The U.S. Current Account Deficit and the Global Economy
- International Monetary Fund
- Published Date:
- December 2004
Lawrence H. Summers
I am deeply honored by the invitation from the Per Jacobsson Foundation to deliver this prestigious lecture and join the list of distinguished lecturers who have been part of this series, and am also very glad to see so many friends and colleagues.
I last had the chance to attend the Annual Meetings in the fall of 2000. Some things have stayed the same, some things have changed. Then there was a close U.S. election, then there was great concern about rising oil prices, then there was a hope of collective efforts regarding debt relief for the poorest countries. These things all give a sense of déjà vu.
But things have also changed. The geopolitical environment is profoundly different than it was at that time and far more preoccupying than it was at that time. And perhaps in part for that reason, the other major difference—the very substantial increase in the pattern of global imbalances in general and in the U.S. current account deficit in particular—is perhaps receiving less attention than it should.
It is this question of global imbalances and the U.S. current account deficit that I want to address this afternoon. The U.S. current account deficit is currently running well in excess of $600 billion at an annual rate, in the range of 5.5 percent of GDP. It represents well over 1 percent of global GDP and absorbs close to two-thirds of the cumulative current account surpluses of all the world’s surplus countries. All of these figures are without precedent. The United States has never run such large current account deficits and no single nation’s deficit has ever bulked nearly as large relative to the global economy. At a minimum, such a unique imbalance deserves careful scrutiny.
I want to ask, and try to answer, four questions regarding the U.S. current account deficit in my remarks this afternoon. As I sequence the questions, there will be some suggestion as to my answers. First, is the U.S. current account deficit a transient that will self-correct without a major discontinuity in the global economy? Second, is the large and growing U.S. current account deficit a sign of economic vitality or an incipient problem? Third, is the U.S. current account deficit and the associated reliance on official intervention from nations dependent on exports to the United States a desirable or sustainable state of affairs? Fourth, what is to be done?
The U.S. Current Account Deficit in Perspective
Let’s turn, first, to a simple look at the current account deficit. In round numbers, the United States is importing 16 percent of its gross domestic product, and it is exporting a bit less than 11 percent of its gross domestic product. So exports are about two-thirds of imports, or imports are about 50 percent larger than exports.
This has a disturbing arithmetic implication. If the global economy grows in a completely balanced way, with all imports and exports rising in proportion to the size of the global economy, the U.S. current account deficit deteriorates. Each $1 increase in U.S. exports is associated with an additional $1.50 in U.S. imports. Or, to put the point differently, if over a five-year period, U.S. imports and exports both grow at 6 percent (about in line with historic averages) and the U.S. economy grows at 3 percent, the current account deficit on this account alone will increase by close to 1 percent of GDP. But the situation is actually rather worse than this for what appear to be reasons of population growth and which appear at the current moment to be reasons of structural productivity growth. U.S. economic growth, in fact, significantly exceeds the economic growth rate of the United States’s major trading partners, and therefore imports are sucked in by economic growth at a rate that exceeds the impact of foreign economic growth on U.S. exports.
There is yet another consideration, quantitatively probably more important though somewhat more mysterious, pointing in the same direction. For reasons that economists poorly understand, it was first noted in 1969 that the elasticity of U.S. exports with respect to foreign economic growth is less than the elasticity of U.S. imports with respect to domestic economic growth. That is to say, our growth sucks in more imports than their growth sucks in exports. This is known as the Houthakker-Magee effect. Lacking a convincing explanation, economists have predicted that this was some kind of anomaly that would go away in the next 30 years, and it has not, as yet. Consistently, the output elasticity of foreign imports is far less than in the case of American imports. Balanced growth means a deteriorating current account deficit. Growth is likely to be unbalanced in a way that exacerbates the situation. And even with balanced growth, there are likely to be unbalanced impacts on imports and exports.
Taking all this together, one can understand the conclusions that almost all the modelers reach with respect to the evolution of the U.S. current account deficit in the absence of any discontinuities. The most recent such study I had a chance to read is that of Nouriel Roubini and Brad Setser, who predict that, without a discontinuity, the U.S. current account deficit will rise to 6.5 percent of GDP in 2006, and 7.8 percent of GDP in 2008. Using a different model and somewhat different methodology, Catherine Mann predicts a current account deficit that could approach 13 percent of GDP by 2010, without some sort of major discontinuity. Note carefully, these are not forecasts of what will occur—they are predictions of what would happen without discontinuous changes in growth rates, patterns of demand, or relative prices of different countries’ output.
There is much that one can argue about in the forecasts and the models, but I’m aware of no credible argument that without some form of discontinuity, the U.S. current account deficit will not increase from its current high level.
Interpreting the U.S. Current Account Deficit
Second question: Is this current account deficit a sign of vitality or of incipient problems? There is a standard set of things that the finance ministers of countries with significant current account deficits say, suggesting that such deficits are somehow a sign of economic strength. Perhaps the sharpest formulation that I have heard is: “We live in a country that capital is trying to get into. Would you rather live in a country that capital is trying to get out of?”
It is important in examining a current account deficit to understand its roots. Tautologously, a current account deficit is the difference between national savings and national investments; or, equivalently, between net national savings and net national investment, removing the effects of depreciation. It is natural to look at the U.S. current account deficit and ask whether its level and its deterioration is better attributed to increased investment or to reduced saving.
In the last year, the net national savings rate of the United States has been between 1 and 2 percent. That is to say, if one adds personal savings, corporate savings, and government savings—in this case government dissavings—and calculates them as a ratio of NNP, one is left with a figure between 1 and 2 percent. This represents a substantial deterioration over the last five years. It represents the lowest net national savings rate in American history, and I believe that of any major nation. At 1.5 percent, the national savings rate is about half of what it was in the late 1980s and early 1990s, when national saving was last a major item on the U.S. policy agenda. In fact, net investment has declined over the last four or five years in the United States, suggesting that all of the deterioration of the current account deficit can be attributed to reduced savings and increased consumption rather than to increased investment.
A second question that one asks in looking at a large current account deficit in the classic emerging market context is whether that investment which is taking place is taking place in the traded goods sector, where it is generating the export capacity that can ultimately service debt, or whether investment is being allocated increasingly to the non-traded goods sector. Here, too, the record is clear: an unusual interest rate environment and heavy foreign competition in manufacturing have changed the composition of investments in the United States substantially toward the non-traded goods sector as manifested particularly clearly in the dramatic increases in the price of residential real estate in and around most major American cities.
There is another way to look at this constellation of issues: in terms of the type of capital inflows which are financing the current account deficit. Here, too, there is much that one can cavil about. There are significant discrepancies that some people in this room probably understand, but I do not, between BIS figures on central bank accumulation of reserves and U.S. Bureau of Economic Analysis figures on official financing of the current account deficit.
But the dominant picture, it seems to me, is quite clear. Brad Setser calculates from the most recent BIS annual report that the central bank reserves of Asia, inclusive of Japan, increased from about $1.1 trillion to about $1.8 trillion from the end of 2001 to 2003—a $700 billion increase in reserves—with Japanese reserves increasing by $265 billion, Chinese reserves by $191 billion, Indian reserves by $52 billion, and the reserves of the newly industrialized economies of Asia by $163 billion. The reserves of Taiwan are in excess of the reserves of all of Latin America.
The evidence from this year is less clear, and there is some suggestion in the data that the extent of reserve accumulation may have tailed off somewhat. But the basic picture that a large fraction of the U.S. current account deficit is being financed by foreign central bank intervention is not one that can be argued with.
There are different ways of describing this system. Michael Dooley, Peter Garber, and David Folkerts-Landau have referred to it as a kind of new Bretton Woods system. Catherine Mann has described it as codependency. Another term for it is “international vendor finance.” It is, I think, relatively clear what is going on. A substantial number of countries are maintaining a fixed or quasi-fixed exchange rate through very substantial exchange rate intervention and enjoying strong export performance to the United States as a result. It was the common advice and, in my judgment, the correct advice and the correct lesson to learn from the experiences of the mid-1990s and the late 1990s, that emerging market countries could profitably have more reserves than was formerly thought to be the case, and that the capacity to roll your own massive IMF program from reserves was something on which policymakers could sensibly put a premium. I believe we are well past the point in many countries where reserve accumulation can sensibly be attributed to a prudent insurance motive with respect to the prospect of future financial crisis.
Nor, looking at the interest rate held on these reserves—which is negative in real dollar terms and negative in local currency terms—if an appreciation lies in the future, is it reasonable to suppose that the primary motivation for the accumulation of these reserves is that it represents the highest-value investment opportunity open to the societies that are investing on such a scale in these reserves. Large current account deficits are likely to grow, replacing declining saving with substantial finance from abroad coming through the official sector.
The next question to ask is whether an arrangement of this kind is desirable, if it can be maintained, and, if desirable in some elements, is likely to be sustainable over the medium and the long term. I have already hinted at some of the concerns. Consider, first, whether it is desirable if it could be maintained. For the United States, an arrangement of this kind, even if it could be maintained indefinitely, carries with it two very substantial risks. One risk is the incipient protectionist pressures that are generated by a large trade deficit, and a trade deficit that is associated with financial practices that have as part of their motivation the promotion of exports. There is no question, in my judgment—and it’s a judgment I came to painfully but fairly confidently—that protectionist pressure in the United States is on an upwards trend. You see it in political debates over trade. You see it in the current furor over outsourcing, which commands so much attention in this political year.
The second risk of a system of this kind, even if it could be maintained for a significant number of years, is a more amorphous one but one that is no less serious. Inevitably, dependence on foreign governments for short-term financing has to raise questions and create vulnerabilities in both the economic and political realms. The question can fairly be asked: How long should the world’s greatest debtor remain the world’s largest borrower? I have previously used the term “balance of financial terror” to refer to a situation where we rely on the costs to others of not financing our current account deficit as assurance that financing will continue. The term may overdramatize the problem, but this is surely a situation of concern.
If these arrangements are problematic, even if they could be maintained for the United States, what of the countries that are providing finance on a substantial scale to the United States with respect to the current account deficit and the maintenance of their exchange rates? Here, too, there are two substantial questions of desirability, even apart from the question of sustainability. A great deal of money is being invested at what is almost certainly a very low rate of return. To repeat, the interest earned in dollar terms on U.S. short-term securities is negative. If even a modest appreciation of a country’s currency lies in the future, the valuation losses on reserves will be quite substantial in local terms.
Second—and the significance of this will vary from country to country—there is, as always with fixed exchange rate regimes, the loss of domestic monetary control and the difficulty of maintaining monetary policy in response to domestic conditions. One sees this very clearly, particularly in the Chinese case where the external constraints exert an important impact on domestic monetary policy. It is noteworthy that much of the speculative bubble in Japan during the late 1980s that had such a catastrophic long-run impact on the Japanese economy was driven by liquidity produced by a desire to avoid excessive yen appreciation.
The current system is also problematic for Europe. With a limited number of currencies that are flexible against the dollar, the flexible currencies will bear disproportionately the impact of any changes in sentiment regarding dollar assets.
This, then, is an arrangement that is not without its virtues—low-cost finance for America at a time when savings are low, strong exports and a very competitive traded goods sector for those who are providing the finance—but it is one that also has very important costs. And then there is the question of the sustainability of the arrangements.
The comparison is made to the Bretton Woods regime, but as Barry Eichengreen notes in his very perceptive paper on this subject, the Bretton Woods regime did end. The Bretton Woods regime took place at a time when the United States had a current account surplus, not a current account deficit. And the Bretton Woods regime took place in an era of substantially less capital mobility than the one we expect—and the one we enjoy—today.
When and in what way does this regime of rising current account deficit increase U.S. debt reliance on official sector finance? Whether it will end on its own accord is impossible to predict, but I would suggest that if one is seeking to draw lessons from the last 15 years of monetary experience, here is one that is very powerful: fixed exchange rates with heavy intervention have enormous capacity to create an illusory sense of stability that could be shattered very quickly. That is the lesson of Britain in 1992, of Mexico in 1994, of emerging Asia in 1997, of Russia in 1998, and of Brazil in 1998.
To be sure, there are important differences between exchange rates that are being protected against pressure for depreciation and exchange rates that are being protected against pressure for appreciation. But the basic point, I believe, remains the same: governments find that a calculus of costs and benefits that weighs political costs highly does not support an adjustment of an ultimately unsustainable exchange rate, absent crisis or financial pressure to do so. By the time adjustment can be justified, it is too late, and the costs, when the adjustment takes place, are very serious. This is a movie we have seen many times in the international financial system: large current account deficits, current account deficits that replace savings and are financed officially in a system that is uneasy in its consequences and unlikely to endure indefinitely as debt accumulates.
Observations on the Road Ahead
What is to be done?
There is the strategy of averting one’s gaze, of not seeking to raise questions, of hoping that the flows will continue, and waiting until the day when they don’t. I would suggest that this is a risky course. It is not a course whose consequences are easy to predict when there will be a problem, if there will be a problem in the next several years, whether the problem will take the form of a need to raise interest rates sharply in the United States, a dramatic increase in protectionist pressures, difficulties of monetary control in the Asian economies. None of these can be predicted with confidence, but it seems to me that the risk of adverse consequences along the road is such that it is prudent to think about managing the situation in some planned way.
What, then, do I have to recommend in terms of what is to be done?
Of course, one of the great glories of being an academic rather than a policymaker is that one doesn’t have quite the same obligation to pose specific remedies, that one can take refuge in exhorting others like those in this room. But I would make five observations with respect to what is to be done.
The first thing I would say is that I will have served my purpose and achieved my objective if I have stimulated thought about the fact that there is a major issue here of exit from this regime that needs extensive consideration, and I am more confident of that conclusion than I am of any other particular conclusion that I’m going to offer.
Second: It seems to me that if these matters are to be reflected upon internationally—and for reasons I will make clear in just a moment, I believe they are matters that require global rather than just domestic consideration—it is unlikely that the G-7 is the appropriate grouping for the totality of that reflection. The G-7, after all, is made up of the United States, Japan, Canada, and four European nations, and the largest part of the current account deficit that is of concern, and of the exchange rates that are fixed or quasi-fixed—the international vendor finance that I referred to—are coming from outside of the G-7.
It would be my hope that reflection would take place in global fora, and it would seem to me that we have come rather more quickly than I had expected in the late 1990s when I was involved in the establishment of the G-20 Group to a situation where the G-20 can consider issues of global economic coordination that are not predominantly about what industrial countries are going to do to bail out emerging market countries that are in trouble, but rather with respect to the overall global economic strategy for sustained growth.
Third observation: A significant increase in U.S. national savings is a necessary but not a sufficient response to the imbalances that I have described. It is a necessary response because it is difficult to imagine an adequate rate of U.S. investment or an appropriate rate of interest in the United States with 1.3 percent or anything close to it in net national savings, and the removal of substantial foreign official finance.
It is surely a concomitant of a healthy adjustment in the U.S. current account deficit that national savings increase, but it is not a sufficient response to the global imbalance. After all, a current account deficit can decrease for only two reasons: it can decrease either because of a change in the quantity of goods demanded in the country in question or because of a change in their relative price. A moment’s calculation will reveal that any attempt to adjust a large part of the U.S. current account deficit by simply slowing down growth in the U.S. economy will involve a slow-down in growth that would be unacceptable in the United States and would have very severe consequences for growth globally. Therefore, an appropriate adjustment must also involve a change in their relative price.
There is no mechanism through which an increase in U.S. national savings will lead to an adjustment if quasi-fixed exchange rate policies are maintained by a substantial fraction of the United States’s major trading partners. Indeed, the impact of an increase in the U.S. national savings, with no other change, could be a need for greater intervention and greater official finance with greater monetary consequences for the nations that are seeking to peg their exchange rates.
Fourth observation: An additional concomitant of a healthy strategy must be some adjustment and, I believe, coordinated adjustment of exchange rates that are currently quasi-fixed to the dollar. The argument is sometimes made that the current system runs the risk of pressure arising for its end, not unlike the pressures that arose at the end of the Bretton Woods system, where some nation will weary of holding and accumulating dollar assets, given what they judged to be the low expected returns, and will make a currency adjustment, and then others seeing the hand-writing on the wall will follow. That is one aspect of the situation.
Another aspect, which I would guess looms larger, is that unilateral exchange rate adjustments will run into very substantial competitive pressures against those who allow a unilateral exchange rate appreciation. And so exchange rate appreciation is likely to be more politically acceptable and more economically acceptable if it does not take place unilaterally but takes place in a coordinated fashion. This, too, is a remedy that doesn’t work very well on its own, and exchange rate adjustment in Asia and an associated substantial reduction in foreign purchases of U.S. treasury bills would put American interest rates and the American recovery at significant risk with, again, risks to the global economy.
When one thinks about this problem carefully, I think one decides that both elements of the necessary adjustment process—an increase in U.S. national savings and an adjustment of currently fixed exchange rates—are much easier and much healthier with the complementary measure, and that is of course the classic case for policy coordination: when measures which are desirable taken jointly are risky taken individually. It would be my hope that in the not-too-distant future, through some set of discussions, through some set of coordination, tacit or explicit, we find our way to such a set of adjustments.
To be sure, there is enormous short-run functionality and comfort in the current system, but it seems to me that our experience with current account deficits of this kind, on the one hand, and quasi-fixed exchange rates, on the other, suggests that it is healthiest to make adjustments before, rather than after, there is great pressure to do so.
Thank you very much.