Finance & Development, September 2009, Volume 46, Number 3

The recovery has started. Sustaining it will require delicate rebalancing acts, both within and across countries

Olivier Blanchard

PDF version

IN normal recessions, however disruptive they are to businesses and jobs, things turn around predictably. The current global recession is far from normal.

Usually, to fight a recession, the central bank lowers interest rates, which results in increased demand and output. People resume buying durable goods such as appliances and cars. Firms start delayed investment projects. Often, an exchange rate depreciation gives a boost to exports by making them cheaper. The lower-than-normal growth during the recession gives way to higher-than-normal growth for some time, until the economy has returned to its normal growth path.

But the world is not in a run-of-the-mill recession. The turnaround will not be simple. The crisis has left deep scars, which will affect both supply and demand for many years to come.

Supply-side problems

Olivier Blanchard: "Some parts of the economic system have broken."

Some parts of the economic system have broken. Some firms went bankrupt that would not have in a normal recession. In advanced countries, the financial systems are partly dysfunctional, and will take a long time to find their new shape. Meanwhile, financial intermediation—and, by implication, the process of reallocation of resources that is central to growth—will be impaired. In emerging market countries, capital inflows, which decreased dramatically during the crisis, may not fully come back in the next few years. Changes in the composition of world demand, as consumption shifts from advanced to emerging economies, may require changes in the structure of production. In nearly all countries, the costs of the crisis have added to the fiscal burden, and higher taxation is inevitable.

All this means that we may not go back to the old growth path, that potential output may be lower than it was before the crisis.

How much has potential output decreased? It is difficult to tell: we do not see potential output, only actual output. The historical evidence is worrisome, however. The IMF’s forthcoming World Economic Outlook presents evidence from 88 banking crises over the past four decades in a wide range of countries. While there is large variation across countries, the conclusion is that, on average, output does not go back to its old trend path, but remains permanently below it.

The possible good news is that the trend itself appears to be unaffected: on average, crises permanently decrease the level of output, but not its growth rate. So, if past is prologue, the world economy likely will return to its past growth rate. But, especially in advanced countries, the period of above-average growth, characteristic of normal recoveries, may be short-lived or nonexistent.

Demand-side issues

Just achieving “normal” growth, however, may be hard because of demand problems. The forecasts now predict that growth will be positive in most countries, including advanced countries, for the next few quarters.

But there are two caveats to this news:

• Growth will not be quite strong enough to reduce unemployment, which is not expected to crest until some time next year.
• These positive growth forecasts are largely predicated on a combination of a fiscal stimulus and inventory rebuilding by firms, rather than on strong private consumption and fixed investment spending. Sooner or later, the fiscal stimulus will have to be phased out. And inventory adjustment will also naturally come to an end.

The question, then, is what will sustain the recovery.

Two rebalancing acts will have to come into play. First, rebalancing from public to private spending. Second, rebalancing aggregate demand across countries, with a shift from domestic to foreign demand in the United States and a reverse shift from foreign to domestic demand in the rest of the world, particularly in Asia.

Rebalancing public and private spending

The fiscal response to the crisis was to increase government spending, lower taxes, and accept much larger fiscal deficits. Given the collapse of private demand, and the inability to reduce interest rates below zero, governments clearly chose the right response. But large deficits lead to rapid increases in debt, and, because debt levels were already high in many countries, such increases cannot go on for long. As large deficits continue, debt sustainability comes increasingly into question. And with this comes the risk of higher long-term interest rates, both because of anticipated crowding out of private borrowers by government borrowers and because of a higher risk of default.

How much longer can the fiscal stimulus continue? On its own, in most advanced countries, probably not very long. The average ratio of debt to gross domestic product (GDP) for the G-20 advanced economies was high before the crisis, and is forecast to exceed 100 percent in the next few years. (The situation is substantially different in a number of emerging market countries, where debt was much lower to start, and where there is more room for deficit spending.)

An important qualifier is “on its own.” The stimulus can be prolonged if, at the same time, structural measures are taken to limit the future growth of entitlement programs—whether from rising health care costs or from the effect of aging populations on retirement costs. The trade-off is fairly attractive. IMF estimates suggest that the fiscal cost of future increases in entitlements is 10 times the fiscal cost of the crisis. Thus, even a modest cut in the growth rate of entitlement programs can buy substantial fiscal space for continuing stimulus.

Eventually, however, the fiscal stimulus will have to be phased out, and private demand must replace it. The source of that demand—whether consumption or investment—is a crucial issue.

Rebalancing demand across countries

The United States was not only at the origin of the crisis, it is central to any world recovery. Consumption represents 70 percent of total U.S. demand, and its decline was the main near-term cause of the fall in output in this crisis. The ratio of U.S. household saving to disposable income, which was close to zero in 2007, has increased to about 5 percent. Will the saving rate go back to its 2007 level? That would not be desirable, and is unlikely.

On the one hand, some of the increase in saving in the last year probably reflected a wait-and-see attitude on the part of consumers, an attitude that will go away as the smoke clears. On the other hand, the saving rate tends to go up as output and income expand. And even if financial wealth returned to its pre-crisis level—be it in housing (which seems undesirable and unlikely) or in stocks—and output returned to its trend path, U.S. consumers would still probably save more. The reason is that the crisis has made them more conscious of tail risks—events that are unlikely to occur but, when they do, have devastating consequences.

Before the crisis, it was an article of faith that housing prices rarely, if ever, decreased (a belief that was a main contributor to the crisis). Another article of faith, one backed by stronger historical evidence, was that investors could count on stocks yielding an annual rate of return of 6 percent. Last year’s decline in the stock market showed that those yields cannot be taken for granted, and that more saving may be needed to ensure a safe retirement. Thus, U.S. consumers are likely to save more, at least until they forget the lessons of the crisis. The best guess (and there is little more to go on) is that the U.S. household saving rate will remain at least at its current level. That means a 5 percentage point decline in the ratio of consumption to disposable income relative to the pre-crisis period, or about a 3 percentage point drop in the ratio of consumption to GDP. Put simply, 3 percent more of U.S. aggregate demand will have to come from something other than consumption.

Will it be from investment? This also seems unlikely. Housing investment, as a percentage of GDP, was too high in the years preceding the crisis, and it will take a long time to get rid of the backlog of houses. Until that happens, housing investment will be low. Will fixed investment, again as a percentage of GDP, be higher after the crisis than it was before? Probably not. Capacity utilization is at a historical low, and will take a long time to recover. While banks may be solvent now, they are still tightening credit, and tight lending standards are likely to last a while. Less-efficient financial intermediation will affect not only the supply side, but also the demand side. Again, historical evidence from “creditless” recoveries suggests that investment will be weak for a long time.

Can low interest rates help?

It is likely that, at any given interest rate, U.S. private domestic demand will be weak for a long time, weaker than it was before the crisis. Note, however, the qualifier “at any given interest rate.” This appears to offer room for some optimism. The short-term riskless rate is lower now than it was in the pre-crisis years. Over the three years before the crisis, the average nominal U.S. treasury bill rate was 4 percent, while the average inflation rate was 3 percent. That resulted in a real—that is, after-inflation—rate of 1 percent. Today, the treasury bill rate is roughly zero and inflation expectations appear anchored around 2 percent. That implies a real rate of around –2 percent—that is, 3 percentage points below its pre-crisis level. The Federal Reserve can leave the policy rate—the federal funds rate—at zero if it needs to, and, because inflation expectations are more likely to increase than to decrease, real rates are likely to remain negative. An old rule of thumb is that a 1 percentage point lower real rate that is expected to remain so for some time leads to a roughly 1 percent increase in aggregate demand. A decrease in the real rate of 3 percentage points would seem sufficient to offset the caution of consumers and firms and sustain the recovery.

But it may not be. What matters for demand is the rate at which consumers and firms can borrow, not the policy rate itself. As was clear during this crisis, the rate at which consumers and firms borrow often is a lot higher than the policy rate. Risk premiums on U.S. BBB-rated bonds, for example, are nearly 3 percentage points higher than before the crisis. This higher risk perception may well be an enduring legacy of the crisis. (The Great Depression led to a large increase in the risk premium on stocks, which lasted for the better part of four decades. But the Depression lasted a long time, and this crisis appears unlikely to have the same psychological impact.) Higher risk premiums, then, could undo, at least in part, lower policy rates. U.S. policymakers cannot count on low interest rates alone to deliver a sustained U.S. recovery.

Can Asia help?

If the U.S. recovery is to take place, if the fiscal stimulus must be phased out, and if private domestic demand is weak, then U.S. net exports must increase. In other words, the U.S. current account deficit must decrease. That means that the rest of the world, now in substantial surplus, must reduce that current account surplus. Where should this reduction come from?

It is natural to look first at the countries with large current account surpluses. Among them, most prominently, are Asian countries. And most prominent among them is China. From the point of view of the United States, a decrease in China’s current account surplus would help increase demand and sustain the U.S. recovery. That would result in more imports from the United States, which would help sustain world recovery.

Why might China be willing to go along? Because it may well be in its own interest: China’s growth has been based on an export-led growth model that relies on a high saving rate, leading to low internal demand, and a low exchange rate, leading to high external demand. The model has been highly successful, but is leading to the accumulation of extremely large reserves, and pressure is building to increase consumption. The high rate of saving reflects the lack of social insurance and the resulting high precautionary saving by households, limited access of households to credit, and governance issues in firms that lead them to retain too high a proportion of their earnings. Providing more social insurance, increasing household access to credit, and improving firms’ governance are all desirable on their own, and would lead to both lower saving and higher internal demand. If such an expansion of demand runs into supply-side constraints, this higher internal demand would have to be partly offset by lower external demand, meaning an appreciation of the Chinese renminbi (RMB) at least in real terms. Both higher Chinese import demand and a higher RMB would increase U.S. net exports.

Other emerging market Asian countries also run large current account surpluses. Their motivations vary—some want to accumulate reserves as insurance, others chose an export-led growth strategy that incidentally affects the current account and reserve accumulation. Many of these countries could decrease saving, public or private (as the dramatic decline in household saving in Korea since the 1990s demonstrates), and allow their currency to appreciate. That would lead to a shift from external to internal demand and to a reduction in their current account surplus.

Their incentives, however, are weaker than China’s. Having substantial reserves has proved very useful in the crisis. Swap lines from central banks, and multilateral credit lines—such as the “flexible credit line” created by the IMF during the crisis—could reduce the demand for reserves. But swap lines and credit lines might not be renewed, and so do not offer quite the same degree of safety as reserves. (Establishing arrangements to substantially reduce reserve accumulation would also both be highly desirable in the long run and help to sustain the recovery in the short and the medium run.) Thus, countries that have adopted an export-led growth model may reassess that policy and give more weight to internal demand, but any change is likely to be gradual.

To get a sense of magnitudes, another rough computation is useful. The GDP of emerging Asia is roughly 50 percent of U.S. GDP (with the ratio projected to increase to 70 percent in 2014). So, if all their trade was with the United States, Asian countries would have to lower their current account position by 4 percent of GDP to improve the U.S. current account by, say, 2 percent of GDP (under the assumption of a 3 percent shortfall in the ratio of consumption to GDP, minus a 1 percent increase coming from lower real interest rates). Since emerging Asia’s trade is not all with the United States, the adjustment would likely have to be even larger. This raises the question of whether other countries can and should play a role.

What role for non-Asian countries?

A number of other countries, including some advanced countries, also have current account surpluses. For example, Germany’s surplus for 2008 is half China’s (although it is shrinking fast); Japan’s surplus is one-third of China’s.

Should Germany, for example, reduce its surplus? It cannot follow the same route as that suggested for China—that is, a currency appreciation accompanied by a decrease in saving. Because it is part of the euro area, Germany cannot engineer an appreciation on its own. And, on the demand side, it suffers largely from the same problem as the United States: it has limited room on the fiscal side, and it is not clear that it is either desirable or feasible to get German consumers to save less. Germany could, however, improve productivity in its nontradable sector, which would be in its interest. This would, in time, lead to a reallocation of demand toward nontradables and reduce its current account surplus. The same argument applies to Japan. But, because such structural reforms are politically difficult, and because their effects take place slowly, it is likely to be a slow process—too slow to provide substantial support to the recovery over the next few years. So, if rebalancing is to come soon, it probably has to come largely from Asia, through a decrease in saving and an appreciation of Asian currencies vis-à-vis the dollar.

What if rebalancing does not happen?

This tour of the world suggests three conclusions:

• First, the crisis is likely to have led to a decrease in potential output. One should not expect very high growth rates in the recovery.
• Second, sustained recovery in the United States and elsewhere eventually requires rebalancing from public to private spending.
• Third, sustained recovery is likely to require an increase in U.S. net exports and a corresponding decrease in the rest of the world, coming mainly from Asia.

One can question all three conclusions.

On the supply side, the effect on potential output is highly uncertain. After all, despite the pessimistic historical evidence, some countries have emerged from banking crises without experiencing a visible impact on potential output (on the other hand, though, some countries have seen a long-lasting negative impact not only on the level of GDP, but also on its growth rate).

On the demand side, the fiscal space in advanced countries may be larger than expected, allowing the United States to sustain longer-lasting deficits and a higher debt level than currently forecast without raising market concerns about debt sustainability. If this is the case, rebalancing private and public spending can be phased in more slowly if needed, allowing more time to achieve a rebalancing of world demand. Alternatively, private demand in the United States may be stronger: U.S. consumers could return to their old ways and save less. That would help the recovery and avoid the need for a major adjustment of net exports, although it would re-create in the longer run some of the problems that caused the current crisis. Or it could be that the world decouples—that Asia, for example, is able to return to high growth, while recovery in advanced countries falters. But the crisis, and the strong export links that turned a U.S. shock into a world recession, suggests that decoupling, although possible, is unlikely.

If, however, one accepts the argument that both rebalancing acts are likely to be necessary for a sustained recovery, the next question is whether they will take place. It is clear that they may not, at least not on the scale needed. If, for example, Asia is unwilling to reduce its current account surplus and U.S. net exports do not substantially improve, weak U.S. private demand may lead to an anemic U.S. recovery. In that case, there would likely be strong political pressure to extend the fiscal stimulus until private demand has recovered.

Were that to happen, one can imagine various scenarios: political pressure may be resisted, the fiscal stimulus could be phased out, and the U.S. recovery might falter. Or fiscal deficits might be maintained for too long, leading to issues of debt sustainability and worries about U.S. government bonds and the dollar, and causing large capital flows from the United States. Dollar depreciation may take place, but in a disorderly fashion, leading to another episode of instability and high uncertainty, which could itself derail the recovery.

Sustaining the nascent recovery is likely to require delicate rebalancing acts, both within and across countries. An understanding of the issues and the dangers, and some coordination across countries, is likely to be as crucial during the next few years as it was during the most intense part of the crisis.