The 2008 financial crisis left no doubt that ill-considered debt can cause major damage not just to an individual country, but to the global economy.

You might think that by now, six years later, balance sheet repair would have taken debt below pre-crisis levels. However, debt burdens are substantially greater in almost all countries. McKinsey's latest analysis, Debt and (Not Much) Deleveraging, captures this reality.

Global debt has grown by $US57 trillion since 2007, raising the ratio to GDP from 269% to 286%. The ratio has increased in all advanced countries, although this average hides some big variations (such as Japan and Spain, with over 500% and 400% respectively).

Financial sector leverage has grown slowly and even contracted slightly in the crisis countries, simplifying the multiple layerings of debt that proved so fragile in 2008. Offsetting this improvement is the dramatic expansion of government debt, boosted by the 2008 financial bail-outs, fiscal stimulus in 2009 and the impact of the slow recovery on budgets. Reversing this rise in government debt would require Herculean budget austerity, amounting to 3-4% of GDP in Japan, Spain and Portugal, and not much less in France and the UK. Deleveraging would not only present a mighty political challenge, but would further dampen current feeble global economic growth.

Household debt has fallen in some of the financially troubled economies (US, UK, Ireland and Spain), in part as a result of default and rescheduling. Elsewhere, it has risen, including in Australia, which is high on global rankings.

All this looks pretty worrisome. But this is not the first time the panic button has been pressed on global debt, and last time it was a false alarm: Reinhart and Rogoff claimed that there was a critical cut-off point for sustainable government debt. It turned out their data didn't support the claim.

Any simple debt rule will mislead (this was Reinhart and Rogoff's main sin, not the careless use of data) but it's indisputable that debt is rising quickly and more debt creates vulnerabilities. But we need to go behind the aggregate figures to see why debt has risen and where the greatest dangers lie. 

More leverage was part-and-parcel of the process of financial deregulation which began in the 1980s, a desirable process that opened up the benefits of borrowing to a wider community. For example, American household debt has risen from 15% of income just after World War II to nearly 100% today because the financial sector now does a better job of meeting households' legitimate needs. But sub-prime lending to NINJAs ('no income, jobs or assets') was not a necessary part of financial development. The 2008 crisis was a failure of prudential supervision and a misplaced faith in the self-regulating capacity of financial markets.

Simple debt/income ratios are a misleading indicator of risk when borrowers have sound assets to match their debt. Governments which use their borrowing to fund useful infrastructure will be in a better position than those which borrowed to fund pensions and welfare (or, for that matter, to rescue failing banks). High leverage based on real estate collateral will be safe unless there is an unsustainable asset-price boom. When the debt belongs to high-income borrowers, high debt-servicing ratios are sustainable.

In short, inter-country comparisons are no more than a starting point in risk analysis.

Similarly, government debt has to be put in context. It's true that Japan would need to shift its budget dramatically towards surplus to get its stratospheric debt ratio down. But much of the debt is held by government institutions (including by the Bank of Japan) and most of the rest is held by stable domestic investors.

The McKinsey report provides a specific example of the benefit of case-by-case detailed study: China.

China refutes the idea that debt has to grow quickly in order to stimulate growth. In China's double-digit growth decades before 2007, debt grew slowly and remained tiny. It has accelerated sharply in the slower-growth period since 2007, so rapidly as to raise universal concern. Even though its total debt in not high as a percentage of GDP, China is probably headed for some uncomfortable financial fall-out: most countries have had some kind of financial crisis during the phase when the embryonic financial sector was growing fast to catch up to the real economy. 

The issue is not whether China has potential financial problems. It's whether it can sort them out without a significant crisis. On the positive side, the biggest banks are state-owned, making bail-out easier. There are no foreign-debt concerns. The central government is not heavily indebted, and could absorb significant  losses. There is still room for substantial urbanisation and upgrading of the housing stock as incomes grow. On the negative side, local governments seem difficult to discipline. At some stage, the grossly abnormal pace of housing construction will have to slow dramatically as the stock of housing comes to match the population's need and the catch-up phase thus draws to an end.

Above all, the McKinsey report reminds us how little we know about the links between debt and economic performance. Japan operates with an apparently unsustainable level of government debt. Denmark operates successfully with mortgage debt three times the global average. Singapore is almost at the top of the debt list but isn't a concern.

We don't have much of an analytical framework. We know that the 'zero net debt' view (additional expenditure by borrowers is cancelled out by lenders' reduced expenditure) doesn't capture the reality, but we can't tie down the relationship between debt and spending. We know that more debt makes a country more vulnerable to cyclical excesses and disruptive reassessments, just as more international trade makes a country more vulnerable to the vicissitudes of global trade. But being able to borrow and lend — shifting purchasing power from those with no immediate spending requirements to those with productive opportunities — ought to make the economy work better.

We don't know what a safe level of debt for governments or households might be, and if we did, we don't know how to enforce such limits while keeping the economy fully employed. Few of us saw the 2008 crisis coming. The one thing we know for sure is that we won't see the next one beforehand.

Photo courtesy of Flickr user epSos.de.