VoxEU Column Global crisis

No ordinary recession

This recession is different. Balance sheets of consumers, firms, and banks are under strain. The private sector is bent on reducing debt and this offsets Keynesian stimulus more than standard flow calculations would suggest. Bank deleveraging is by far the most dangerous. Fiscal stimulus will not have much effect as long as the financial system is deleveraging.

This is not an ordinary recession that differs from other recent episodes simply by being somewhat more severe. It differs in kind.

The end of the Cold War brought a decline in military spending and a recession which impinged most heavily on the states, like California, where the military-industrial complex was an important part of the local economy. The nationwide unemployment rate rose from 5.25% in 1989 to 7.5% in 1992. It then fell every year reaching just under 4% in 2000. The “free market” took care of the recession of the early 1990’s. Resources moved from the defence industries, trickling into other uses through innumerable channels. The federal government did not need to take a hand. Beginning in 1993, the federal deficit in fact shrank every year turning into a modest surplus in 1998. That was a very ordinary recession.

If the current situation were at all similar we would expect a recession in residential construction with unemployment among construction workers and mortgage brokers. Naturally, recent boom areas would be hard hit but we would expect resources gradually to trickle into alternative employment. Instead, we are threatened by a veritable disaster.

Balance sheet recessions

What is the difference? It resides in the state of balance sheets. The financial crisis has put much of the banking system on the edge – or beyond -- of insolvency. Large segments of the business sector are saddled with much short-term debt that is difficult or impossible to roll over in the current market. After years of near zero saving, American households are heavily indebted.

The holes that have opened up in the balance sheets of the private sector are very large and still growing. A recent estimate by Jan Hatzius and Andrew Tilton of Goldman Sachs totes up capital losses of $2.1 trillion; Nouriel Roubini thinks the total is likely to be $3 trillion. About half of these losses belong to financial institutions which means that more banks are insolvent – or nearly so – than has been publicly recognised so far.

So the private sector as a whole is bent on reducing debt. Businesses will use depreciation charges and sell off inventories to do so. Households are trying once more to save. Less investment and more saving spell declining incomes. The cash flows supporting the servicing of debts are dwindling. This is a destabilising process but one that works relatively slowly. The efforts by financial firms to deleverage are the more dangerous because they can trigger a rapid avalanche of defaults (Leijonhufvud 2009).

Three examples

Richard Koo (2003) coined the term “balance sheet recession” to characterise the endless travail of Japan following the collapse of its real estate and stock market bubbles in 1990. The Japanese government did not act to repair the balance sheets of the private sector following the crash. Instead, it chose a policy of keeping bank rate near zero so as to reduce deposit rates and let the banks earn their way back into solvency. At the same time it supported the real sector by repeated large doses of Keynesian deficit spending. It took a decade and a half for these policies to bring the Japanese economy back to reasonable health.

The US Great Depression saw no consistent policy of deficit spending on adequate scale in the 1930’s. War spending not only brought the economy back to full resource utilisation but also crowded out private consumption to a degree (Barro 2009).1 The deficits run during the war meant that:

  1. At war’s end, the federal government’s balance sheet showed a debt of a size never before seen, but also
  2. The balance sheets of the private sector were finally back in good shape.

At the time, a majority of forecasts predicted that the economy would slip back into depression once defence expenditures were terminated and the armed forces demobilised. The forecasts were wrong. This famous postwar “forecasting debacle” demonstrated how simple income-expenditure reasoning, ignoring the state of balance sheets, can lead one completely astray.

The lesson to be drawn from these two cases is that deficit spending will be absorbed into the financial sinkholes in private sector balance sheets and will not become effective until those holes have been filled. During the years that national income fails to respond, tax receipts will be lower so that the national debt is likely to end up larger than if the banking sector’s losses had been “nationalised” at the outset.

The Swedish policy following the 1992 crisis has been often referred to in recent months. Sweden acted quickly and decisively to close insolvent banks, and to quarantine their bad assets into a special fund.2 Eventually, all the assets, good and bad, ended up in the private banking sector again. The stockholders in the failed banks lost all their equity while the loss to taxpayers of the bad assets was minimal in the end. The operation was necessary to the recovery but what actually got the economy out of a very sharp and deep recession was the 25-30% devaluation of the krona which produced a long period of strong export-led growth. Needless to say, the US is in no position to emulate this aspect of the Swedish success story.

US predicaments

Strong contractionary forces are at work in the US emanating both from the capital and the income accounts. Stabilisation requires major policy actions on both fronts.

  • First, the financial system must be recapitalised so as to remove the relentless pressure to deleverage from the banks.
  • Second, a spending stimulus sufficient to reverse the rapidly worsening decline in incomes must be administered.

When the entire private sector is bent on shortening its balance sheet and paying down debt, the public sector’s balance sheet must move in the opposite, offsetting direction. When the entire private sector is striving to save, the government must dis-save. The political obstacles to doing these things on a sufficient scale are formidable.

If banking system losses are of the magnitude estimated by Goldman Sachs or Roubini, the banks need capital injections of at least another $200-300 billion. Even if injections equal to all their losses could be effected, the banks might still want to contract, now that they know how dangerous their leverage of yesteryear was.

The American public understands clearly that the present disaster was fashioned on Wall Street (albeit with some stimulus from Fed policy). Outright bail-outs are a “hard sell” therefore. But the American ideological taboo against “nationalisation” also stands in the way of dealing with the matter in the straightforward way that Sweden did. The present administration, like the last, would like to recapitalise the banks at least partly by attracting private capital. That can hardly be accomplished as long as the value of large chunks of the banks’ assets remains anybody’s guess. Government guarantees against (some part of) losses that may be incurred might solve this problem. But it would be a strangely contrived way out of a political impasse.

Fiscal stimulus will not have much effect as long as the financial system is deleveraging. Even if that problem were to be more or less solved, the government deficit would have to offset both the decline in industry investment and the rise in household saving – a gap that is rising as the recession deepens. Here, too, the public is sceptical and prone to conclude that a program that only slows or stops the decline but fails to “jump start” the economy must have been a waste of tax payers’ money. The most effective composition of such a program is also a problem.

Almost all American states now suffer under self-imposed constitutional balanced budget requirements and are consequently acting as powerful amplifiers of recession with respect to both income and employment. The states will spend everything they get, as will a great many local governments. The point of the stimulus package is to increase spending. Income maintenance for unemployed and other low income households will also be effective.3 Tax cuts will produce considerably less spending per dollar than these other programmes. However, the political prospects seem to portend a less than ideal program mix.

Perils, present and future

If government programs end up not being large enough to turn the recession around, we have to look forward to a deflationary period of indeterminate length. If they do succeed, however, severe inflationary pressures may surface quite quickly.

The US ratio of federal debt to GNP is not particularly high at this time. But it does not take into account the very large off-balance liabilities of entitlement programs. Since the present crisis began, moreover, the Federal Reserve System and other federal agencies have made bail-out, loan and credit guarantee commitments totalling many trillions of dollars with uncertain eventual implications for the consolidated federal balance sheet.

Much will depend on the willingness of the nation’s foreign creditors to continue to accumulate or at least to hold dollars at low rates of interest. Should this willingness falter, inflation will be hard to contain.

There is much to fear beyond fear itself.


Editors' note: This column is a Lead Commentary on Vox's Global Crisis Debate where you can find further discussion, and where professional economists are welcome to contribute their own Commentaries on this and other crisis-linked topics.



Robert Barro, 2009, “Government Spending is no Free Lunch” (Wall Street Journal, January 22).

Leijonhufvud, 2009 “Two Systemic Problems,” CEPR Policy Insight No. 29, January

Richard C. Koo, 2003, Balance Sheet Recession: Japan’s Struggle with Unchartered Economics and its Global Implications, Singapore: Wiley


1 The crowding out at full employment, Barro thinks, “most macroeconomists would regard … as a fair [test case] for seeing whether a large multiplier ever exists” (italics added). Most macroeconomists will presumably agree that World War II was not a free lunch but are not likely to agree to Barro’s test case inference.

2 There were plenty of bad loans but at that time they did not have the non-transparent “toxicity” that “sliced and diced” CDO’s were to generate in the present crisis.

3 Milton Friedman’s negative income tax proposal seems to have become anathema to conservatives by now. It would work in our present situation.

945 Reads