“You can’t cut debt by borrowing.” How often have you read or heard this comment from “austerians” (a nice variant on “Austrians”), who complain about the huge fiscal deficits that have followed the financial crisis?

The obvious response is: so what? Shifting debt from people who cannot support it to those who can – the population at large, both now and in future – seems to make a great deal of sense if the alternative is an economic collapse that leads to a loss of output and investment now and so of income in the long term. Indeed, under the latter alternative, even the fiscal deficits may end up little, if any, smaller if one tries to slash them, as the UK could be about to discover.

Before leaping to that conclusion, however, let us approach the issue of de-leveraging – or debt reduction – analytically. Between 1994 and 2007, total US non-financial private debt rose from 118 per cent of gross domestic product to 173 per cent, the highest level in US history. Over the same period, US financial sector debt rose from 54 per cent of GDP to 115 per cent. A great deal of this leveraging up of the economy (matched elsewhere, notably in the UK) was based on false premises: borrowers and lenders thought that the assets against which they had borrowed would be worth more than turned out to be the case.

How, then, can people reduce their indebtedness or restore their net worth, after an unforeseen fall in asset prices? There are three mechanisms: sale; bankruptcy; and frugality. Let us consider each of these, in turn. But remember that, at the global level, debt cancels out: net debt is zero. So, in paying down debt, one is also reducing credit by an equal amount.

People with assets that they no longer wish to hold and debts they no longer wish to bear, can sell the former to repay the latter. If this is to cancel debt, then the ultimate purchaser needs to be a creditor. Sale makes this a voluntary transaction.

This path to de-leveraging is going to be part of the story. But when the predominant asset is housing, as it is now, the willingness of creditors to purchase will be limited. By and large, people who wish to buy houses are young and have limited liquid assets. Most creditors already own houses. In theory, houses could be sold to cash-rich foreigners. But that, too, is going to be a limited avenue for economy-wide deleveraging in most countries. (In Spain, however, sale to cash-rich foreigners seems a more plausible solution, since much of the past construction was designed for their use.)

The second approach is mass bankruptcy. In this case, creditors are forced to write down their loans to the value of the asset. That is clearly an important part of any de-leveraging. But since highly leveraged financial intermediaries stand between the ultimate creditors (households) and the ultimate debtors (other households), mass bankruptcy is going to wipe out the capital of intermediaries. That is likely to trigger panic, as losses cascade across the financial system.

Organising such a bankruptcy procedure, to allow for a mass adjustment of claims, is indeed one of the necessary conditions for managing a financial crisis efficiently. But it is going to be politically and technically complicated. In the end, however, a substantial part of the debt and the corresponding credit should be eliminated in this way. The big policy decision is how far the state wishes to socialise the losses of creditors. The answer will certainly include some socialisation, since governments insure deposits in financial institutions.

The third approach is repayment. Under any imaginable resolution of the debt overhang, some people are going to seek to pay down their loans. Indeed, a great many are going to try to do so: those who dislike the idea of bankruptcy, including the stigma; and those whose assets are worth not much less than their loans. To these groups of higher savers should be added those who are simply poorer than they thought they would be and so decide to save more.

While the highly indebted and the newly “asset-poor” have good reason to spend less than before the crisis, creditor households have no reason to spend more. Indeed, the collapse in interest rates in a slump lowers their incomes and so is quite likely to make them want to cut back on their spending, too. The aggregate effect of these changes in behaviour is, of course, a rise in the desired household rate and so the desired financial surplus of the household sector.

It is a matter of simple logic, that, since the financial balances of the household, corporate, government and foreign sectors must sum to zero, a rise in the surplus of the household sector must be offset by an offsetting move in other sectors.

During a post-crisis recession, the surplus of the corporate sector always rises, as it has done now, because managements slash investment. In the current crisis, increases in the surpluses of the non-financial corporate sector in high-income countries have been particularly large. In fact, non-financial corporate sectors were running substantial financial surpluses in the high-income countries before the crisis and are running still bigger surpluses now.

A shift in the foreign surplus means a shift towards surplus in the current account of the debt-burdened economy. That takes time. It also requires changes in the balance between saving and investment in the rest of the world. In practice, surplus countries do not want to make the adjustments needed to allow the US, UK and other former deficit countries run huge current account surpluses at full employment levels of income. So this way out is also largely blocked, alas.

When one has eliminated everything else, it turns out that the only sector both able and likely to offset a large move of the household sector towards financial surplus in a post-crisis slump is the government. Indeed, that is exactly what has happened.

My conclusion, then, is the exact opposite of the conventional wisdom with which I began: the only way that the private sector can de-leverage, when large economies are in a post-crisis recession, is for the government to leverage. The economy, as a whole, cannot de-leverage in any other way, other than via accelerated mass bankruptcy, which would certainly deepen the recession, if not create a depression. If the government tried to eliminate its deficit over night, it would have to drive the private sector back towards balance (or achieved a massive shift in the external balance very swiftly). In the context of excessive debt, that is only going to happen if private sector incomes are so squeezed that paying down their debt is no longer feasible. But in this situation, mass bankruptcy and a slump again becomes a likely outcome.

The latter is, indeed, what now threatens peripheral European countries forced to reduce fiscal deficits at exactly the time when their households are trying to pay down their debts and corporations are slashing investment. I fear the outcome of this hair-shirt policy, which is likely to break the will of some countries and, quite possibly, the eurozone itself.

So the least bad way to deal with a huge debt overhang has three elements: facilitate mass bankruptcy of the hopelessly over-indebted; lower interest rates, so making it easier for the indebted to carry and pay down their debt; and accept large fiscal deficits as a way of sustaining the incomes of those trying to pay down debts. The recommended alternative of slashing the fiscal deficit while the private sector tries to slash its debt suffers from a fallacy of composition: it is impossible for all sectors of the economy to spend less than income at the same time.

Of course, as this process proceeds, private debt should fall and public debt rise, relative to GDP. Is this a big problem? In some countries, the answer will be: yes. These countries will have to go through massive debt restructuring in the private and, quite possibly, public sectors. But other countries, notably the US, are perfectly able to run large fiscal deficits, financed, if necessary, by the central bank. At the end of this multi-year process of private sector debt restructuring and repayment, the private sector will be in balance once more and able and willing to spend. Meanwhile, the higher level of debt can be carried quite easily. So long as the real interest rate on government borrowing is not much above the real growth rate, stabilising the level of public debt to GDP does not even require a primary fiscal surplus.

Now, assume, that in this newly restored economy, the fiscal deficit is largely eliminated. Then, over time, the ratio of public debt to GDP can be brought down through the normal process of economic growth. Making structural changes in fiscal policy that control spending in the long run makes this more credible.

In short, not only can we deal with the private sector debt overhang by increasing the fiscal deficit, but we must do so. It is the only way of avoiding a deep slump and the immense disruption of mass bankruptcy. But this is not to preclude debt restructuring, as well. It is important to develop ways to restructure private debt, too. But, for this to happen, we must be prepared to impose more losses on financial intermediaries and so on their creditors.

Analysis of the economy is not the same thing as analysing a single household. What is true of the latter is not true of the former. The unwillingness to recognise this truth will lead to serious policy mistakes.