By Paul M. Romer
From Policy Options, July-August 1994
For information contact: gsb_newsline@gsb.stanford.edu
4/97
This talk was delivered April 5, 1997 at a conference sponsored by the Stanford Alumni Association and the Stanford Business School Alumni Association held at the Hotel InterContinental in London.
Thank you.
As a new member of the faculty, I can say that it's a pleasure to be affiliated with an institution like Stanford. It's also a pleasure to become a kind of adopted member of the extended family of Stanford graduates that you represent.
I decided that I would start this morning by telling you about the relationship between accountants and economists. In the business school, and in other organizations, accountants are paid more than economists. Because the intellectual prerequisites for the two disciplines are fairly similar, you might expect some economists or would-be economists to switch over and become accountants. Absent some kind of barrier, this kind of supply response should ultimately drive down the salary differential between the two fields. But there is a barrier. It turns out that many economists wanted to be accountants, but didn't have enough personality for the job.
This helps explain what my role is this morning. My job is to give you a pedantic lecture that will make all the rest of the speakers seem scintillating and entertaining by comparison. With this in mind, I decided to give you a brief account of the history of thought in macroeconomics. Now economists are often accused of being unfeeling, but we don't believe in unnecessary cruelty. To serve my role, there's no need for me to go boring on for the full hour. I'll stop a little more than 1/2 way through the hour and take some questions. Be sure to have some questions ready, because if you don't, I'll tell more economist jokes.
Now, any history of thought in economics has to start with Adam Smith. I will not depart from tradition. When Smith wrote his famous treatise on national economic policy, he convinced the intellectuals of his day that monetary tokens do not constitute the true wealth of a nation. Instead, he argued wealth lies in tangible assets such as tools, structures, and improved farm land. He concluded that a nation becomes more wealthy if its citizens and its government all refrain from excessive consumption today and save for the future by producing more of these assets. In the classical approach to macroeconomic policy that he helped create, saving and capital accumulation are the central concerns. Excessive spending on current consumption, especially by the government, is the most serious threat to sustained economic growth. As a result, the economy could suffer from a "paradox of thrift." By trying to save too much, consumers could cause a recession that would lead to less total saving and investment. What this offered was the promise of something for nothing. If the government spent more, both savings and consumption could increase.
In the early part of this century, John Maynard Keynes argued that this classical fear of excessive consumption was misplaced, even dangerous. To Keynes and his followers, the biggest risk was that consumption spending might be too low, not too high. The danger he saw was excess supply and insufficient demand. Keynes argued that the classical aversion to deficits stood in the way of the required increase in government spending. By the middle of this century, the classicals were in full retreat and Keynesians had carried the day. Since then, macroeconomic policy in many industrialized countries has been based on the Keynesian strategy of using monetary and fiscal policy to increase "aggregate demand" and current consumption.
In the last 15 or 20 years, dissatisfaction with the results from Keynesian demand management has grown. The classical approach to macroeconomic policy has made something of a comeback. Many economists once again are calling for smaller government deficits, higher private savings and more rapid capital accumulation. They argue that excessive government spending reduces capital accumulation and that capital accumulation is the key to growth.
For someone who is not an economist, this oscillation between calls for more saving and more spending is more than a little frustrating. When the newspaper reports that consumer spending has increased- and that private savings has decreased- is this good for the economy or bad? Moreover, to people who are careful students of business, science and history, there is something suspect about this entire discussion. Doesn't the key to economic success lie in a nation's ability to introduce valuable new goods, to improve the quality of existing goods, and to find more efficient ways to manufacture and deliver these goods? If so, why do economists seem to devote so much attention to monetary and fiscal stimulus on the one hand and purchases of existing capital goods like fork lifts on the other? Where is the discussion of innovation, invention, discovery and technical progress?
A recent branch of work in economics that goes under the label of "endogenous," "neoSchumpeterian," or simply "new" growth theory validates some of the concerns about innovation and discovery expressed by people who are not economists. It suggests that both the classical save-more and the Keynesian spend-more macroeconomic policy prescriptions miss the crux of the matter. Neither adjustments to monetary and fiscal policy, nor increases in the rate of savings and capital accumulation can by themselves generate persistent increases in standards of living. This recent work suggests that the most important job for economic policy is to create an institutional environment that supports technological change. This sounds simple enough, but policy-makers must also resist the temptation to impede change when it causes temporary disruption. In a modern industrial democracy, achieving a balance between support for economic progress and tolerance of economic change is not a simple task.
Within the economics profession, recent work on the theory of long-run growth is described in terms of mathematical equations and statistical analysis. It is possible, nevertheless, to give an interpretation of this work in terms of a metaphor that is as concrete as the image of a family that needs to save (the metaphor invoked by classical economists) or of a pump that needs priming (the one favored by the Keynesians). The description I'll offer makes use of the tired rhetorical device of a sports metaphor but relies on an unfamiliar sport. Imagine that an economy is like a swim team. Creating a higher standard of living is like helping swimmers achieve faster times in their races.
As background, I have to explain that swimming coaches use two approaches to give swimmers a short-run performance boost just before a race: tapering and blood doping. Blood doping refers to the process of drawing blood from a swimmer weeks or months in advance and then re-infusing the blood just before the race. This raises aerobic capacity because hemoglobin is the carrier of oxygen in the blood. Performance falls when blood is withdrawn. As the body replenishes its stocks of hemoglobin, performance returns to its previous level. Then when the withdrawn blood is re-infused, the swimmer will swim faster because she has a temporary excess of hemoglobin and an increased oxygen carrying capacity. But there are risks. If too much blood is infused, the swimmer can experience serious medical complications from "sludging" and excessive clotting.
The other method for increasing performance in the short run is a taper. A taper refers to the process whereby an athlete follows a training schedule that tapers off in intensity in the weeks before an important race. Most areas of competitive sport are governed by the iron law of athletics: no pain, no gain. The taper, however, seems to exploit a "paradox of training" that parallels the paradox of thrift cited by Keynesians. Training hard just before a race seems to reduce race times, just as, so the Keynesians claimed, an increase in savings could reduce capital accumulation.
Despite the pejorative sound of the term, blood doping can be a valuable medical procedure. If someone knows in advance that he will have to undergo a surgical procedure that has a high risk of blood loss, it is a good idea to store up some blood in advance. If he suffers from bleeding, injections of his own blood can be a crucial life-saving measure. Symmetrically, injections of money can in some circumstances bring an economy back to life. Economists now understand that a major contributing factor to the depression of the 1930s was the decision by the Federal Reserve Board to stand idly by and watch as the financial system in the United States hemorrhaged. Between 1930 and 1933, the money supply fell by about a third as the banking system collapsed.
Although transfusions are valuable during real emergencies, this does not commend blood doping as a regular part of athletic training. Transfusions can be very effective in averting death, but they cannot permanently raise an athlete's performance by much. In the same way, the power of monetary policy to cause a sharp recession or to speed recovery from a recession does not imply that we can use frequent adjustments in the supply of money or the level of interest rates to achieve steady increases in standards of living. Our experience with monetary stimulus is like that of a swim team with a medical trainer who has been using blood doping aggressively, achieving some short-run increases in performance. But he pushes the team right up to the limit where the complications from sludging and clotting (i.e., inflation) become unacceptable. Then the trainer withdraws blood, causing a sharp reduction in performance. This creates pressure for more short-run stimulus, and the boom-bust cycle repeats itself.
If blood doping is like monetary stimulus, tapers are like government deficits. A reduction in training may give a short-term boost, but it impedes progress toward better aerobic conditioning and more strength. Conditioning, strength and technique all improve with the total distance that a swimmer covers in practice. Shorter workouts during a taper reduce this total. In most cases, government deficits have a positive effect on short-run output, but at the cost of a reduction in national saving and the rate of accumulation of capital. The supposedly free lunch of the paradox of thrift turns out to be a trade-off of short run benefit against long run cost. In the long run, less capital will translate into fewer inputs that we can use in production. Of course, tapers and deficits, like injections of blood and money, do have their uses. If you are a coach training a team for an important swim meet, or if you are the leader of a nation that must fight a war, you may be quite willing to give up some future performance to get a short-term boost. But there is no free lunch.
Now, it will come as no surprise to you that the popularity of both coaches and politicians goes up when they institute a taper or undertake fiscal stimulus. The short run gains seem to loom larger than the distant costs. As a result, in the politics of macroeconomics, both the left and the right have developed their own distinctive versions of fiscal stimulus. Politicians on the left promise an investment-led recovery of growth driven by deficit-financed government spending. Politicians on the right promise a supply side recovery that is driven by deficit-financed tax cuts. Part of the return in the economics profession to a more classical policy stance can be explained as a reaction to these political pressures and the persistent deficits that they have caused in modern industrial democracies.
You might think that left wing promises of costless increases in spending and right wing fantasies about self financing tax cuts have been completely discredited. But the political power of the free lunch persists. Stimulative monetary policy and government deficits can both be useful policy tools. But in the long run, they cannot make a nation rich. Unfortunately, neither can the opposite of large deficits - an extremely high rate of savings and capital accumulation. We now know that the classical suggestion that we can grow rich by accumulating more and more pieces of physical capital such as fork lifts is simply wrong. The problem an economy faces (but that a family putting its savings in the bank can ignore) is what economists call "diminishing returns." In handling heavy objects, a forklift is a very useful piece of equipment. When there were few forklifts in the economy, the return on an investment in an additional forklift was high. But as we increase the total number of forklifts the value of each additional forklift drops rapidly. Eventually, additional forklifts would have no value and become a nuisance. The return on investment in an additional forklift diminishes and eventually becomes negative. As a result, an economy cannot grow merely by accumulating more and more of the same kinds of capital goods.
This process of diminishing returns has an analog for a swimmer. A swimmer invests in pool time just as an economy invests in capital. She cannot continue to improve her times year after year merely by training harder to improve aerobic conditioning and physical strength. The physics of drag in water and the physiology of the human body impose an upper bound on the speed at which anyone can move through the water using a particular swimming stroke. There is a limit to how far a brute force approach to training can take a swimmer.
In swimming, as in macroeconomics, we nevertheless continue to see steady improvement. From the 1950s up through 1990, the number of world records that were broken in an average year has remained about the same. This improvement derives fundamentally from the same source in both swimming and the economy - cumulative improvements in technique. In swimming, there has been a dramatic improvement in the techniques that swimmers use to propel themselves through the water. They exactly parallel the improvements in the techniques of production that have raised our standards of living.
One of the advantages of using a swimming metaphor to think about the sources of long-run improvement in performance is that it is relatively easy to see what the technical innovations are and why it takes a combination of big breakthroughs and small refinements for real progress to take place. For 400 years (from about 1500 to 1900) a stroke that resembled what we now know as the breast stroke was thought to be the most efficient swimming stroke in the Western world. This stroke, in which the arms remain underwater even during the forward recovery, was used throughout most of the nineteenth century in racing competition. The first person to swim the English Channel did so using the breast stroke in 1875. For readers who do not appreciate how inefficient the breast stroke is, with its underwater recovery of the arms and legs and with the chest plowing through the water like a barge, swimming the Channel with the breast stroke today would be like running a marathon backwards - a stunt you do to get publicity because it was so senseless.
Around the turn of the century, English and Australian swimmers copied techniques used by native peoples in Australia and Ceylon and developed what is now known as the free style or crawl stroke. Because the arms recover out of the water, this stroke has inherent efficiency advantages over the breast stroke. As early as 1844, a version of this novel kind of stroke was demonstrated by two native American Indians at an exposition in England, which was then the center of worldwide swimming competition. But the inherently inefficient breast stroke had been refined, and early attempts to use an over-arm recovery of the crawl stroke were crude and awkward. (One of the biggest initial stumbling blocks was the lack of a kick that would work with the new arm stroke.) As a result, the efficiency advantages of having the arms recover out of the water were not generally recognized for more than 50 years.
One of the chronic problems that people have in understanding the potential for discovery is that we extrapolate our current circumstances both forward and backward in time. We imagine most people expect that the basic swimming strokes have been known for centuries and are surprised to learn that something as obvious as the crawl stroke was discovered at about the same time as quantum mechanics. People also tend to underestimate the enormous gulf that separates the discovery of abstract principles from their effective application. The basic physical laws of fluid dynamics were established in the 1700s but even with all the sophisticated laboratory experiments done today, there is still disagreement among coaches and swimmers about what constitutes the most efficient swimming stroke. New techniques continue to be proposed. Aspects of the stroke are still being refined. For example, only in the 1950's - less drag when the body slices through on its side instead of plowing through like a barge.
When a new technique like the crawl stroke is discovered, it takes lots of investment in pool-time for the swimming community to uncover these refinements. It also takes lots of pool time for any individual swimmer to master what is already known. In the same way, when a new type of capital good like the railroad or the digital computer is invented, it takes lots of investment to reap the benefits from these discoveries. But in either case, the return on additional investment falls rapidly in the absence of a steady flow of new techniques and new refinements. Without discovery of new techniques, diminished returns sets in rapidly.
This lesson shows the limitations inherent in the return to a classical emphasis on savings and investment. Especially in discussions of education and human capital accumulation, it is easy to be seduced by a suitably extended version of the family finance metaphor. If I am a high school educated worker and if all of my children become college educated professionals, then their standards of living should be higher than mine. I can invest, therefore, not by putting money in the bank, but by sending my children to college.
This makes sense at the level of the family, but in the economy as a whole, a strategy based solely on human capital accumulation eventually runs into the same limits from diminishing returns that arise with physical capital accumulation. Suppose that there had been no innovation and no technological change during the nineteenth century. We could have accumulated more human capital by increasing the fraction of the population that was high school educated or even college educated. We could also have accumulated more physical capital in the form of sailing ships, water wheels and ox carts. But eventually we would be forced to admit that we had little use for one more college graduate who is employed driving one more ox cart. The increases in standards of living that we achieved in the last century were possible only because of the discoveries and innovations that let new physical capital and new human capital be put to work in high return activities.
All economists agree that governments need to pay attention to the textbook fundamentals of monetary and fiscal policy. They also agree that governments should encourage, or at least avoid discouraging, the accumulation of capital. But these steps by themselves are not enough. A government must create an environment that fosters progress- and change- in the techniques we use.
For a government, there are two parts to any strategy for creating this kind of environment. The first is the one that most people think of in discussions of technological progress. The government can support the research and training missions of university-based science and engineering. Market incentives by themselves cannot simultaneously solve the problems of discovering and distributing knowledge. In the area of knowledge creation mechanisms like royal patronage and its modern descendent, the government research grant, can help. The government pays people to discover new things and to share their knowledge freely. Subsidies for education and training can indirectly assist in this process. Governments, however, cannot do it all. Market incentives must guide the process of discovery. Some kinds of university-based research and training have a greater likelihood of generating increases in economic value than others. If the private sector and universities are working well together, the market incentives can guide researchers and teachers toward valuable new areas, as happened in the United States when the opportunities in the emerging petroleum refining industry led to the creation of schools of chemical engineering.
For a nation as a whole, an effective institutional arrangement for supporting technological advance must therefore support a high level of exploration and research in both private firms and in universities. Moreover, it must support a high degree of interaction between these two domains. Both people and ideas must move readily between them. If they do not, the university research can become sterile and irrelevant. Private sector efforts can lose the steady flow of new talent and new ideas that sustains its creativity.
This support for universities is critical to the emergence of new technologies, and no university illustrates the benefits of this arrangement as well as Stanford. But this is not the halt of the policy process that I want to emphasize this morning. It is equally important that governments not impede change. Unfortunately, we cannot have growth without change, and change is disruptive. The citizens of industrial democracies increasingly demand that governments protect them from disruption. From a long-run perspective, the most serious side effect of the Keynesian revolution in macroeconomic policy may have been the intellectual legitimacy it lent to the impulse to blame the government for every individual economic misfortune. After all, if the government controls powerful levers that can be used to prevent job loss and unemployment, someone who loses a job would seem to have a strong claim of negligence against a government that fails to act.
In economics, the real force of the athletic slogan of "no pain, no gain" lies not in an admonition to consume less and save more. Rather, it comes from the warning that in many cases, things have to get worse before they can get better. If you talk to a swimmer, she will tell you that each time she tries a new variation in her technique, her performance falls. She typically has perfected the details of the old, less efficient stroke, much as the English swimmers had perfected the inherently inefficient breast stroke. When she makes a change, she initially gets many of the details wrong and performance falls. Then slowly the efficiency improves as she adjusts all the details to get the best performance out of the new stroke. If swimmers were never willing to tolerate temporary reductions in performance, they would not be able to experiment and develop better technique.
Joseph Schumpeter called the equivalent process of experimentation, deterioration and improvement that takes place in an economy "creative destruction". A new or improved product typically replaces an existing one. The creation of economic returns for inputs used to make the new good is associated with the destruction of returns for inputs that made the old good. Workers who used to produce the old good have to be shifted into some new activity. At best this results in a spell of unemployment. At worst, it can lead to a permanent loss in income.
If the old activity had been a particularly profitable one (the production of mainframe computers, for example), stockholders of the company engaged in this activity will lose when the new good (the personal computer) comes along. If the company that made the mainframes shared some of its profits with its workers, paying them well above the market wage, these workers may also suffer permanent reductions in income when the new personal computer industry destroys the profits of the mainframe manufacturers. The individual workers who lose their jobs may never again find jobs that pay them more than the market wage. The job market is characterized by constant turnover, with new hiring at some firms and job losses at others. Some workers suffer permanent decrease in income when these changes take place. Much of this turnover takes place during an economic expansion. The rest of it gets concentrated during recessions, when the vulnerability of older firms becomes plainly evident. In politics, discussion usually focuses on the individual worker, but the same forces show up at the level of the firm. Some of you listening today will have to lead firms that are forced to shrink. If the law allows, you may decide to fire some workers. This is not a sign that you are evil or incompetent.
Innovation and change create winners and losers. Job losses and pay cuts, divestitures, cutbacks, and bankruptcies are the most visible, and politically most powerful symptom of this disruption. But we know that for society as a whole, innovation, discovery and technological change offer large net gains because the new goods or processes are more efficient and more valuable than the old ones. A world with personal computers opens up opportunities that were inconceivable in the era of the mainframe. If we had not tolerated disruption in the past, we would still travel in ox carts.
Macroeconomic policy measures like fiscal or monetary stimulus cannot change the fact that many older firms need to shrink to make room for the new firms to grow. Aggressive stimulus measures can perhaps change the timing of the job losses company cutbacks that this implies, but they cannot avoid them. If these policy measures delay the process whereby an existing firm comes to terms with new competitive realities, they might make things worse. More aggressive measures like bailouts, government loans, nationalization of failing firms and prohibitions on firing workers are even more likely to have perverse effects in the long run. The ultimate contraction is likely to be associated with more job turnover and income loss than it would if firms had responded sooner. And in the process, the underlying engine of growth may be throttled back.
This perspective suggests that at least some of the effort expended by governments to prevent recessions and avoid job losses may be misguided. The circumstances of the 1930s may have been exceptional (in part because of government actions that made the contraction so severe). Stimulus measures that were appropriate then may not be useful in the ordinary course of events. Such efforts may only postpone processes of adjustment that are inevitable. There is a growing suspicion, for example, that high levels of long-term unemployment in continental Europe may be the unintended consequence of policies designed to fight unemployment and recessions, not the result of the recessions themselves. A more productive approach to policy would be to make the adjustment process as efficient and painless as possible, and to maintain the conditions that lead to rapid entry of new firms that compete for workers.
The greatest challenge for policy makers may therefore be one of political leadership. Each year, the electorate voices new demands for security and protection from disruption. Meanwhile, evidence mounts that government efforts designed to provide security and prevent disruption have had a corrosive effect on the operation of markets, on the processes of entry and exit by firms and ultimately on the kind of competitive spirit that any successful competitor - in the world of athletics or the world of work - must cultivate. Many workers suffer losses and setbacks during their careers, yet still have access to material advantages that would have been unattainable by even the very rich just two generations ago. Most swimmers lose races, but still benefit from participation in the competition.
To succeed in narrow economic terms, but also in broader human terms, leaders must give people the confidence to compete. They must encourage people to believe that economic change brings real opportunity together with risks that are real, but manageable. Leaders must not cater to the demands of a generation that seems to believe that each person is entitled to a job, a house in the suburbs and two cars just for serving time in school - as if everyone deserved a trophy just for putting in a specified number of hours in the pool. They must inspire people to dive in, strive for each small improvement in technique, compete to the best of their abilities.
Nations that can sustain a policy stance that tolerates, or perhaps even fosters, the process of creative destruction can count on sustained economic growth that will carry them into the next century. Those that are most successful in creating institutions that foster discovery and innovation will be the worldwide technological leaders. Through mechanisms like free trade and transfers of technology by multinational firms, nations that are less successful in the cultivation and commercial exploitation of science and technology can still follow comfortably along in the wake of the leaders. But nations that try to resist change by protecting inefficient firms, impeding flows of goods and ideas, and making a high level of income an entitlement instead of a reward will slowly be left farther and farther behind.
In closing, let me pose the question that often goes unasked in this kind of discussion. Would being left behind be so terrible? Perhaps the growth we get is not worth the disruption that we have to endure to get it. I believe that the growth is worth the disruption but I think it would be healthy to talk more openly about the costs and benefits.
In many people's minds, the advantages that come from economic growth are exaggerated. An increase in our level of income or even an increase in its rate of growth does not lead to as much of an increase in human happiness and satisfaction as we expect. Many politicians seem to speak as if all our problems would be solved if we could just increase our trend rate of growth by 1/2% or 1% per year. This surely can't be right. Leader probably believe this last century, but our growth rate in this century did increase and we are still having trouble balancing the government budget. My guess is that the direct material benefits from economic growth are overstated but that the psychic benefits are underestimated.
Imagine for a minute that we lived in a world where there was no potential for growth in income per capita. The kind of world that most people lived in our evolutionary past. In such a world, the only way to achieve a higher standard of living for yourself or your children is to see to it that someone else and his children experience a lower standard of living. We would be stuck in what game theorists call a zero sum game. History shows us that we humans are quite capable of causing harm to our neighbors and their children if doing so is the only outlet we have for pursuing our ambitions.
Economic growth makes it possible for you to help yourself and your children without hurting anyone else. It is not a panacea, but it does make it logically consistent to have hope and to believe in charity.
For me, this would be a lot to give up just to avoid job loss and disruption in our lives.
Thank you.