taxing away the foreign assets of its citizens. At the very least the government can fund itself by forcing its residents to divest themselves of their foreign assets and buy (domestic) government bonds instead.

However, in this case the government faces the temptation to default on its foreign debt while its citizens can still enjoy the returns from their foreign assets. This temptation will be the stronger the more difficult it is for the government to tax the foreign assets of its residents.

The importance of this point was illustrated by the case of Argentina, where the country as such did not have a large net foreign debt. The private sector had large foreign assets while the government had about the same amount of foreign liabilities. However, Argentina went bankrupt with little net foreign debt because wealthy Argentines had spirited their assets out of the country and thus out of the reach of the government while poor Argentines refused to pay the taxes needed to satisfy the claims of the foreign creditors.

However, when the foreign assets of the country are held not by households but by institutions, such as pension funds, they can be taxed. While there might be limits to the extent to which a government can in reality tax the foreign assets of its citizens (depending in what form they are held) it is clear that the government of a country with a net positive asset position has more opportunities to extract resources for the service of public debt than a country with a large net negative asset position.

Another reason why the economic adjustment to domestic debt is much easier is that interest payments for domestic debt go to residents. Thus higher interest payments are – from the perspective of aggregate demand – like shifting money from the left hand to the right hand. The higher cost of servicing the debt does not alter the consumption possibilities of the country as a whole. Higher interest rates mean ”just” higher taxes for some and higher income for others, and the two groups might even partially overlap.

This is quite different when public debt is held by foreigners. In that case higher interest rates shift more money out of the country. This implies a higher transfer to foreigners and thus requires an adjustment in the trade balance.

Politics matters too

The political debt dynamics are quite different for domestic than for foreign debt. In the case of domestic debt there is a constituency that will vote for governments that want to avoid default. This is not the case for foreign debt; defaulting on ”foreigners” might actually be highly popular.

To see the point, take another simple example, this time of a nation’s citizens all belonging to the same family. The young generation works and pay taxes; the old generation lives off their savings. Assuming all government debt is domestic, the retirees will be relying on domestic government bonds for their retirement.

Think about voting on default in this situation. The old will clearly vote for tax increases (on the working young) to pay for debt service necessary to avoid default. The young will oppose. The outcome will depend upon empathy and electoral headcounts, but the main point is that close to half of voters will naturally be opposed to default. Indeed, in today's Europe the median age of the voting population is not far from the average effective retirement age. 

Now, re-run the thought-experiment with the retired bondholders as foreigners where the anti-default bloc cannot vote on the issue. If the country has a lot of foreign debt (as in the case of Greece and Portugal), it will be much more difficult to obtain popular consent for the austerity measures needed to transfer resources to foreigners.

There are thus both economic and political economy reasons why the amount of foreign debt of the country should be an important, maybe even dominant, determinant of risk premia on government debt.

Current accounts to the foreground

An instinctive reaction of many economists is that in a monetary union “regional” current-account deficits should be irrelevant. Indeed many would argue that one of the purposes of creating a monetary union was to transform national current-account deficits into the equivalent to regional deficits within a country and thus make them irrelevant. In a frictionless market this might indeed be the case. If residents of one member country want to consume or invest more, they just try to obtain financing on the capital market. As long as they can get this financing they can run a deficit.

Reality, however, is much more complicated.

  • In theory, the capital markets should lend only to creditworthy borrowers so that no debt overhang could ever arise.

The markets would simply stop lending when the debt of a borrower has reached the limit of his debt service capacity.

  • In reality, the capital market is subject to booms and busts.

During the boom credit is available cheaply and risk appears low.

The boom also creates its own fundamentals because the countries that receive large capital inflows record high growth rates and thus appear very good risks because a high growth rate indicates that the country should be able to service ever-increasing debt levels. However, when the bust hits, credit availability is restricted, sometimes even rationed, and the growth rate of countries that relied on capital inflows to sustain consumption and/or investment suddenly falls, thus reinforcing doubts about their creditworthiness.

In recent research with Cinzia Alcidi (Gros and Alcidi 2011), I provide some rough calculations of the magnitude of the turnaround in the “fundamentals” (the difference between the growth rate and the interest rate) between the boom (up to 2007) and the bust (2009/2010). This general phenomenon is well known in the literature on emerging markets and is called the ”sudden-stop” problem (Kopf 2011). ”Sudden stops” are very costly because they usually imply not only that the government cannot finance itself on the market, but also that banks and enterprises of the country concerned lose access to funding in credit markets.

The potential for ”sudden stops” implies that there exists an externality. During normal times, and even more during a credit boom, each individual national actor will try to obtain as much foreign credit as possible (given his own inter-temporal preferences and expected future income). However, the potential for a “sudden stop” and thus systemic crisis is related to the amount of debt the entire country has accumulated. This implies that governments in heavily indebted countries, or countries that are approaching the threshold beyond which a financial crisis becomes much more likely, should tax credit to domestic residents.

Implications for the Eurozone crisis

My argument that foreign debt is more important than public debt has a number of implications the ongoing Eurozone crisis:

  • Ireland, which is on course to run a current-account surplus this year and which has an overall small net international debtor position (about 20% of GDP) should be able to withstand the crisis much more easily than countries like Portugal or Greece which are still running sizeable current-account deficits and which have very large net international debtor positions (about 100% of GDP).
  • It might be more important for Greece (and Portugal) to achieve an external (current-account) surplus than a fiscal (primary) surplus.
  • In order to restore the creditworthiness of a country policymakers need to cut its foreign debt, not just public debt in general.

Given that by now only about one-half of all Greek government bonds are held by private foreign residents, it will be difficult to solve the problem by defaulting on government debt alone.


Barrios, Salvador, Per Iversen, Magdalena Lewandowska, Ralph Setzer (2009), “Determinants of intra-Eurozone bond spreads during the financial crisis”, European Economy, Economic Paper 388, November.

Corsetti, Giancarlo (2010), “The ‘original sin’ in the Eurozone”, VoxEU.org, 9 May.

De Grauwe, Paul (2011), “Governance of a Fragile Eurozone”, CEPS Working Documents, CEPS, Brussels.

Gros, Daniel and Cinzia Alcidi (2011), “Adjustment Difficulties and Debt Overhangs in the Eurozone Periphery”, CEPS Working Documents, Brussels.

IMF (2010), Global Financial Stability Report, April 2010, International Monetary Fund.

Kopf, Christian (2011), “Restoring financial stability in the Eurozone”, CEPS Policy Brief, CEPS, Brussels, March.

Pagano, Marco (2010), “Fiscal crisis, contagion, and the future of euro”, VoxEU.org, 15 May. 

1 Unfortunately different measures of foreign debt are often quite different. For example, for Ireland the sum of past current accounts suggests that the net debt of the country amounts only to about 20% of GDP. But the official data on the international investment position of the country (which is dominated by the foreign owned Irish subsidiaries of big multinational companies) shows a net negative position equivalent to almost 100 % of GDP.

2 The nature of the tax needed to pay off public debt might be different if public debt is due to banks because in this case the government would have to tax the holders of bank deposits.

3 The US is of course an exception since it enjoys the “exorbitant privilege” of having its foreign debt denominated in its own currency.


5,669 Reads