My post earlier today described new research from the European Central Bank which showed that GDP growth becomes seriously impaired when government debt-to-GDP reaches 90-100% and that some GDP components become affected at the 70-80% level. While this is useful, it is also worth considering some issues that can make comparing debt-to-GDP ratios across countries less reliable. These issues with the simple debt-to-GDP ratio include: off-balance sheet government liabilities; demographic projections; the holders of the debt; and the currency of the debt.
On the first, many governments have made future spending commitments that are not counted as part of the debt, ie social security and medicare in the US or public sector pensions in the UK. Spending on these is likely to increase substantially in the future and should be included (at their present value) within the ratio.
With regard to the second, the situation in the US (with a growing population) is much better than Japan (with a declining population) as tax revenues from more people will be available to service the debt.
On the third, debt held by citizens is considered less problematic than debt held by foreigners (on this measures the US fares worse than Japan) as interest and principal payments represent transfers abroad which are then much less likely to spent and recycled within the domestic economy.
Finally, the currency of the debt is also important. The US, for example, is able to create money to service its debt and therefore is not at risk of default, though this would lead to currency depreciation (a defacto default/write-off for foreign holders and something that would make US citizens relatively poor than those abroad). Greece on the other hand, cannot print its own currency (nor can many emerging market countries print the currency of their debt as most borrow in USD) and so is in a much more perilous situation.