Ken Rogoff at Harvard and Carmen Reinhart at Maryland have written a series of papers (and a book) on the subject, using a comprehensive dataset of 70 countries over two centuries. They measure countries external and internal debt levels, and identify debt and banking crises, as well as episodes of hyperinflation and currency crashes. The sample is small (290 banking crises and 209 sovereign defaults in 14,700 country-years of data), and there are some issues surrounding the somewhat ad-hoc way that the authors classify episodes of hyperinflation and currency crashes. However, the fact that it encapsulates virtually all modern experiences with sovereign defaults makes it an interesting exercise, and one worth treating seriously as a guide to what we might expect.
The analysis yields several answers about sovereign debt, all of which generate concerns about the current state of the world economy. To begin with, prior to the onset of banking crises (such as the one that the US, Europe and the UK have just experienced), the external indebtedness of nations undergoes sudden surges. These surges are mostly in private debt (capital inflows, banking sector debt), and in short-term debt, although these liabilities seem to be converted into government liabilities pretty soon after the crisis. This happens for two main reasons: first, governments often bail out private participants, as we have seen in the most recent case. Second, tax receipts fall sharply in the aftermath of crises (more on this later). This finding on its own is not very worrying, although it does tell us that high debt levels in countries are perhaps a useful warning indicator for prospective banking crises.
More worrying are two related findings. The first is that banking crises tend to precede sovereign debt crises. This makes sense, given the earlier-mentioned tendency of government liabilities to rise on account of banking crises. It is also possible that this tendency arises because of currency depreciations that accompany banking crises, which in turn result in governments net foreign currency-denominated liabilities becoming larger. This suggests that the furore over the banking crises that we have just been through might be intensified if sovereign defaults materialise as they have in the past.
The second worrisome finding pertains to the nasty consequences of financial crises, namely housing price collapses, significant declines in GDP and high levels of unemployment. To be more specific, the numbers cited are average real house price declines of 35%, reductions in output from peak to trough of 9%, and rises in unemployment of around 7% over four years. These numbers are staggering by any measure. If we were to take these statistics at face value, we would have real reasons to be concerned about the likelihood of sovereign defaults and the economic destruction that would invariably ensue.
My reading of the situation, however, is that it is not necessarily all doom and gloom. The sample of defaults that we can study is relatively small, and times (and policies) have changed significantly over the last two centuries. We have more flexible monetary policy in many parts of the world and it has become a relatively powerful instrument of macroeconomic policy. This is also important in light of the fact that the findings that I discussed earlier are not conditioned on the specific policy responses that were implemented during any of the crises. It could very well be the case that conditional on appropriate policy responses, declines in output are much smaller, and unemployment doesnt increase quite as much as the unconditional numbers cited above. Finally, global policy coordination has definitely been better this time around than, for example, during the Great Depression when the imposition of the trade-crippling Smoot-Hawley tariffs by the US that exacerbated the length and depth of the prevailing crisis.
All of these mitigating factors lead to more benign predictions about the consequences of the current banking crisis. It is by no means certain that sovereign defaults will materialise, and even if they do, the resultant output reductions could be much less pernicious than we have seen in the past. The bottom line Dont panic yet, and hope for the best.
The author is a financial economist at Sad Business School, University of Oxford