The United States and the Global Adjustment Process
by Lawrence H. Summers, Harvard University
Speech at the Third Annual Stavros S. Niarchos Lecture
Institute for International Economics
March 23, 2004
Institutions like the Institute for International Economics make an enormous difference. John Maynard Keynes famously observed that almost everything statesmen said in his day represented the distilled frenzy of a defunct scribbler. Today the world moves faster. Almost everything that statesmen say represents the content of a fax from a think tank within the last several months and no think tank more frequently than the Institute for International Economics. When Alexis de Tocqueville wrote about the United States, he said that one of its great strengths was the number of associations and organizations that had that brought people together to solve problems and that the connections that it forged represented an important national asset. Increasingly, we need to solve problems and address issues on a global scale. And the same kind of coming together of individuals that has been a source of strength for the United States is going to be necessary on a global scale. That’s one of the reasons entities like the Institute for International Economics make such a very great contribution.
I was reminded of that earlier today when Fred Bergsten told me about how the European Union at a recent summit, no doubt motivated in part by conviction—and, no doubt, motivated in part by what we used to call “deliverables” when in government—decided that they would form a European Center for International Economics, modeled on the Institute for International Economics. That is a true compliment to the work that you, Fred, and all your fellows have done. And it seems to me that at a moment in the world where misunderstanding by the United States and misunderstanding of the United States are at an all-time high, the work of an organization like this is especially important.
Preparing to speak here tonight was a mild Rip van Winkle experience for me. I have not left controversy or politics behind in leaving Washington, but for the last three years, neither the US current account deficit nor the current state of the budget deficit nor even the fluctuations of the dollar have been a preoccupation. When I last was engaged in these questions, I was concerned by issues of the current account deficit and the adjustment problem of a global economy. I observed in 1999 or 2000 that the global economy depended on the US economy, that the US economy depended on the consumer, that the consumer depended on the stock market, and that the stock market depended increasingly on 30 or 40 stocks. I also remarked that the main thing we have to fear is the lack of fear itself and that the world economy could not fly forever on a single American engine. Some of what I observed then is still true today. The world is, if anything, more dependent on a single American engine, as the US current account deficit has widened significantly since the late 1990s. Risk premiums, as they were then, look very low today, but unfortunately it is no longer true that the main thing we have to fear is the lack of fear itself. Rather, there is much in the current American economic situation and its implications for the global economy that should concern and preoccupy us.
I am reluctantly convinced that the most serious problem we have faced in the last 50 years is that of low national saving, resulting dependence on foreign capital, and fiscal sustainability, which has far-reaching implications for the US and the global economy. Why do I make such strong statements? Look carefully first at figure 2. It plots the rate of net investment in the United States—that is, the growth in equipment and structures after subtracting depreciation—and the level of net national saving—the resources that Americans are saving net of the amount that the federal government is borrowing. The difference between them, of course, represents US borrowing on the national scale—the current account deficit. Figure 2 shows several things quite clearly. The current account deficit has widened sharply over the past four years, relative to an unprecedented rate of 5 percent of US net national product. More than 100 percent of the deterioration of the current account deficit is accounted for by a drop in the level of national saving, and the US net national saving rate—the savings relative to income that Americans are putting aside for the future—was 1.3 percent of our national income in the latest available data, the lowest level since the postwar period and about 35 percent of the lowest level previously reached in the early 1990s, when inadequate saving and dependence on foreign capital were an economic preoccupation.
Why has net national saving declined so severely? If you look at figure 1, which plots private saving and national saving, the answer is very clear. Private saving has been trending downward for many years, with breaks in the trend associated with periods when the stock market performed poorly. But the clear change in national saving, which has declined precipitously in recent years, comes from the increase in the federal budget deficit, which accounts for the fact that the United States now has that lowest rate of national saving in its history. Indeed, the federal deficit now absorbs three-quarters of the private saving generated by the rest of the economy.
Is this low level of national saving, and its links to the budget deficit, a serious problem? If it were a cyclical phenomenon or if it were a reflection of transient expenditure needs, then it would not be a major problem. We would simply expect the situation to work through with time. To answer that question, one needs to look at realistic long-term fiscal projections that reflect, in a reasonable way, a business-as-usual economic strategy for the United States. There are different ways of making such estimates, and reasonable people can differ as to how they should be made.
The estimates I find most convincing in this regard are those of Bill Gale and Peter Orszag at the Brookings Institution. They take the baseline Congressional Budget Office forecast and make three adjustments. They assume that expiring tax provisions that Congress regularly extends will be extended. They assume that the alternative minimum tax will not be allowed to become a tax that is part of the life of a third of all Americans and the majority of significant taxpayers, and they assume that discretionary spending, which has historically risen faster than income, will simply rise with inflation and the number of people in the United States. When they make that calculation, they do not exclude Social Security or Medicare; they simply make the forecasts. The budget deficit over the next 10 years is just comparable, relative to GDP, to the budget deficit in 2003 that drove the trends just described.
As someone with some experience in these forecasts, I find these estimates quite optimistic. They do not take account of the enormous serial correlation in revisions of budget forecast estimates, which have been revised every year in a negative direction for the last four years. They assume that the current prevailing estimates of the costs of the prescription drug benefit in Medicare will prove accurate. They assume the absence of any new major initiative on either the spending or tax-cutting side by the United States for the next six or seven years. They could turn out to be right. It could turn out that GNP growth will average above 3 percent over the next 10 years and that the United States will not have another recession. It could turn out that there will not be a major interest rate spike. But if one looks at these revised estimates, it seems much more likely that they will prove to be too optimistic rather than too pessimistic.
There is little reason to expect private saving to increase, giving its secular downward trend. There is also little reason to expect federal dissaving to decrease; perhaps it will even increase over the next decade, with the United States saving 1.3 percent of its income. The arithmetic can work out in various ways: One is to invest only 1.3 percent of income, with the associated consequences for productivity; the other, as has been done in the United States, is to borrow and invest. I submit that there is no long-run healthy path that is consistent with anything like a national saving rate of 1.3 percent of GNP.
Why do I say this? First, there is the question of sustainability. Are the current level of investment, the current quantity of borrowing, and the current account deficit sustainable? Anyone who proclaims on this question with great confidence is making a mistake. We do know that the US current account deficit is running at 5 percent of GNP, which has traditionally been a danger point. We also know that the dollar is uniquely a reserve currency, and, as has been recently pointed out, financial markets have become ever more flexible.
But there are classic tests for knowing when a rising current account deficit should be of greater concern and when it should be of lesser concern. When the current account deficit is rising to finance investment, that is a matter of less concern; when it is rising to finance consumption, that is a matter of more concern. The US current account deficit is rising to finance consumption. When the current account deficit is rising and investment is becoming increasingly concentrated in the traded goods sector, that is a reason for lack of concern; when investment is becoming increasingly concentrated in the nontraded goods sector and away from the capacity to produce for the international sector, that is a cause for greater concern. And investment in the United States is tilting ever more toward real estate. When long-term private investors are financing a current account deficit, that is a matter of less concern; when short-term investors are financing that current account deficit, that is a matter of greater concern. When official providers of finance account for an increasing share of the current account deficit, that is a matter of particular concern. They now account for nearly half of the financing of the US current account deficit. In short, no one can know the answer to the sustainability question, and the United States is in many ways sui generis, but all of the traditional warning indicators point to the fact that the current 5 percent of GNP current account deficit has all the hallmarks of a particularly serious situation.
But let us imagine that it were possible at this level of national saving to continue to borrow on this substantial scale to finance investment and that this situation were sustainable into the indefinite future, a situation of which none of us can be confident. Is it healthy for the US economy or for the global system? I would suggest not for three reasons.
First, it’s not our capital. The saving rate is what reflects the accumulation of wealth by Americans, and if we are not saving, regardless of how much investment we finance, the returns from that investment will not be available for the United States. Now Pete Peterson would argue—and I think he is right—that the saving rate that is appropriate today in the United States is one that is substantially greater than the rate that was appropriate 10, 20, or 30 years ago because the baby boom generation will begin retiring in 2011. Even if he is not right in that supposition, it is hard to see why wealth and savings should be lower than they have been at any point historically, or how they are going to get better automatically.
Second, a situation of substantial dependence on foreign capital and a substantial current account and trade deficit, when it has taken place in the United States, has historically at every point been associated with a substantial increase in protectionist pressure. Whether the protectionist pressure derives from the relative level of the dollar or the relative level of the trade deficit is a question that econometricians and politicians can debate—I don’t think the data really permit a distinction—but it is hard to believe that the protectionist pressures would be as serious as they are if the United States did not have a trade deficit of the current magnitude, and it is hard to believe that trade deficits of the current magnitude will not lead to increases in protectionist pressure in the future.
The third troubling aspect of this dependence on foreign capital is its geopolitical significance. Here it is most difficult to speak with definitiveness. There is surely something odd about the world’s greatest power being the world’s greatest debtor. In order to finance prevailing levels of consumption and investment, must the United States be as dependent as it is on the discretionary acts of what are inevitably political entities in other countries? It is true and can be argued forcefully that the incentive for Japan or China to dump treasury bills at a rapid rate is not very strong, given the consequences that it would have for their own economies. That is a powerful argument, and it is a reason a prudent person would avoid immediate concern. But it surely cannot be prudent for us as a country to rely on a kind of balance of financial terror to hold back reserve sales that would threaten our stability.
Notice that what I have said leaves open the question of how the exchange rate will adjust or should adjust. Those outcomes, probably not within the control of policymakers, will influence whether low national saving affects US dependence on foreign capital or its level of investment. My point is that there is no resolution with the current low rate of national saving that speaks to healthy growth, strength, and preparation for an aging society.
What is to be done domestically, and what does this mean for the rest of the world? The most potent and reliable way to increase national savings, in the judgment of almost every economist who has looked at the question, is to reduce federal dissavings, which experience suggests translates, almost dollar for dollar, into increased national savings. Budget deficits and debt finance are not an alternative way of financing government expenditures; they are a way of deferring tax increases or subsequent expenditure cuts at substantial cost in interest and ultimately in the allocation of national resources. At the same time, the downward trend in American private saving must surely be a cause for concern. A prudent government and prudent employers will seek to ensure that Americans planning for retirement have a sense of how long that retirement will be in a world of rising life expectancies, how expensive that retirement may be in a world of rising medical care costs, and how uncertain the financing of that retirement may be in a world where financial markets provide uncertain returns.
Increasing US national saving, then, is crucial to the hope that the United States will be able to maintain the strong economic performance of the last 15 years. What are its implications for the global economy? It should first be said that an increase in the US national saving rate represents a decline in global demand. An increase in US national saving, with an associated reduction in the US current account deficit, means that, after many years of US demand exceeding and growing more rapidly than US output supply, a period is ahead of us when the US output will increase more rapidly than demand. The resulting gap between global demand and supply comes at a time when almost no other part of the world appears capable of increasing demand, even at a rate that is commensurate with the increase in supply.
Notice that exchange rate manipulation and adjustments, even if they could be controlled or willed by policymakers, do not address the global demand-supply imbalance created by an increase in US saving. They serve simply to redistribute it from one country to another. For example, the scenario advocated by many, in which increased US global saving is associated with a depreciation in the dollar, may offset the adverse demand impact in the United States of increased saving by switching the demand for expenditure from foreign goods to American goods, but only at the cost of increasing the demand-supply imbalance abroad. A healthy global adjustment process requires a healthy US economy, which requires increased national saving, which in turn requires measures that replace the demand that is lost from increased national saving. Indeed, even with the assumption of constant US national savings, the global economy today appears to be suffering more from the deflationary pressures associated with too little demand than the inflationary pressures associated with too much demand.
In times past, it has been common to suggest that increases in US national saving can be accommodated and their impact on demand can be offset through substantial increases and easing of US monetary policy. This does not appear in prospect at the current time, and that makes increases in demand abroad an imperative. There are policy challenges in Europe, where policymakers too often confuse the supply and demand elements of economic policy. Measures that stimulate aggregate supply and the economy’s capacity to produce are very desirable in their own right, but unless they are associated with increases in investment demand, they actually worsen the problem of imbalances. For Europe, a combination of structural reforms that increase incentives for investment demand and a willingness to use expansionary monetary policies seems appropriate. Similar measures seem appropriate in Japan, where the challenge of assuring adequate aggregate demand has yet to be fully met without reliance on the external sector, even after a very long time. As for the emerging market nations of Asia, the central question is not the behavior of the exchange rate but rather the full constellation of economic policies that bear on the creation of supply versus the creation of demand. For the last decade, the United States has enjoyed a period of import-led growth, which has enabled others to enjoy export-led growth or to enjoy modest performance fueled by exports. That import-led growth is neither likely nor desirable to continue, which means the growth plans of others that rely on export-led growth will need to be adjusted in the years ahead.
Again, these problems are not small. The country that is more at the center of the economic system than it has been at any point since the immediate postwar period is saving less than it ever has. Its adjustment is important and will have far-reaching implications. If it happens suddenly and without control, the consequences are likely to be very serious for the cyclical performance of the United States and the global economy. Ultimately, the consequences of these adjustments being mismanaged are likely to be profound for the global integration process that we all regard as so very important.