We say "the debt of Greece" (or of "Spain", "France" etc.) while we talk about the debt of the Greek State (or of the Spanish, the French State). But the debt of a State and the debt of a country are two different things.
Moreover, we evaluate the level of indebtedness of a State by the ratio "gross debt / GDP", selected in the Maastricht agreements. This ratio is a chimera, a monstrous concept, because it compares a stock of an actor (the level of the gross debt of a State) to a flow of another actor (the annual output of a country, measured by its GDP). Yet we learned in elementary school that all proportion must include things of the same nature ("do not divide leeks by turnips," said our teachers). It is surprising to see so many economists discourse pedantically on such a misleading indicator.
The "gross debt" of an economic actor is less significant than its net debt, the gap between the value of its debts and the claims he has over others.
Parenthetically, there is nothing wrong with debt: any productive actor (the government itself produces services), whose work involves anticipation and investment, borrows and holds claims on other actors. The theory defines the optimal level of debt but, reality being never what the theory assumes, the most important criterion is credibility: an actor is credible if its creditors, confident in his ability to repay, are ready to renew their loans without requiring a higher risk premium.
The ratio that assesses the credibility of an actor is "net debt / value of assets" because one who possesses an important asset may, if necessary, sell it in order to repay. If we are careful we calculate this ratio by weighting debts according to their due, claims and assets according to their liquidity.
The national accounts do not publish such a ratio: one would have to delve into the accounts to find the data required, and neither the citizen nor even the economist possess the technical know-how - furthermore some essential data are missing: how to evaluate this component of the assets of a State that is his ability to raise new taxes?
What matters however is less the credibility of a State than the credibility of the country which, besides the State, includes businesses and households: a State may, in fact, when compelled, drain the other actors of the country – politically painful and difficult, but possible operation.
In the net debt of a country, internal debts are canceled since a debt of an actor is a debt of another: one has to consider only the debts and claims with other countries. Then the appropriate ratio to assess the credibility of a country is "net debt to foreign countries, all players combined / assets of the country," applying possibly the same rules of caution as above.
National accounts provide an indirect measure of the net debt: the current account, balance of interest paid and received by the country, is roughly proportional to the net debt. As it is not easy to assess the value of the assets of a country, this value can be assumed proportional to GDP: it is equivalent to assuming (another rough approximation!) that the value of the annual output of a country is a constant fraction of the value of its capital.
The ratio "current account / GDP", even if rough, is more consistent than the chimera "gross debt of the State / GDP" because it divides a stream with another stream and these two streams are related to the same actor.
In December 2009 the OECD predicted for 2010 a current account surplus of 5.4% of GDP for China, 4.5% for Germany, 2.8% for Japan and a deficit of 2.1% for France, 2.3% for Italy, 2.4% for the United Kingdom, 3.4% for the United States and Canada. In light of this indicator, who is "strong" and who is "weak"? (see Christian Sautter, « Oh ! La vilaine dette publique ! » , Lettre à nos amis, n° 407, March 5, 2010).
One found higher ratios when considering smaller countries: surplus of 18.6% for Norway thanks to gas, 10.2% for Switzerland, 8.2% for Sweden and the Netherlands, deficit of 10% in Greece (before his austerity plan), 11% for Portugal.
In Japan, the ratio "gross debt of the State / GDP" is 225%: this is far superior to the 80% of France and to the 60% of the theoretical maximum in the Eurozone, but this is not dangerous because 94% of Japanese government debt is financed by Japanese creditors who are faithful and insensitive to the ratings given by the rating agencies – in fact, as we have seen, the current account of Japan is positive.
To each good indicator one can assign a limit beyond which a warning signal lights. The example of Japan shows that it is impossible to define such a limit for the ratio "gross debt of the State / GDP": hence this chimera adds no information and has no practical meaning. But it has consequences...