The Map and the Territory

by Alan Greenspan

  Advances in the speed and efficiency of the transportation of goods increase GDP. Aluminum plate is more valuable at an aerospace assembly plant than at the plate mill. The spread in the 1870s between the prices of cattle at ranches in Texas and those at the railheads in western Kansas narrowed sharply with the substitution of rail transport for expensive, legendarily long cattle drives. With the onset of transcontinental railroads, the time it took to move goods from the East Coast to the West Coast fell from six months to six days.33 The number of workers required to move a ton of freight one mile fell materially, a development that added significantly to lower costs and hence higher national value added.

  Aside from the production and transportation of goods, remarkable innovative advances in business services also added importantly to output. Quickened communication of information that facilitates the arbitraging of markets improves the accuracy of the vast system of relative prices, which, in turn, assists in directing our scarce savings into the most efficient technologies. As noted in Chapter 6, with the introduction of effective telegraphy (1844) and especially the transatlantic cable (1866), the cost and timeliness of acquiring information that determined key decisions in producing and moving a good to its final place of consumption fell sharply, and standards of living rose commensurately.

  Perhaps most important, information significantly lessens the amount of real resources required to produce any level of output because it reduces short-term uncertainties. Before this generation’s revolution in information availability, most twentieth-century short-term business decision making had been hampered by uncertainty created by inherent delays in the transfer of information. In college I worked part time at a department store to make ends meet. I was astonished by how many people and how much time the store employed trying to keep track of inventories. By the time the inventory count was completed, it was already out of date. Owing to such paucity of timely knowledge of customers’ needs and of the real-time size and location of inventories, businesses generally required substantial programmed redundancies in the use of energy, materials, and people to function effectively. Because decisions were made from information that was days or even weeks old, production planning required costly inventory safety stocks to respond to the inevitable unanticipated and misjudged levels of demand. Through most of the twentieth century, determination of the amount of inventories on hand often required laborious and time-consuming counting of individual items. Inventories often built up for weeks before management had become sufficiently informed to curtail production.

  Today, executives have real-time accounting for inventories, receivables, and payables, and can act promptly to address emerging imbalances.34 Clearly, the remarkable surge in the availability of real-time information in recent years has enabled business management to remove large swaths of inventory safety stocks (thanks to just-in-time inventorying) and programmed redundancies of labor. That means fewer goods and worker hours are absorbed by activities that, while perceived one or two generations earlier as necessary insurance to sustain valued output, in the end, most often produce nothing of enduring value themselves.35

  These developments emphasize the essence of information technology: the expansion of knowledge and its obverse, the reduction of short-term uncertainty. As a consequence, risk premiums that were associated with many forms of business activities have permanently declined, reducing the amount of capital required to back up the information systems. In short, information technology raises output per hour for the total economy, in part by reducing hours worked on activities needed to guard productive processes against the unknown and the unanticipated.

  It’s worth singling out one innovation that both increases the effectiveness of labor hours and at the same time enhances the climate-related quality control of output: air-conditioning. Willis Carrier introduced the first modern form of air-conditioning in 1902, and it quickly widened location options for plants and revolutionized manufacturing and commerce in America’s South.

  THE PHYSICAL DOWNSIZING OF OUTPUT

  For generations Americans had reveled in acres upon acres of massive motor vehicle assembly plants, ever taller skyscrapers, enormous dams, and longer bridge spans. I never ceased being in awe of America’s industrial might in the 1950s as I drove along the south shore of Lake Michigan toward Chicago, passing one huge steel mill complex after another. But running parallel, starting in the early postwar years, was a growing view that ever larger and more resource intensive production had an upside limit. I recall walking the streets of downtown Pittsburgh in the late 1950s with coke oven ash crunching under my feet. Pollution and the environment were terms that rarely captured public attention, but that was about to change.

  In the years immediately following the war, belching smoke stacks were symbols of industrial progress. I remember pondering in my youth where all the waste from the coal emitted from our chimneys ended up. I assumed, as did most of my acquaintances, that big blue skies and deep blue seas had an unlimited ability to absorb and cleanse. It was never true of course; industrial production imposed unpriced externalities on society. Industrial polluters were getting the services of waste disposal for “free.” But industrial progress was very high on the nation’s agenda, and pollution was inexorably tied to progress, hence tolerated. The same view appears to be held widely in modern-day China, although it may be changing.

  In 1962, biologist Rachel Carson’s Silent Spring detailed the effect of the widely used pesticide DDT and other chemicals on the environment. Largely as a result, the Environmental Protection Agency (EPA) was born in 1970. A couple of years later, the Club of Rome commissioned a book, The Limits of Growth, evaluating the ever larger consumption of the world’s raw materials and speculating about possible future chronic shortages and inflationary imbalances. That forecast did not materialize, but the notion of ever greater waste creation from economic growth fueled by ever more humungous structures and physical goods remained.

  No doubt, the emergence of the environmental movement led to some shift in tastes away from ever greater material consumption and especially away from consumption most likely to result in a degradation of the natural environment. This likely provided some spur to the downsizing of real output. But by far the more important factor driving down the physical component of output and driving up the conceptual content of output was the discovery of the electrical properties of silicon chips and the development of the integrated circuit. This technological development and all the innovations that followed in its wake revolutionized the structure of advanced economies. The fabrication of integrated circuits required negligible quantities of physical materials such as silicon, a natural resource in widespread abundance.

  DOWNSIZING AND PRODUCTIVITY

  The discovery of the electrical properties of silicon and transistors created a world in which the creation of economic value has shifted dramatically toward conceptual and impalpable values, with decidedly less reliance on physical heft.

  Three quarters of a century ago, our radios were bulky and activated by large vacuum tubes. Years later, owing to the insights that followed from modern electronics, the same function was served by pocket-sized transistor packs. Today we have iPhones that also serve as cameras, flashlights, GPS, portable media players, and docks for a seemingly endless (and growing) list of applications, all in a small handheld device. Moreover, there were other significant technological downsizing developments well beyond the silicon chip. Metal beverage cans are now rolled to thinner tolerances than was conceivable decades ago. Lightweight fiber optics replaced vast tonnages of copper. Space-heating technology enabled reduction in the fabric weight of apparel because people didn’t need to wear warm clothes indoors. Advances in architecture and engineering, and the development and use of lighter but stronger materials, now give us the same working space in newer buildings with far less concrete and steel tonnage than was required decades ago.

  Even the physical quantity of goods consumed in creating econom
ic services has been affected. Financial transactions that historically were buttressed with reams of paper are now memorialized electronically. The transportation services industry now moves more goods with greater convenience, while consuming substantially less fuel per ton mile of transportation.

  The considerable increase in the economic well-being of most advanced nations in recent decades has come about without much change in the bulk or weight of their gross domestic products. The weight of nonfuel raw materials that are consumed in the United States has not been growing measurably since the late 1970s (Exhibit 8.5), with raw tonnages no greater today than they were three or four decades ago. This means that increases in the conceptual components of GDP—that is, those components reflecting advances in knowledge and ideas—explain almost all of the rise in real GDP in the United States, and presumably elsewhere in the industrial world. Services, at constant prices, however, with no physical weight, are roughly the same share of real GDP as they were in 1949, and hence have contributed only modestly to downsizing. The brunt of the change has been in the downsizing of goods.

  So the economy has gotten lighter without question. Quantifying this, however, is not all that straightforward. Similarly, measuring the weight of GDP is difficult. We cannot put the total GDP on a scale and read off the number of tons. We do have estimates of the total (nonfuel) tonnages of raw materials that enter into goods production, and if we assume that the physical weight of input is proportional to the tons of output, we would have an unambiguous measure of GDP output tonnage. In practice, however, technological advances have probably improved the ratio of output to input.36

  But the key measure of output—real GDP—raises a series of questions. Real GDP is supposed to measure nominal GDP in dollars of constant purchasing power. That requires a measure of price that reflects and incorporates the ever-changing quality of goods. A new car purchased in 2013 almost certainly has many features undreamed of by the purchaser of a new car in 1998. I am astounded at how close we are to being able to vocally direct our cars while sitting back and enjoying the ride. There is a vast economic literature on how to convert nominal dollars into constant (real) dollars, and unending changes over the years in the statistical techniques employed. What the Bureau of Economic Analysis seeks is a quality-adjusted price for goods and services, most recently a measure of the quality-adjusted real output in 2005 dollars.37

  To estimate the weight in tons of U.S. GDP goods, I employed the data compiled by the U.S. Census Bureau on both the value and weight of imports and exports transported by water and air back to 1953. Those data enable us to calculate the weight of both imports into the U.S. production process and exports ultimately destined to contribute to production abroad, relative to the constant 2005 dollar value of imports and exports.

  For U.S. imports since 1955, the ratio of weight to real (constant 2005 dollar) value has fallen by a relatively stable 3.1 percent annual rate. Since 1977, the ratio has declined by an average of 4.6 percent per year (Exhibit 8.6). Assuming that this ratio can be applied more broadly to all private real goods output in the United States creates a tonnage estimate for aggregate output that is broadly consistent with an independent compilation of raw commodity inputs into the production process, such as iron and copper ore, cement, and steel scrap (Exhibit 8.7).38 Both estimates indicate that the weight of private GDP output has leveled off since the 1970s. Of course, the composition of imports is not the same as the composition of the economywide goods GDP. But there is little evidence that this ratio is significantly biased by the shifting composition of imports, at least relative to tonnage estimates.

  THE TURNING POINT

  The correlation between growing economic activity and growing weight of real GDP apparently peaked in the late 1970s. In recent years, the conceptual contribution to economic activity has reflected importantly the explosive growth in information gathering and processing techniques, which have greatly extended our capability to substitute ideas for physical volume.

  In the years ahead, telecommunications and advanced computing will doubtless take on an even greater role. By expediting the transfer of ideas, information technology creates value by facilitating the substitution of intellectual for physical labor in the production process, much as the American railroads in an earlier time created value by transferring physical goods to geographic locations where relative shortages made those goods more valuable. At the turn of the last century, for example, we created economic value in the United States by moving iron ore from Minnesota’s Mesabi Range down to furnaces in Pittsburgh, where it was joined with West Virginia coal to produce steel. In today’s environment, economic value is increasingly created by fitting ever smaller silicon chips ever closer together with still larger data capacity than earlier, much bulkier units. At least to date, Moore’s law still prevails. (Though as Gordon Moore himself recognized, miniaturization has physical limits that some see approaching in the near term.39)

  THE BENEFITS

  Two clear benefits of the economy’s “weight loss” are the reduced depletion of the world’s finite natural resources in the context of growing populations and the expansion of international trade. Obviously, the smaller the bulk and the lower the weight, the easier it is to move goods across national boundaries. High-value computer products are a major and increasing factor in global trade.40

  Also implicit in the downsizing of products is the increased integration of many of the world’s production facilities. Inflationary bottlenecks tend to emerge when domestic productive facilities are pressed to capacity by burgeoning domestic demand. But if additional supplies from other world producers can be made readily and quickly available, such pressures can be significantly eased, effectively reducing the level of domestic capacity required for any given global demand for a commodity. The cost of moving gravel across continents makes it difficult to envisage foreign gravel pits as backup for excess domestic demand. But the ease with which downsized electronic components can be moved essentially integrates much of the world’s electronic component capacity. Misplaced or displaced production facilities become a much smaller problem.

  Thus, as we move beyond the current crisis and the general downsizing of economic output continues, worldwide production and inventory controls become far more feasible and price pressures associated with production dislocations less likely. One can only imagine the downsizing that will emerge with maturing nanotechnology and 3-D printing. Goods or their electronic versions may eventually be moved in a manner that brings to mind Star Trek’s instantaneous teleportation devices.

  MILLENNIA OF STAGNATION

  Many, if not most, innovations fail and are quickly forgotten. But innovators keep trying. We are continually pressing forward. Life and our competitive propensity seemingly require it. However, if innovation and productivity growth is a human propensity, why was it essentially stagnant for nearly two millennia prior to the eighteenth century’s Age of Enlightenment?41 Clearly, the propensity to enhance our material state of living is a necessary but not a sufficient condition for the actual achievement of such growth.

  Economic growth requires that the interactions between market participants are governed by a rule of law in which the rights of ownership are effectively enforced; it implies a legal framework (for example, contract law) that facilitates the free interchange of goods and services in a society. Economic growth also requires a degree of abstinence from consumption to provide the savings required to fund a sufficiently large addition to our capital stock that, in turn, is needed to leverage human inventiveness on the path to rising levels of productivity. The insights of the eighteenth century’s John Locke, David Hume, and Adam Smith, among others, developed such a framework that spread rapidly throughout the then-developed world, accelerating standards of material well-being twentyfold in the subsequent more than two centuries. Real per capita GDP of the Western world, according to economic historian Angus Maddison42, which had been virtually stagnant during the millennium precedi
ng the eighteenth century, has accelerated at a 1.7 percent annual rate since 1820. Productivity trends will have more to do with material standards of living in, say, 2030 than any other single economic statistic.

  STATISTICAL APPENDIX 8.1

  To place both productivity and patent issuance on a comparable basis, I treat the annual series of output per hour (a flow) from the perspective that it is the stock of cumulative embedded technology. In essence, the level of productivity is best thought of as being a measure of the previously accumulated level of knowledge.43

  I then assume that the rate of depreciation of the real capital stock, as estimated by the BEA, can be substituted (not implausibly) for the annual rate of depreciation of the “stock” of accumulated technology (output per hour). The gross additions to the stock of technology are then estimated by adding the net change in the technology stock during the year to the amount of depreciation estimated from the BEA rate of depreciation of the related real capital stock. That series is shown in Exhibit 8.8 along with annual patent issuance. Both series are measures of gross new additions to the infrastructure that produces productivity. Patent issuance, as I concluded, does parallel gross additions to the stock of productivity but does not lead it.

  NINE

  PRODUCTIVITY AND THE AGE OF ENTITLEMENTS