Is Capitalism Too Productive?

Paul Krugman. Foreign Affairs. Volume 76, Issue 5. September/October 1997.

One problem capitalism does not suffer from is being too productive for its own good. Imagining problems that do not exist has real costs. To speak to European advocates of the global glut theory is to be struck by their fatalism: they seem to have given up on the idea of making the European economy grow. And this fatalism seems almost completely ineffectual. Meanwhile, American global glutters seem to spend half their time complaining that nothing good can be done unless the country gives up the idea of a balanced budget, and the other half converting their doctrine into a new justification for good old-fashioned protectionism. It is a bit funny, but also quite sad: those who preach the doctrine of global glut are tilting at windmills, when there are some real monsters out there that need slaying.

Draining the Global Glut

The great majority of those who voted for Lionel Jospin’s Socialists in the French elections earlier this year were surely voting against, not for: they were protesting high unemployment and the aloof austerity of Alain Juppe’s conservative government, not endorsing the specifics of the oppositions program. Nonetheless, Jospin’s elevation to prime minister is a remarkable event. Sooner than anyone might have expected, a radical economic doctrine has emerged from obscurity to become, in principle at least, the official ideology of a major advanced nation’s government.

Let me give that economic doctrine a name, and call it global glut. It may be summarized as the view that capitalism is too productive for its own good-that thanks to rapid technological progress and the spread of industrialization to newly emerging economies, the ability to do work has expanded faster than the amount of work to be done. In its milder forms, the global glut doctrine involves the belief that policies should aim at increasing demand rather than supply; thus its American advocates have opposed efforts to eliminate the budget deficit or increase national savings, claiming that such efforts will actually reduce the economy’s growth. In its more extreme forms, the doctrine calls for outright reductions in the economy’s capacity, in particular through “work-sharing” schemes that reduce the length of the workweek. And it was this extreme form that was a central plank of Jospin’s program: he called for a mandated reduction in France’s workweek from 39 to 35 hours.

Heterodox doctrines, in economics and elsewhere, often fail to get adequately discussed in their formative stages. Both the intellectual and the political establishment tend to regard them as unworthy of notice. Meanwhile, those doctrines can seem compelling to large numbers of people, some of whom may have considerable political clout, financial resources, or both. By the time it becomes apparent that such influential ideas-say, supply-side economics-demand serious attention after all, reasoned argument has become very difficult. People have become invested emotionally, politically, and financially in the doctrine, careers and even institutions have been built on it, and its proponents can no longer allow themselves to contemplate the possibility that they have taken a wrong turn.

The doctrine of global glut has probably not yet reached that point. Even Jospin’s Socialists could still quietly drop work-sharing from their program without much of a backlash, provided they deliver in other ways. And with only a few exceptions, American intellectuals and politicians who have been flirting with some version of the doctrine can still modify their views without too much embarrassment. But now is the time to discuss the doctrine seriously, before it becomes a dogma impervious to logic and evidence.

Let me not be too coy about my verdict. The idea of a global glut does not stand up to examination; it is conceptually confused, and its advocates seem oddly unaware of even very basic facts. But it is evidently a doctrine that many intelligent people find compelling; it is therefore important both to lay out the case against a global glut and to try to understand why the doctrine is so appealing, despite its intellectual vulnerability.

Too Much of a Good Thing?

Concerns that capitalism might be too productive for the good of its workers, perhaps even for the good of the capitalists themselves, have been with us almost since the beginning of the Industrial Revolution. The slump that followed the end of the Napoleonic Wars, and the brutal displacement of handweavers by the power loom, forced even classical economists to consider the possibility that the introduction of improved technology might have some adverse effects. Thus David Ricardo’s The Principles of Political Economy and Taxation (1817), which among other things first demonstrated rigorously the benefits of free trade, also included a chapter entitled “On Machinery” about the possibility of technological unemployment.

More or less modern concerns about excessive productivity emerged in the 1930s and 1940s. It was natural that some observers would tie the lack of jobs during the Great Depression to the widespread introduction of mass production techniques in the 1920s. Keynesian economics, which legitimized concerns about overall inadequacy of demand, provided an intellectual framework for the idea. During the late 193os and early 1940s, many economists subscribed to a doctrine-often referred to as “secular stagnation”-that was quite similar to global glut. Secular stagnationists pointed out that well-off families tended to save a higher fraction of their income than poor ones; thus, they argued, per capita consumer spending would not keep pace with growth in per capita income. The economy could therefore maintain full employment only if investment spending grew much more rapidly than income, which seemed unlikely. So the secular stagnationists predicted a return to depression conditions once World War II was over, and a tendency toward ever-growing unemployment rates over time.

Postwar developments in economic theory, together with the experience of the postwar boom, pretty much eliminated secular stagnationism among professional economists. The doctrine did live on in more popular writings, often bound up with concerns about the impact of “automation,” the supposedly imminent widespread replacement of workers by machines. But by and large, concerns that capitalism might simply be too productive remained dormant from the early 1950s to the 1980s.

The rise of the current doctrine of global glut can be traced to three main changes. First, mass unemployment has reemerged in Western Europe, although not in the United States. Most economists and business leaders blame the long-term rise in European unemployment on “Eurosclerosis,” a condition brought on by excessive taxation and regulation. But many of Europe’s trade union leaders and some of its political leaders have concluded instead that there simply is not enough work to go around, and that this problem is becoming ever more acute as labor productivity rises. Thus proposals similar to Jospin’s work-sharing scheme have been common on the European scene for a decade or more. Demands for work-sharing became particularly intense when Europe went into an economic slump in the early 199os; the concept even received a somewhat blurry endorsement in a 1993 European Commission white paper, Growth, Competitiveness, Employment.

Second, there is a widespread perception that productivity growth in the advanced countries, especially in the United States, has accelerated. What turned out to be a temporary period of jobless recovery” in 199192-that is, of growing GDP but rising unemployment-apparently had a permanent impact on some economic writers, convincing them that the traditional link between growth and jobs had been severed.

Finally, the spread of industry to newly emerging economies and the rapid growth in exports from those economies has fed the sense that global productive capacity is growing headlong, far too fast for demand to keep up. Indeed, some global glut adherents, notably William Greider, not only believe that growth will lag behind capacity, but warn that the growing gap between potential supply and demand will provoke an economic crisis like that of the 1930s, with output plunging. The doctrine of global glut, then, is a response to some real changes in the world economy.

Is There Really a Global Glut?

To get some perspective on the global glut, it may be useful to focus on three propositions that are crucial to the doctrine, one about supply and two about demand (because we need to ask somewhat different questions about demand in advanced and in developing countries). The propositions are:

  1. Global productive capacity is growing at an exceptional, perhaps unprecedented, rate.
  2. Demand in advanced countries cannot keep up with the growth in potential supply.
  3. The growth of newly emerging economies will contribute much more to global supply than to global demand. Are these propositions true about the present, or likely to be true about the future?

Productive Capacity

Popular economic discussion tends to be driven by impressions and anecdotes, rather than statistics. There is something to be said for this attitude, especially in times of rapid change: statistics designed to track the last generations economy can easily miss crucial aspects of what is happening right now. On the other hand, anecdotes tend to emphasize the exciting and different and to miss the continuous. The fundamental things stay the same as time goes by, but you may not remember this if you focus too much on the interesting anecdotes.

The belief of commentators like Greider that there is huge growth in productive capacity seems to be driven largely by tales of individual industries that clearly do have a problem of overcapacity, notably automobiles. The problem with such stories is that in an unpredictable world, the emergence of excess capacity in particular industries that overestimated their market prospects is a normal hazard of business as usual-as is the emergence of capacity shortages in industries that underestimated demand. At times the balance is clear: at the bottom of a severe recession, or even after a more modest slump like 1990-91, there is a clear predominance of excess supply. But at the moment there are shortages of some things, such as office space in many cities, coffee beans, and skilled machinists, at the same time that there are excess supplies of others, like auto assembly plants. Which stories should we regard as defining, and which as incidental?

Another source of the belief in surging productive capacity is the impression of breakneck progress created by the triumphs of digital technology, where Moore’s Law-which says that the price of a given computation halves every i8 months-implies a state of permanent revolution. But there is more to production than computation. Even in manufacturing, reports of the workerless factory have repeatedly proved premature; the much-hyped arrival of industrial robotics in the 198os turned out to be no more than a modest addition to the arsenal of production techniques. Evidence of a productivity revolution is even harder to find outside the manufacturing sector. The paperless, secretaryless office, for example, is something that is always about to happen but somehow never does.

For what they are worth, official estimates of the overall rate of capacity growth in advanced countries are distinctly unimpressive. The Organization for Economic Cooperation and Development and the International Monetary Fund both estimate the rate of growth of potential GDP (the amount an economy can produce at normal rates of capacity utilization) in the advanced economies as a group at no more than 2 to 3 percent annually-about the same as in the previous zo years, and well below the growth rates of the 1950s and 1960s.

But can such estimates be trusted? Technological enthusiasts are quick to point to qualitative improvements that normal estimates of real GDP miss, like the convenience of automatic teller machines. On the other hand, such poorly measured quality improvements were arguably equally important in earlier decades; it is anybody’s guess whether the undermeasurement is greater now than before.

There is one point nobody would dispute: the stunningly rapid industrialization of some developing countries, mainly in Asia, has contributed to an acceleration in the growth of world capacity.

Averaging these dreary official estimates of potential growth with the impressive actual growth rates in Asia suggests that overall productive capacity in the world’s market economies is growing at something close to 4 percent per year. This is better than the 3 percent or so growth in the 1970s and 1980s, but somewhat less than in the 1950s and 196os. Perhaps the estimates are more wrong now than they were then, but the case for unusually high growth in productive capacity is weak at best, and there is no case for the more breathless claims about unprecedented expansion.

Demand in the Advanced Countries

Even if the growth in capacity is more modest than the hype would have it, the capacity will go unused unless consumers provide sufficient demand. Conventional economists do not see any problem here. After all, if the growing capacity of advanced countries is in fact used, income will rise in line with that capacity; and if income rises, why won’t consumer demand rise too? Somewhat oddly, none of the main proponents of the global glut doctrine offers any direct refutation of this standard view. Greider, Rifkin, and even Heilbroner (who should at least be aware of the conventional argument) seem to take it for granted that if growth in supply is too rapid it will outstrip growth in demand. Judis has pointed out that the 1930s did happen, proving that demand need not always equal supply, but that argument says nothing about whether growing capacity itself (as opposed to, say, monetary mismanagement) will provoke a slump.

Reading between the lines, however, it seems likely that the global glutters, like the secular stagnationists of yore, have in mind the idea that as income rises, consumers become saturated-that having bought all their necessities, they become reluctant to increase spending even if their income rises. This idea seems confirmed by ordinary experience: well-off families typically do save much more of their income than the poor. Some of the global glutters follow this logic a bit further, suggesting that consumer demand is also limited by the maldistribution of income within advanced countries, or at least within the United States. Struggling families would spend more if they could, but income gains go instead to the rich, who hoard income instead of spending it.

It is evidently a persuasive story. But it runs up against an awkward fact: consumer spending in the United States has kept up with rising income for a very long time. Real per capita income today is roughly triple what it was in the late 1940s, the heyday of secular stagnationism. And that income is substantially more unequally distributed. Yet the fraction of personal income that consumers save is actually lower now than it was then. Somehow or other, most Americans, including many of those in the upper echelons of the income distribution, have found ways to spend their income without becoming saturated.

If you think about it, this is not all that amazing. Most readers of Foreign Affairs surely know people with annual incomes of $300,ooo or more. Indeed, a fair number of readers probably meet that description themselves. In reality, how hard is it to find ways to spend that money? A really nice home, a second home or nice vacations, private colleges for the children, two good cars . . . Yet even if median family income in the United States grows at z percent per year, it will take a century before the median family has an income equivalent to $300,000 in today’s dollars.

It is true that people with above-average incomes tend to save more than people with lower incomes. But economists have long realized that this fact is mainly a sort of statistical optical illusion. In any one year, the class of people with high incomes includes a disproportionate number of people who are doing better than usual and saving for a rainy day, while the class of people with low incomes includes a disproportionate number of people who are doing worse than usual and drawing down their savings. But when people’s normal or typical (or, in the jargon of economists, “permanent”) income rises, as it does as the economy grows, their spending grows right along with it. In short, it is hard to see any justification for the belief that consumer demand in advanced countries cannot keep up with growing capacity.

The Third World

It is when we come to the role of newly industrializing economies that emotions and confusion run highest. Perhaps the best way to sort things through is to present the story as a Panglossian conventional economist might tell it, see how the global glut advocates differ with that story, and look at the relevant facts.

The agreed starting point is that there has been a rapid increase in the productivity of labor in some developing countries. Hundreds of millions of workers who were previously of little account in the global economy are now an important productive resource, adding to the world’s productive capacity.

The conventional economist sees no problem in this development. To be sure, the world’s productive capacity is increased, but higher productivity will mean higher incomes. And since one can hardly claim that consumers are already saturated in such poor nations, demand will increase along with supply. The newly industrializing economies, says conventional theory, will in general spend about as much as they earnor, what is the same thing, import as much as they export.

Actually, a conventional economist might well predict that emerging economies will in general run trade deficits. After all, one might expect these economies to attract inflows of investment from abroad. And a country that attracts a net inflow of capital must, by definition, be a country where domestic investment exceeds domestic savings, and which will therefore have spending in excess of its income. But a country that spends more than it earns must run a trade deficit, importing the difference. (More precisely, it would run a deficit on the current account, which includes trade in services and the income from past investments).

But how can this be? Won’t countries that have very low wage rates be extremely cost-competitive, and therefore run trade surpluses? Well, the conventional economist has a pat answer for that too: wages will rise in line with productivity, so that those countries won’t be all that cost-competitive after all.

To believers in an emerging global glut, this all seems naively optimistic. A representative statement of their position came in the scholar Alan Tonelson’s review in the June 15 New York Times Book Review of former Under Secretary of Commerce Jeffrey Garten’s The Big Ten: The Big Emerging Markets and How They Will Change Our Lives. Tonelson asserts that “consumer markets in these emerging markets are likely to stay small for decades … Two reasons stand out. First, largely because most of these countries have huge foreign debts to pay off, and therefore need to discourage consumption by their populations, they are pursuing export-led growth strategies. Second, if they don’t keep wages and purchasing power low, they will have difficulty attracting the foreign investment they require, both to service debt and to finance growth.” In short, emerging economies will not add nearly as much to global demand as they will to global supply.

This hypothesis seems clear enough. Yet if one rereads Tonelson’s remarks, or similar passages in Greider’s new book, and also does some research-say, by buying a copy of The Economist on the newsstand and opening to the “emerging market indicators” on the last pageone receives a bit of a shock. Tonelson seems to believe that newly industrializing economies are currently engaged in “export-led” growth-that is, that they are selling much more to advanced countries than they buy in return-and that it is naive to expect this pattern to change. Presumably, then, he believes that these economies are currently running large trade surpluses. Table 1, however, shows the trade and current account balances of Garten’s “big ten”: seven are running trade deficits, as is the group as a whole, and all but one are running current account deficits. More broadly, of the 25 emerging markets covered in The Economist, 17 are running trade deficits, 20 current account deficits. This is export-led growth?

Let me put it this way: if you read what the global glut advocates have to say, you might have the impression that right now the newly industrializing countries are running huge trade surpluses, and that it is a dubious theoretical prediction that at some future date they will begin to import as much as they export. But the truth is just the opposite: the emerging economies as a group are running trade deficits, and it is merely a speculative prediction on the part of the global glutters that at some future date they will start to run large trade surpluses. (And it is a speculative prediction that seems to have some conceptual problems too. Notice that Tonelson appears to say that these countries will run large surpluses because they must keep wages low to attract foreign investment. He seems unaware that a country that attracts a net inflow of foreign investment must, strictly as a matter of accounting, run a current account deficit.)

There is a similar inversion of the relationship between fact and hypothesis when we turn to the behavior of wages in newly industrializing economies. Perusing the global glut writers, you might have the impression that thus far wages in emerging economies have failed to rise along with productivity, and that assertions that they will rise in the future are purely optimistic speculation. In fact, those developing economies that have had rapid productivity growth in past decades have also, without exception, had rapid increases in wages. Japan, we should not forget, was once a low-wage country, and as recently as the early 1970s many Western critics accused it of deliberately keeping wages low in order to foster a trade surplus. Today hourly compensation in manufacturing in Japan is higher than in the United States. And as Table 2 shows, the original four Asian “tiger” economies have experienced huge wage increases over the past two decades. In other words, the evidence to date is that wages do in fact rise along with productivity, and it is the assertion by global glutters that they will not do so in the future that is pure speculation.

All this is fairly puzzling. The advocates of the global glut doctrine see themselves as more realistic than conventional economists; they see themselves as rejecting an abstract framework that gives excessively optimistic assurances in favor of a clear-eyed view of the world as it really is. Yet the data bear out the supposedly naive views of conventional economists, while the global glutters turn out to be speculative theorists who assert, without justification, that the future will be entirely different from the past and tresent.

There must be something peculiarly compelling about the idea of a global glut-so compelling that it persuades intelligent people not just to reject conventional economics but to overlook inconvenient facts.

A Failure of Imagination

There are probably many readers who, despite the previous discussion, still find it hard to shake the notion that there is now or soon will be a massive global oversupply of goods. After all, world productive capacity is increasing steadily-even if not quite as rapidly as some accounts would have it-and it is hard to imagine how all that capacity can be used. That, I believe, is at the heart of the global glut doctrine’s appeal: it is hard to imagine what a much more productive world economy would look like. But the deficiency is in our imaginations, not in the real economy, which will have no trouble using that capacity.

The misconceptions that make the idea of a global glut seem plausible probably begin with a fallacy of composition. If you look at individual industries in isolation, both productivity growth and globalization can easily seem to be job-destroying forces. The U.S. steel industry has experienced a dramatic rise in output per worker since 198o; the result has been a sharp drop in the number of steelworkers. Many developing countries have become exporters of clothing; the result has been a decline in garment industries in the advanced countries. So won’t a further rise in productivity throughout the economy and the further industrialization of the Third World mean job losses throughout the economy?

What this kind of reasoning misses is the indirect effects of these changes. Productivity gains in steelmaking may reduce the number of jobs in the steel industry, but they create jobs elsewhere, if only by lowering the price of steel, thereby releasing money to be spent on other things. Advanced countries may lose garment industry jobs to developing-country exports, but they gain other jobs producing the goods those countries buy with their new export income. To observe that productivity growth in a particular industry reduces employment in that same industry tells us nothing about whether productivity growth in the economy as a whole reduces employment in the economy as a whole. As long as consumers find a way to spend the increased income that is generated by economic growth, there is no reason for such growth to lead to any inadequacy of demand. And all the evidence suggests that people will find something to buy with whatever income they have.

But what, exactly, will they buy? That is a natural question, but an unfair one. Suppose that you had approached an economist in 1840, when most Americans were farmers, and textiles dominated the still small manufacturing sector, and informed him that 150 years later some two percent of the labor force would grow all the food and less than one percent produce all the cloth. And suppose you had demanded that he explain what everyone else would do for a living. He could not have given a very good answer, but he could with justice have argued on general principles that the economy would find something useful for them to do. (Global glut advocates tend to focus in particular on the loss of jobs in manufacturing and find it hard to imagine that the service sector could make up for that loss. Yet we have already seen how it can be done, since the U.S. economy is already overwhelmingly a service, rather than goods, economy. As a Russian emigre neighbor put it, “I don’t understand this country-it seems very prosperous, but I don’t ever see anybody making anything.”)

A somewhat different failure of the imagination prevails in thinking about the populations of the newly industrializing economies. Here it is obvious what consumers will demand: the things that people in advanced countries already have. What seems hard for Westerners to visualize is a world in which Chinese and Indonesians earn decent wages. After all, they have always been poor. And old images of the “teeming hordes” may play a role: a distinguished diplomat, journalist, and author told me, “I can’t see how their wages can rise, no matter how productive they become-there are just so many of them.” But one cannot argue that an increase in the productivity of Chinese workers is a massive shock to the world economy because they are so numerous, and at the same time argue that it is too small a force to raise the wages of so many people. Again, the deficiency is in our imaginations, not in the real world.

Finally, lurking behind the global glut argument may be an old fallacy about the relationship between consumption and investment. One thing capitalist economies do is accumulate capital. Not all of their productive capacity is used to produce consumption goods; some of it is used to produce capital goods that will expand future productive capacity. But not all of that future productive capacity will be used to satisfy consumers-some of it will be used to expand capacity even more, and so on. Now it is easy, if one puts it that way, to start to think of the whole thing as a sort of chain letter or Ponzi scheme: surely, you may imagine, at some point this business of building machines to build machines must come to an end, with disastrous results. (Something like this seems to be what Greider has in mind when he talks of the “manic logic” of capitalism). In fact, as Karl Marx could have told you, there is nothing unsustainable about an ever-growing capital stock in an ever-growing economy. It has worked for the last 50 years, and there is no end in sight.

Tilting at Windmills

None of the preceding should be taken as a declaration that all is right with the world economy. There are severe problems: inequality in the United States (exacerbated, though not caused, by imports of laborintensive products), unemployment in Europe (also slightly exacerbated by labor-intensive imports, but mainly due to rigid labor markets and bad macroeconomic policy), a Japanese economy struggling to overcome the consequences of a burst financial bubble, a number of newly industrializing countries facing potential crises due to financial excesses and lax bank regulation, and so on. On the whole, the condition of humanity-as measured by such crude but crucial indicators as life expectancy and malnutrition among children-is far better now than 20 years ago, largely because of economic growth in the Third World, but there are many shadows in the picture. One problem capitalism does not suffer from, however, is being too productive for its own good.

Imagining problems that do not exist has real costs. To speak to European advocates of the global glut theory is to be struck by their fatalism: they seem to have given up on the idea of making the European economy grow. And this fatalism already seems, at the time of writing, to have left the new Jospin administration almost completely ineffectual. Meanwhile, American global glutters seem to spend half their time complaining that nothing good can be done unless the country gives up the idea of a balanced budget, and the other half converting their doctrine into a new justification for good old-fashioned protectionism.

It is a bit funny, but also quite sad: those who preach the doctrine of global glut are tilting at windmills, when there are some real monsters out there that need slaying.