A year ago this time, in early October of 2008, the world was on the edge of a financial and economic abyss. Those very close to the events in the financial sector were terrified. The world at large had not yet fully comprehended the magnitude of the disaster. The October 2008 World Economic Outlook had predicted that global GDP growth would have attained 2.7 percent in 2008 and would be 1.9 percent in 2009. Compare that to the 2.1 percent realized in 2008 and the most recent projection of -2.3 percent for 2009. Projections have been revised upward over the last few weeks, but the loss of output in 2009 will still be much greater than what was projected a year ago. The real point, however, is that it could have been much worse. What happened on Wall Street in September of 2008 was the financial equivalent of the Cuban missile crisis of 1962. We came very close to a complete meltdown … as the world had come very close to nuclear war in 1962. But the meltdown did not take place—a very vigorous policy response in the major economies and concerted action by the major central banks forestalled a much worse disaster.1

A year ago this time, in early October of 2008, the world was on the edge of a financial and economic abyss. Those very close to the events in the financial sector were terrified. The world at large had not yet fully comprehended the magnitude of the disaster. The October 2008 World Economic Outlook had predicted that global GDP growth would have attained 2.7 percent in 2008 and would be 1.9 percent in 2009. Compare that to the 2.1 percent realized in 2008 and the most recent projection of -2.3 percent for 2009. Projections have been revised upward over the last few weeks, but the loss of output in 2009 will still be much greater than what was projected a year ago. The real point, however, is that it could have been much worse. What happened on Wall Street in September of 2008 was the financial equivalent of the Cuban missile crisis of 1962. We came very close to a complete meltdown … as the world had come very close to nuclear war in 1962. But the meltdown did not take place—a very vigorous policy response in the major economies and concerted action by the major central banks forestalled a much worse disaster.1

The topic today, however, is not the past but the future. Will the world return to the kind of growth we had in the 2002-07 period—averaging about 3.2 percent at market prices and 4.3 percent at purchasing power parity prices? Those five years were a period of very rapid growth, close in pace to the growth experienced during the post-World War II reconstruction period. Or are we likely to experience much slower growth for a number of years? I am not talking about 2010—I am trying to look forward to the next five to ten years. What kind of economic growth can the world expect over the next decade? Predictions diverge widely, in line with the disagreements in the economic literature on trend reversion of GDP.2 Until recently many observers have stressed that we should not expect world growth to resume at precrisis pace in the near future. The majority view has been that not only will we not experience the kind of accelerated post-depression growth observed, for example, in the U.S. economy in the 1930s, but there will be several years of below-trend growth after the great crisis of 2008-09. The process of deleveraging will take time. The U.S. consumer will no longer be able to play the locomotive role of the past. A more-regulated financial sector facing higher capital requirements will be unable to provide financing as easily as in the past. These are some of the arguments most often cited to express caution and a subdued view on future growth. More recently, as the data point to a stronger recovery of output in the second and third quarters of 2009 than what was expected in most countries, many observers now see a “good snapback” in the second half of 2009, but then a “slow crawl” in the coming years, at least in the advanced economies.3 In terms of the literature on trend reversion, most observers probably agree that the crisis of 2008-09 qualifies as a candidate to be one of the “structural breaks,” different from more minor recessions for which there is some evidence of trend reversion, at least in some countries. What I would like to do today is to share my perspective on this question of medium-term growth prospects. Clearly a great deal of modesty is in order. As Michael Spence, Chairman of the World Bank-sponsored Commission on Growth and Development in which I participated, has written in the postscript report that will be discussed later today, here in Istanbul: “the crisis was a humbling experience for anyone who seeks to understand and explain the world economy.” 4 Let me stress that not everyone got it equally wrong. There were warnings about excess leverage, unsustainable housing prices, bubble features in many classes of asset markets, and global imbalances. But very few predicted the nature and extent of the crisis. When it comes to predictions, modesty is in order for economists. All I aspire to do is to review some of the arguments and make some tentative suggestions. Let me also add, here at the outset, that GDP growth is a very imperfect measure of human progress, as has most recently been explained again, and with updated detail, by a commission led by Joseph Stiglitz and Jean-Paul Fitoussi. The fact that this lecture focuses on the traditional GDP growth measures should not be interpreted as implying that GDP is a sufficient measure of economic and social progress. That very important issue is not, however, discussed in this lecture. Another very important issue not discussed is the issue of climate change. If there is no global collective action to reduce carbon emissions, world growth over the next decades will increasingly be facing a climate challenge and climate constraint. When growth prospects over the next few decades are discussed, climate has to be part of the discussion. In this lecture, we focus on the next five to ten years. While preventive action is needed starting in this coming decade, and while the relationship between climate policy and growth is already important, I have not addressed it in this lecture. A longer-term discussion of global growth prospects would have to include serious analysis of climate-related issues, including the need to drastically reduce the carbon intensity of GDP.

The Recent Debate about Growth HAs Focused Only On Demand-Side Factors

The worldwide economic debate over the last 18 months has been conducted almost entirely from an aggregate demand perspective. This is of course understandable, as the financial crisis of 2008 led to an unprecedented worldwide collapse of private demand. High oil and commodity prices did add a supply-side element to the making of the crisis in mid-2008, but the dominant initial cause of the downturn was the decline in demand in the advanced economies caused by the steep fall in asset prices and the credit crunch in the United States and Europe. The second wave of demand declines quickly came through the trade channel, amplifying the first and reaching the emerging market and developing economies. Steep declines in aggregate demand led to contractions in the demand for exports, with multiplier effects throughout supply chains and across borders. The most export-oriented economies, except those with large current account surpluses and reserves acting as buffers, suffered the most. This experience of the crisis with its devastating effects should not make us forget, however, that considered from the supply side, the recent history of the world economy had been one of rapid and indeed accelerating growth. As mentioned before, growth during the 2073-07 five-year period averaged to one of the highest levels ever experienced.

What Are The KEY Sources OF GRowth From The Supply SIde?

It will be useful to briefly recall the main longer-term drivers of growth in potential output. First there is technical progress or the outward shift of the production possibility frontier due to innovation and new knowledge, mostly taking place in the most-advanced economies operating close to the knowledge frontier. Technical progress had been very rapid, particularly in the United States during the second half of the 1990s. There has been some slowing of the pace of productivity growth in the United States in the early years of this decade, probably reflecting less investment, but the underlying shift of the technological frontier is unlikely to slow down as new advances in knowledge are taking place. These range from new waves of progress in information and communication technology and digital production techniques to advances in biotechnology and energy efficiency.

Second, there is the speed of diffusion of knowledge and technology. There is little doubt that globalization—which we can here define as the increase in trade, investment, and information flows—accelerated the diffusion of technology. Globalization reached a new dimension after the fall of the Berlin Wall and the much greater integration of China into the world economy. The emergence also of India as a significant participant in the world economy, after the early 1990s, has now added hundreds of millions of workers and consumers to a much more integrated and interdependent world economy. Over the past two decades old barriers disappeared, information traveled faster than ever, and new technologies could be diffused at unprecedented speed across the globe. Because of the enlarged world market, and because of the nature of new Internet-based modes of production, the potential for exploiting economies of scale has also increased in the case of many economic activities.

The availability of knowledge and new technology can generally translate into increased labor productivity and production only if there is sufficient investment to make it happen and if physical investment is complemented by appropriate human capital formation. The need to embody technology and traditional capital deepening make capital accumulation the third major source of supply-side growth. Human capital indicators are overall improving rapidly, although unevenly across the world. The progress in Chinese secondary and tertiary enrollment rates is particularly impressive. Moreover, the aggregate world savings and investment rate had recently been on an increasing trend primarily because of the increasing weight in the world economy of the East Asian high-saving nations, but with some oil exporters also a strong contributing factor. And this increase took place despite a low and declining American savings rate. If the savings rate in the United States had not declined from about 16 percent to about 13 percent—I am here referring to total gross savings as reported in the national accounts, not to the much lower household savings rate—the increase in world savings over the 2002-07 period would have been even more remarkable. In Asia, plentiful savings made “catch-up” growth all the more rapid, as the great potential for technological diffusion could be translated into unprecedented output expansion through high investment rates. New techniques came embodied in new capital goods, and the high Asian investment rates facilitated diffusion of knowledge. There is every reason to believe that this process will continue for many years, even if we should see a slow decline in the extraordinarily high Chinese savings rate. Most Asian countries as well as the oil exporters of the Gulf will continue to have high savings rates, and their weight in the world economy will increase. The Chinese savings rate will probably come down a little, but in relative terms it will remain huge. In the advanced countries, savings are unlikely to decline. They may even increase in the United States. So in the aggregate the world savings rate will probably continue to increase, at least moderately.

A rapidly shifting technological frontier, rapid worldwide diffusion of knowledge, and plenty of investment, including in human capital, translate into rapid growth of potential output worldwide. This process can produce particularly good results in terms of output, while there is still a lot of underutilized labor in the rural sectors of emerging and developing economies, constituting a fourth source of supply-side growth. Even without appealing to a shifting production function or to technology with economies of scale, it would still be possible to view part of the growth of potential output in many emerging market economies through the lens of a classical Lewis growth model where there are no diminishing returns to investment, as more rural labor can be drawn into the modern sector with low opportunity cost. Economies of scale and technical progress only add to the prospect of accelerating growth in potential output.

There are, therefore, powerful factors supportive of growth in potential output in the world economy, and there is no reason for any one of the main elements causing this trend to weaken in the medium term, if we define the medium term to be about a decade or so. The accumulation of knowledge does not appear to slow down. On the contrary, all indications are that further impressive advances can be expected not only in informatics and digital production, but also in other sectors such as biotechnology, and the field of potential applications is wide. The rate of diffusion of knowledge will also continue to be rapid, as there is continued progress in education and language skills, as well as communication technology. With the increasing economic weight of the nations with high savings rates, the overall potential investment rate will be high. Finally, for the next decade at least, there are still large amounts of “reserve labor” to be deployed in the higher-productivity modern sectors in the developing economies, or indeed through migration, in the advanced economies. Supply-side growth models for the world economy as a whole factoring in all these elements will tend to produce an average annual global growth in potential output close to 4 percent for the coming decade, with GDP measured at market prices, close to or a little higher than what was achieved in the best years of the past two decades.


The problem is, of course, that the death of the business cycle and of Keynesian economics had been announced very prematurely. The presidential address of Robert Lucas to the American Economic Association in 2003 is often quoted as such an announcement, although, to be fair, he announced the end of depressions—not the end of recessions.5 Be that as it may, macroeconomics is not just about aggregate supply. Supply still does not necessarily create its own demand. We have been living through massive output losses because of problems unrelated to the factors determining long-run supply, which I tried to summarize above. The massive decline in asset prices between the summer of 2008 and the spring of 2009, the huge rise in uncertainty, the collapse of “animal spirits,” and the disorganization of the financial sector with accompanying declines in credit led to a worldwide fall in private consumption and investment demand and to the emergence of an output gap of a size the world had not experienced in decades.

“Ex post” supply of course always equals demand, and the adjustment had to take place through a contraction of output. Public policies have limited the decline in output by substituting public demand for private demand on an unprecedented scale. The expansion of fiscal deficits in G-20 countries is likely to average more than 5 percentage points of GDP in 2009, with about 2 percentage points due to discretionary fiscal stimulus. A “hyper-Keynesian” combination of fiscal and monetary policy response helped the world avoid a much bigger disaster. It is true that ascribing recent output increases primarily to fiscal stimulus spending per se is difficult, particularly in the United States, because only part of the stimulus has actually been spent. Monetary policy, including the unconventional part, and the direct interventions in the financial sector also deserve much credit. Moreover, the effect of the decisive and large overall policy response on “animal spirits” should not be underestimated. The “hyper-Keynesian” message was strong, particularly in the United States and China, but also in large parts of Europe and Japan, and it affected animal spirits. The London meeting of the G-20 and the decisions to triple the lending capacity of the IMF, as well as the measures the IMF itself took to expedite lending and make financing available through its new Flexible Credit Line, contributed to this reversal of animal spirits also in emerging markets. The fall and winter panic was overcome, and that, in itself, was probably as important as the actual amounts of government spending or the level of interest rates. Some parts of Europe appeared not to have responded as vigorously, but taking into account the strong automatic stabilizers existing in Europe, the fiscal impulse was actually quite strong even in countries openly worried about the hyper-Keynesian nature of the response. To some degree the countries having done a little less are also enjoying a bit of a free ride from the demand spillovers coming from the countries having done a lot.


There is, now, the issue of when it will be appropriate, in aggregate terms, to start to reduce public demand and wind down the fiscal stimulus as well as to start thinking about some tightening in monetary policy. I think the message from the G-20 meeting in Pittsburgh is, overall, an appropriate one. It is not a message that fiscal stimulus will now start to be withdrawn. The indicators of recovery are far too weak at this point. Unemployment worldwide, in particular, is still increasing. The appropriate interpretation of the message is that policymakers, worldwide, are preparing exit strategies and will be ready to tighten fiscal and monetary policy in appropriate doses if and when economic indicators warrant such action. Such a message is needed, both because it is what should happen, and because here again it is what is needed from the point of view of animal spirits. There is a justified worry that some of the fiscal expansion could be hard to rein in and that there could be pressures in the government debt markets of the largest economies leading to substantial increases in interest rates that would slow down or kill the recovery. A strong message that policymakers are aware of this danger and are, already now, working on preparing exit strategies is a positive one for animal spirits and will be supportive of the recovery, if it is credible. The Istanbul meetings of the IMF and the World Bank can reinforce this message and make it more credible by showing common resolve.

I would like to add two points in this context. The first one is that the dosage and timing of the policy response remains very important and perhaps more difficult than in the winter of 2009-Then, the danger of a world depression was so great that it was necessary to mount essentially as large as possible a response, as quickly as the legislative processes allowed. Now, fine tuning becomes more important, in terms both of timing and of amounts. The second point relates to the inflation unemployment trade-off. I believe, with Akerlof and Schiller,6 that there is indeed such a trade-off, even in the medium term, although it is not a very stable trade-off. The fine tuning of the exit strategies will have to be based on political choices regarding that trade-off. The inflation targets governments and central banks choose should depend on the amount of unemployment projected over the next two to three years. A totally necessary commitment to low inflation need not and should not imply that the inflation target chosen over a two- to three-year period should not at all depend on where we are with regard to employment. Long-run steady-state inflation targets are a different matter. It may indeed be desirable that they be fixed more or less forever … but that should not be the case in the immediate aftermath of a huge crisis.


Let me now turn from these remarks on global aggregates, to a brief analysis of problems relating to the structure of global demand. The stylized story has been that U.S. consumption demand was the locomotive of world demand, providing, in particular, Asia, and there, mainly China, with the necessary outlet for export-led growth, with both Asian capacity creation and some of American consumption financed through a huge Asian savings rate. A healthy recovery implies that American consumers rebuild their balance sheets, and the worry is that the implied higher American private savings rate and reduction of the U.S. current account deficit would mean that the American consumer could no longer be the “driver” of global growth. It is feared that such an interruption of what, for a long time, has been viewed as the growth engine of the world economy would lead to a lasting slowdown of world growth. In the absence of a strong expansion of consumption demand in China and other surplus countries, so the stylized story goes, there would have to be either a return to a large U.S. current account deficit and, again, a low U.S. savings rate, or lower world growth as a whole, as output has to adjust to insufficient global effective demand. As the need for deleveraging and concern about the deteriorating U.S. net debt position will not allow a return to the old global imbalances, it is argued that the only way the world can return to a sufficient amount of effective demand to allow rapid growth is through a change in the export-led growth model of China and to a lesser degree of Germany, with greater domestic consumption in these surplus countries leading to a lasting unwinding of the great global imbalances. Fred Bergsten and Arvind Subramanian put it as follows in a recent opinion piece in the Financial Times: “The U.S. strategy on this issue is not, at least for the moment, consistent with strategies elsewhere. Put starkly, [White House National Economics Council Director Larry] Summers stated that China can no longer behave like China because the U.S. intends to behave much more like China. The world economy cannot have two, or even one-and-a-half, Chinese growth strategies from its most important two economies.” 7

There is no question that some rebalancing of global demand would be beneficial for world growth over the next few years. The deleveraging process in the United States that is necessary to overcome what has been called a “balance sheet recession” will require a higher household savings rate. Moreover the U.S. net debt position, while still modest, has been on a rising trend, and a continuation of large U.S. current account deficits would no doubt create new vulnerabilities for the United States and the world economy. As all the major countries in the world consider their fiscal and monetary policies for the next two years, it is highly desirable that they choose these policies in a manner that would facilitate the unwinding of the global imbalances. In fact, it seems that in 2009, the sum of the absolute values of the current account surpluses and deficits, ignoring signs, of the G-20 economies will be close to half of what it was in 2008. But this sum, which we can take to measure global imbalances, could rise again in 2010 and beyond. If the macroeconomic policy mix in the United States over the coming years were to be more expansionary than the policy mix in China, Japan, and Germany, then global imbalances will rise again. International policy coordination explicitly targeting these imbalances is desirable, and the decisions taken at the Pittsburgh summit constitute a major, welcome step in formalizing a process that should be helpful in that context. If the surplus countries expand by more than the deficit countries, we can enjoy lower global imbalances and therefore greater overall medium-term stability with higher aggregate world growth. If, on the other hand, the burden of adjustment is put strongly on the deficit countries alone, then it will be possible to reach equilibrium only at lower aggregate world output levels.

I do not disagree with this line of argument, but I do think that one should not exaggerate the extent to which rapid expansion of U.S. consumer spending is a sine qua non for world growth. Let me make a few observations in this context.

First, it is something of an exaggeration to say that the American consumer has been the “driver” of world growth, in any long-run sense. The real medium-term drivers of world growth have been high rates of technical progress, rapid diffusion of technology, high Asian investment rates, the exploitation of economies of scale, and the availability of labor and resources. On the demand side, U.S. private consumption amounted to about 17 percent of total world demand in 2008. It grew at a rate of about 3 percent a year in the rapid-growth period preceding the crisis. Suppose that it will grow at only about 2 percent over the coming five years. Such a 1 percentage point decline, compared to the precrisis years, by itself, and holding everything else constant, would mean a 0.17 percentage point decline in the growth of aggregate world demand. If we instead project U.S. consumption to grow at only 1.5 percent a year over the next decade, the immediate decline in total world demand would be about a quarter of a percentage point. This would be significant, but hardly cataclysmic.

My second point is that we should look at the overall structure of current account deficits and surpluses, not just at the United States and China. In that context it is not clear whether focusing on the German surplus makes much sense. Germany is part of the euro zone, and Spain, another part of the euro zone, has had a huge deficit. The euro zone as a whole has not had a significant surplus or deficit. Maybe Texas, in the United States, has had a large surplus and California has had a deficit, but we do not look at such numbers. It would seem to make more sense to talk of the euro zone rather than individual countries in it, when discussing global imbalances. It is true that the euro zone is not a federal country with a common fiscal policy, but it is now both a single market and a single monetary zone. From the point of view of the link between current accounts and the euro/dollar exchange rate, for example, it is the aggregate savings-investment balance of the euro zone that matters. More importantly, some of the rebalancing could and should take place in a way that involves the middle- and lower-income countries. Let us think stylistically of six economic “regions” in the world economy: the United States, the euro zone, other advanced countries, China, the Gulf oil producers, and the other emerging and developing countries. The precrisis Chinese surplus of about 10 percent of Chinese GDP has been largely seen as the counterpart of the U.S. deficit, and it has indeed contributed significantly to the financing of the U.S. deficit.

For the world to have a desirable structure of current accounts, by how much does China have to lower its surplus in the next few years? I would like to suggest that rather than trying to lower the surplus drastically, which could have a significantly negative impact on its growth, China could redirect a significant portion of its surplus savings towards developing countries, allowing them as a group to run somewhat higher current account deficits and finance somewhat more investment. I have done the rough arithmetic. Taking the 2010-15 period, we could have a Chinese current account surplus averaging about 6 to 7 percent of GDP, a U.S. deficit of about 2 to 3 percent of GDP, and an “other emerging and developing countries” deficit of about 3 to 4 percent of their GDP. Arithmetically, this would be consistent with a surplus of about 10 percent of GDP for the Gulf oil exporters (this implicitly assumes quite high oil prices), and a nearly balanced aggregate current account in the euro zone and other advanced countries taken as a whole. Such a scenario could constitute a good transition phase during which the world avoids very sharp swings in growth strategies and macroeconomic balances. The United States would not have to eliminate its current account deficit. China could continue with a somewhat more moderate version of its high savings, export-led growth strategy and manage the transition to a more consumer-driven growth process gradually rather than very quickly. A greater amount of savings would flow to the poorer countries, allowing them to increase their investment and rhythm of development. The average return on those investments in the developing countries would surely be higher than what China can earn in U.S. Treasury bills. The scenario involves gradual change, not sudden, large, and potentially disruptive change, in the structure of world savings and investment.

A key feature of such a scenario that I consider desirable is the larger net capital inflow into the developing countries. As policy frameworks in these countries have improved, it is desirable and should be natural for them to be net capital importers rather than run balanced or surplus current accounts, without this leading to the types of balance of payments crises that these countries experienced in the 1980s and 1990s. And given the very high Chinese savings rate, some of these flows could and should come, directly or indirectly, from China. We have seen the beginning of such a new structure of capital flows over the last two years, including bilateral agreements between China and some Latin American countries to use their own currencies in part of their trade. To reduce and pool risk, some of the suggested capital flows could also be intermediated by the multilateral development banks, which should receive capital augmentations. Some of the excess Chinese savings could contribute to “pooled reserves” for the developing countries, including in the form of precautionary finance at the IMF.8

Such a global demand management scenario would allow a gentle rather than an abrupt unwinding of the global imbalances and a gradual shift in growth strategies rather than a large sudden change, which could be disruptive. If it could be realized, the problem of global imbalances would not be the key major constraint on medium-term world growth. Current accounts are of course not policy instruments, but the outcome of public and private sector spending decisions. For the scenario I have outlined to be realized, coordination of macroeconomic policies and a more active role of the international financial institutions in helping channel resources to the developing countries and in helping them manage the risks of a larger inflow of capital would be necessary. Fortunately, that is exactly what the Pittsburgh summit decisions should make possible.


While current account imbalances may evolve in a direction compatible with rapid worldwide growth, what about “internal” savings and investment balances in key economies? One of the most dramatic and unsettling economic statistics in recent times is that a full two-thirds of all economic gains in the United States during the rapid precrisis growth years accrued to only 1 percent of the population.9 In China, the share of labor income in GDP has declined to only 40 percent! It may well be that it is income-distribution-related internal imbalances that may be a key short-to medium-term demand-side threat to sustained rapid growth. But how come, then, there was such rapid precrisis growth despite these extreme income concentrations at the top? One answer to this question is that it was the unsustainable internal debt accumulated by U.S. households and the illusion of wealth due to the asset price bubbles that sustained consumption in the United States and other countries, despite the worsening income distribution and the lack of real income growth for most households. In China, it was both asset prices and export demand that played a role in compensating for the negative effect of internal income concentration on consumer demand. I do believe the link between income distribution and effective demand is underresearched and not sufficiently debated, particularly when compared to the constant attention given to the global imbalances issue. It is true that in the long run, the structure of supply can adjust to any particular income distribution. In the short to medium term, however, sectoral imbalances reflecting sharp changes in income distribution can lead to macroeconomic problems. It may well be that a continued worsening of income distribution in countries such as the United States and China could be as significant a demand-side threat to global growth as the U.S. and Chinese current accounts deficits and surpluses.


I would like to conclude from all this that there is probably the potential for very rapid growth in the world economy, by historical standards, over the coming decade, because of strong supply-side factors. Whether this growth can be realized does, however, depend on demand-side management both at the national level and through improved global macroeconomic policy coordination. Macroeconomics remains highly relevant, and proactive policies, national and international, can provide large benefits. The world has actually experienced an overall highly successful macroeconomic policy response to a crisis largely due to a previous excessive confidence in self-regulating markets and a neglect of the need for careful regulation and macroeconomic management. Let us not make the mistake of thinking that because output indicators are now improving, the policy response was unnecessary or that, in the years to come, macroeconomic management can be on some kind of autopilot, rather than being responsive to ever-changing circumstances.10