Lawrence H. Summers

Charles Eliot University Professor, Harvard University

December 19, 2007

Risks of Recession, Prospects for Policy

I am speaking here today because I believe that our current economic situation requires a comprehensive program of measures to contain the fallout from problems in the financial and housing sectors and to assure sufficient policy support for economic growth over the next several years. Perhaps because of a failure to appreciate the gravity of our current situation and the problems our political process has in responding quickly and collaboratively to emergent threats, such a comprehensive program is neither in place nor in immediate prospect.

No economic projection put forward with anything like complete confidence should ever be trusted. The current consensus suggesting that growth is likely to be slow over the next several quarters and that the odds of a technically defined recession are in the 40% range is troubling enough given that it means rising unemployment and budget deficits, likely falls in real family incomes and a downturn in plant and equipment spending.

For the last year, the economic consensus, and the policy actions that have flowed from it, has been consistently behind the curve in recognizing the gravity of the problems in the housing and financial sectors and their consequences for the overall economy. This continues to be the case. In my view it is almost certain that we are headed for a period of heavily constrained growth, quite likely that the economy will experience a recession as technically defined and distinctly possible that we are headed into a period of the worst economic performance since the stagflation of the late 1970s and recessions of the early 1980s.

The late Rudi Dornbusch was fond of remarking that in economics “things take longer to happen than you think they will and then they happen faster than you thought they could.” So it has been recently. The related but distinct patterns of excessive valuations in housing markets and excessive complacency in credit markets were pointed out for years by experienced observers. The cracks took longer to appear than many expected and have now proven to be far more structurally damaging than almost anyone supposed.

Economic downturns historically come in two categories. For most of the post war period, economic expansions did not die of old age. They were murdered by the Federal Reserve in the name of fighting inflation. This was the story in 1958, 1971, 1974 and 1982 as sharp increases in credit costs drove the economy into downturns.

Before World War II, and in recent years as inflation has come under control, expansions have ended as a consequence of the workings of the financial system, sometimes in conjunction with oil shocks. After a period of optimism, asset prices expand beyond fundamental values, credit expands, investors embrace financial innovations that allow greater leverage so as to better take advantage of rising asset values. At some point the party ends, asset prices fall, financial structures that once looked impregnable become vulnerable, confidence collapses, propensities to consume and invest fall off, and the economy turns down.

Experience suggests that downturns driven by falling asset prices and credit problems tend to be recognized relatively slowly and to be quite protracted. Two extreme examples are the American experience after 1929 and Japan’s experience in the 1990s after the 1989 asset price collapse. Our last two recessions associated respectively with the bursting Savings & Loan real estate bubble and the NASDAQ collapse revealed gaps of several years between asset price peaks and the restoration of satisfactory rates of economic growth. Nationally housing prices peaked less than a year ago, and credit spreads reached their minimum levels only about six months ago.

History’s caution that situations like our current one are likely to surprise on the downside for a considerable time and prove quite protracted is confirmed by forward looking indications regarding the economy.

• 300,000 home foreclosures were initiated in the first half of last year. The vast majority of them involved mortgages that had not yet reset. Even with recent policy changes up to 1 million foreclosures are expected over the next two years.
• The new and relatively crude futures markets that exist are predicting that peak to trough national housing prices will fall by 24% according to an index that has only declined 6.6% from its peak so far. Already prime mortgages are defaulting at the same rate sub-prime mortgages defaulted 3 years ago.
• Freely traded shares in Real Estate Investment Trusts (REITs) are suggesting that the value of commercial real estate if marked properly to market may be down by as much as 20% and the rate of transactions in commercial real estate has declined by more than half over the last year.
• The most important driver of U.S. economic growth over the past seven years has been consumption which has outstripped GDP growth. The combination of a near zero personal saving rate, lost housing wealth, reduced availability of credit, reduced real incomes caused by rising oil prices, a falling dollar and rising food prices and increased uncertainty constitute a perfect storm depressing consumer spending.
• Even looking out five years the spread between safe liquid Treasury borrowing rates and the rates at which major financial institutions borrow is at well above normal levels. The debt of some of our countries largest and most prominent financial institutions is trading at levels suggesting a market judgment that their odds of defaulting on their debts over the next five years approach one in ten.

Of course it is possible that improved trade performance coming from the falling dollar, the working through of the Fed’s monetary policy actions and typical American resilience will carry us through the next year robustly. But this is not where the preponderant probability lies.

Economic policy making is about balancing risks. I have already suggested that the probability of subpar growth exceeds the chance that growth will be robust. There is an additional crucial point as well. The adverse consequences of policy choices that fail to deal with a potential recession and fail to stimulate the economy and that do not allow for financial repair far exceed the adverse consequences of over-insuring against an economic slowdown.

Consider the costs if we experience even in a relatively mild recession:

• Losses of close to $5, 000 in income for the average family of four quite heavily concentrated among the disadvantaged who are inevitably last hired and first fired along with cutbacks in Medicaid, child welfare and other social safety net programs as state budgets contract.
• A several hundred billion dollar increase in our national debt and a significant reduction and a substantial cutback in investment in plant and equipment, education and R&D
• Hundreds of thousands more foreclosures and greatly increased risks to the financial system.

• Greatly complicated international relations as the our downturn slows the rest of the world economy, the American economic model is called into question, protectionist pressures rise, and the dollar’s centrality to the international financial system is called into question

Of course if a downturn turns into more than mild recession, the risks are that much greater.

Against these risks, what do those who counsel against what they see as imprudent activism worry about? They fear that stating the need for strong action will somehow undermine confidence by laying the problem bare. And they worry that inflation might tick upwards or that those who have made financial errors will be insufficiently punished. .

I only hope that history will see these as the main economic problems faced by whoever is elected President of the United States in 2008.

It is the great irony of financial crisis that the very measures that could have prevented crisis are counterproductive in a time of crisis. Of course it would have been better to have had more fear on the part of lenders, less rampant liquidity, and higher saving two years ago when imbalances were building. But that is not what we need now.

The most urgent priority for policy over the next several months is containing the incipient economic downturn. I am convinced this is possible without giving rise to either excessive complacency in the future or accelerating inflation. I want to briefly sketch what would seem to me on current information to be the appropriate evolution of policy in a number of areas. Of course as data comes in and alternative measures are debated, any particular combination of policies might look less and less appropriate. I will have served my purpose if I have advanced the debate by contributing an example of an ambitious policy program.

Monetary Policies and the Financial System

One former economist official whose advice I sought in preparing these remarks referred to recent events as “adjusting for raised expectations, the greatest failure of risk management in financial history.” This is too apocalyptic. But it is suggestive of the extent to which major financial institutions are unsure of their own and their counterparties creditworthiness.

In normal times the spread between the rate at which the Treasury borrows and the LIBOR rate at which banks lend each other money for 3 months is typically well under half a percentage point. Currently it is about 2 percentage points. In the United States and Europe large and persistent spreads have also opened between the policy rates of central banks and the lending rates at which banks make credit available to each other and to firms and households.

In this environment the dominant risk is a downward spiral in which financial problems curtail credit and spending thereby reducing economic activity, which in turn exacerbates the financial problems, creating a vicious spiral. Once in progress, such a spiral may prove very difficult to arrest. It is much more important to establish credibility that policy is ahead of the credit crunch spiral than to reassure yet again that it is not behind the inflation curve.

I say this not because I am unconcerned about inflation. The achievement of price stability over the last generation is one of the most important factors contributing to improved economic performance. It is a matter of balancing risks. With workers and firms as insecure as they are today, I see little risk of the kind of wage-price cycle that has set off inflation in the past. Data on indexed and nominal bonds suggest that despite what has happened to oil prices and to the dollar there has been no increase in the expected price level several years out. Moreover, failure to contain a credit spiral could cost the economy years of satisfactory economic performance. If I am wrong and policy creates undue inflation pressures, they can be removed at a much less perilous moment.

So far the Fed has responded by cutting its policy rates by a full percentage point and with a number of programs to make liquidity available to banks. The seriousness of the problems is suggested by fact that liquidity provision has not yet made a large dent in the spread between bank and government borrowing rates. While reductions in policy rates have translated directly into lower lending rates, it appears that half or more of their impact has been offset by increases in the spread between policy and lending rates. This means that the apparent easing in monetary policy in recent months has been much greater than the actual easing.

What does this suggest going forward? First it suggests that policymakers should consider focusing attention not on their traditional policy rate but on targeting some more meaningful indicator of the cost of credit to households and businesses (such as 3 month LIBOR). In this way, increases in credit risks will not automatically translate into de facto tighter policy as they do today.

Second, assuring full transparency with respect to the valuation of assets and the recognition of losses and liabilities should be the top regulatory priority. The Japanese experience taught painful lessons about the dangers of government support and encouragement for measures that seek to rearrange balance sheets so as to avoid facing painful realities. Wherever possible assets should be marked to market, not to model, and liabilities should be explicitly recognized.

Third, regulatory policy needs to focus on assuring that financial institutions raise adequate amounts of capital to maintain their activities, even if this is painful for existing shareholders. If a bank is at the point of indifference between reducing the size of its balance sheet and raising capital by issuing shares or cutting dividends, the broader economy is not. Policy in recent months has devoted considerable attention to destigmatizing and indeed encouraging borrowing in one form or other from the Fed. In the months ahead it will be equally important to destigmatize the raising of capital and indeed to insist that institutions raise enough capital to allay credit risks and permit the resumption of normal lending activities.

Fiscal Policies

The success of the Clinton 1993 budget plan in setting off a virtuous circle of growth, reduced deficits, lower interest rates and still more growth – along with a growing sense that short-run stabilization policy is the job of the Fed – have reinforced the economics profession’s growing aversion to the use of fiscal policy to stabilize the economy.

Yet, if economic data over the next several months come in as I fear they will—with increasing signs of recession—several considerations suggest that the policy response should include fiscal as well as monetary stimulus for several reasons.

If policymakers are able to act quickly and effectively, fiscal policy can work more rapidly than monetary policy, which has about a lag of a year between the change in the federal funds rate and its maximum impact. Moreover, the efficacy of monetary policy may well be diminished by capital constraints that limit the ability of banks to lend or by creditworthiness constraints that limit the ability of businesses to borrow.  As important, the extent to which monetary policy can be prudently used in the current environment is limited by concerns about the dollar as well as about the bubble creating effects of very low interest rates.  Finally certain problems—such as the impact of mass foreclosures on affected communities—are not easily amenable to monetary policy.

Fiscal stimulus is critical but could be counterproductive if it is not timely, targeted and temporary.  Gene Sperling’s Bloomberg column this week makes these points strongly.  To respond to an incipient downturn, fiscal policy has to have its impact in a timely manner.  It has to be targeted to assure that increased government borrowing translates directly into increased spending and demand. And, critically, it has to be temporary so that its effects are not offset by higher long-term interest rates.  Indeed from the point of view of stimulus, the optimal package is one that raises spending and the deficit in the short run while reducing the deficit in the long run and thereby reducing long term interest rates.

Any actual fiscal stimulus program would have to be worked out in the context of events as they unfold and should be walled off from longer term policy considerations where actions to assure long term fiscal sustainability are essential.

It is reasonable to suggest that stimulus approaching $50-$75 billion -- roughly in the range of 1/2 of 1% of GDP -- is likely to be appropriate. The largest part of this stimulus should come in the form of tax cuts distributed equally among all taxpayers and recipients of tax refunds.  Other elements of a stimulus package should include extension of unemployment insurance benefits given that long term unemployment is already at recession levels, temporary step-ups in food stamp benefits which can be executed and have effect very quickly, and tax measures to eliminate from taxation the so-called income that homeowners receive when they are foreclosed, a step that has just been passed by the House of Representatives.

In the context of a legislative stimulus program, consideration also should be given to steps that can be taken to help contain energy and food prices.  Such measures both raise consumers’ purchasing power and reduce inflation concerns.  These might include reform of the strategic petroleum reserve to assure that the government stops the practice of accumulating especially scarce oil products at times when markets indicates that current supply is selling at a large premium, and adjustments in policies promoting ethanol to assure that they do not drive up food prices.

Housing and Mortgage Market Policy

Probably the single most important thing economic policy can do for homeowners is to minimize the risk of recession or the severity of recession if it comes. With the bursting of what now can clearly be seen as a pervasive bubble, and the drying up of important segments of the mortgage market, the last thing that the housing market needs is a recession that would reduce incomes of homeowners and potential purchasers. That is why the aggressive fiscal and monetary policy I have just discussed are so important.

But it is also true that problems in the housing sector are an important reason for recession fears and they need to be addressed. The recent teaser-freezer (which freezes the initial teaser rate of some sub-prime mortgages) is a useful step that addresses that relatively small minority of subprime mortgage holders who on the one hand appear very unlikely to be able to get a new mortgages and on the other hand appear very likely to be able to carry their existing mortgage. It is a constructive step but I know of no credible estimate suggesting that it will reduce annual mortgage payments by more than about $5 billion.

It is a perhaps appropriate component of a much broader strategy that recognizes the core problems posed by the sharp decline in housing prices. While the issue of resets is an important one, a much more fundamental problem needs to be addressed. Consider a homeowner who purchased a home for $250,000 putting nothing or next to nothing down implicitly relying on appreciation of the house to service the mortgage. That homeowner finds himself today with a home worth perhaps $220,000 and with the capacity to service perhaps $200,000 worth of mortgage even before any rate reset. If the house is foreclosed, its value will probably decline to $150,000 and adversely affect the neighbors as well.

The best outcome for both borrower and lender is a writedown in the value of the mortgage that allows it to be serviced and at the same time prevents a mutually costly foreclosure just as Chapter 11 of the bankruptcy code prevents the liquidation of overly indebted but viable companies. It is deals of this kind in the subprime, alt A, and prime space that need to be negotiated if families are to be saved the agony of foreclosure and lenders are to maximize their recoveries.

The answer may lie in bankruptcy law reform, standard templates for mortgage restructuring or other means. Various tax and regulatory obstacles to shared appreciation mortgages in which lenders reduce monthly payments in return for a share in a house’s appreciation when it is sold should be removed. Until there is recognition that many individuals who cannot meet their original mortgage obligations are nonetheless the highest value occupants of their homes, we are not going to fully respond to the problems in the housing sector.

Additional steps that should be taken in the next several months include:

• the provision of Federal assistance to those who are foreclosed in locating new rental housing and to communities that wish to purchase foreclosed homes and convert them into rental properties.
• support for an adequate supply of mortgage credit. Proposed increases in the availability of FHA guarantees are a positive development though they are manifestly insufficient to assure an adequate flow of mortgage capital across the entire housing spectrum.
• The Government-Sponsored Enterprises (GSEs), Fannie Mae and Freddie Mac, should be granted significant temporary increases in their portfolio limit so that they can perform their market stabilizing function at the time it has been most needed in two generations. They should also be freed on a temporary basis from punitive capital requirements and the conforming loan limit should be increased to about $600,000.

It is of course possible that developments in the housing sector will prove less serious than I fear and that not all of these measures will have been necessary. How serious a problem will this be? A substantial fraction of the originators of subprime mortgages have gone bankrupt. If I read the political winds correctly, those who remain will face greatly enhanced regulation. The concern that too many homeowners will learn from these events that it is a good idea to excessively lever up their homes seems less than paramount at this point. On the other hand, if policy remains behind the curve families in communities across the country will bear the brunt of the errors.

Conclusion

While it has not been my topic this morning, I trust that extensive efforts will be made to learn from painful experience. Most obviously and visibly there is the need to protect vulnerable people from the kind of predatory lending practices that have been all too common in recent years. Recent experience also suggest the need for reevaluation of traditional approaches to monetary policy, the regulation and provision of liquidity to different types of financial institutions, the role of the rating agencies and much else.

It has always seemed to me that those of us involved with finance bear great responsibility. There is the great importance of well functioning capital markets and the credibility of the currency. Much more important is the reality that when the economy are successfully managed people’s fortunes are determined by their own choices and efforts. When the wrong economic policy choices are made people’s lives can be wrenched apart as they lose their jobs or their homes or their ability to provide for their family because of complex forces entirely beyond their control.

The economy is at as critical a juncture as it has been in many years. Policy must balance risks at a highly uncertain moment. The lives of millions of people who will never think about countercyclical policy, moral hazard, lending facilities or the federal funds rate may be profoundly affected by the policy choices made in this city in the next few months. I hope they will be made both urgently and wisely.