GUEST BLOGGER Robert J. Gordon




  In August 2012 the noted American economist Robert J. Gordon raised the deliberately provocative question “Is U.S. growth over?” After much debate among economists, Gordon has updated his argument in a recent paper, which this Series provides.

  The first Post showed how growth SLOWED between the periods 1891-1972 and 1972-2007. This Post analyzes HEADWINDS facing attempts to revive growth. The third Post will reiterate that those Headwinds are likely to offset future technological INNOVATION.

  All this should greatly interest Chinese. If future USA growth remains slow, that will strongly affect the PRC. To some extent the future of USA growth will also become the future of PRC growth, eventually. Chinese may wish to do similar analysis of China’s own growth.











  Robert J. Gordon

  Northwestern University and National Bureau of Economic Research

  [This post has been excerpted by Winckler from Robert J. Gordon 2013 “ ‘Is U.S. growth over?’ One year later. Reinterpretations, criticisms, and reflections.” Prepared for 11th Annual Conference, Center for Capitalism and Sociey, Dynamism and innovation in the West: Has decline set in? Columbia University, New York City, September 16, 2013. Here a few simplifications and explanations by Winckler appear in brackets. Non-technical readers can follow the main argument through key statements that Winckler has bolded. More technical readers may wish to attempt similar analysis of Chinese economic growth. REFERENCES to all three Posts were attached to the first Post in this Series. Gordon’s August 2012 paper was summarized in this Blog in its first Post, 130105, Section 3.3.]

  [This sequel paper concludes with] a general overview of several classes of criticisms directed at the earlier paper. These include the future potential of genome-based medical and pharmaceutical research to cure cancer and mental diseases, the role of “big data” in revolutionizing business practices, and the potential productivity-enhancing contribution of small, flexible, and cheap robots as well as driverless cars and trucks. Other optimists point to gas and oil fracking as achieving American energy independence and, through a decline in the relative price of natural gas relative to competing nations on other continents, a revival in American manufacturing, and this ray of hope for the future is assessed here as well.


  The 2012 working paper contained the sub-title “faltering innovation.” This self-inflicted wound has generated most of the negative commentary on my pessimistic view of the future. But there is no need to make any forecasts about future innovation. The reason for this change in emphasis is clear from the raw data on productivity. Let us start from the historical record displayed [in the 130921 first Post in this series] in Figure 2 on the growth rates of productivity, output per person, and hours per person over the four intervals 1891-1972, 1972-96, 1996-2004, and 2004-2012. We can simplify that graph to show only the growth rates for productivity, that is, total economic output per hour.

  This graph continues the technique used in [Figure 2 in the 130921 Post] of displaying the width of each bar as proportional to the length of the time interval. Thus the left bar is much wider than the others because it refers to a time interval of 81 years, while the three right bars cover respectively 24, 8, and 8 years.

  When the time intervals are divided in this way, we discover an important fact about historical productivity growth that has not been widely recognized. American history since 1891 can be divided into two “good” (i.e., green) intervals and two “bad” (i.e., blue) intervals. In my interpretation IR #2 [the Second Industrial Revolution] of the late 19th century and all of its complementary inventions that extended until 1972 created an unbroken chain of 81 years in which productivity grew fast enough to double every thirty years. There is a distinct watershed at 1972 for reasons discussed below, and the following 40 years were characterized by the two blue periods as shown interrupted for only 8 years by the green interval associated with the invention of the internet, web, and e-commerce.

  The post-1972 slowdown which lasted until 1996 created its own enormous literature on the cause or causes of the slowdown, containing speculation about energy prices, infrastructure, education, and much else. Robert Solow in 1987 captured the frustration created by the slow pace of productivity growth by quipping “We can see the computer revolution everywhere but in the productivity statistics.” But soon the revolution began to pervade the productivity statistics – it took the experts a few years to wake up, but the post-1995 revival was first noticed in early 1999 and then popped into widespread recognition with the benchmark revisions of the National Income and Product Accounts that occurred in late July, 1999.

  Soon thereafter the objective measures of the internet-led productivity revival faded away. The boom of the late 1990s was driven by an unprecedented and never-repeated rate of decline in the price of computer speed and memory, and a never-since matched surge in the share of GDP devoted to information and communication technology (ICT) investment. The ICT share in GDP declined along with the stock market in 2000-2002, leaving a subsidiary mystery of why productivity growth continued to be so strong until early 2004. Convincing research by Erik Brynjolfsson and his colleagues suggested that there was a substantial lag between the production and purchase of computer investment and the learning curve of how to use all the new equipment efficiently and productively. One of many examples was the installation in airport lobbies of electronic check-in kiosks, a product of late 1990s technology that was not implemented until 2001-04.

  The further time advances beyond that golden 1996-2004 interval, the more clear it becomes that the effect of the internet in creating IR #3 [Third Industrial Revolution] was temporary, apparently only eight years in the statistics. But computers were creating revolutions long before 1996, with computerized bank and phone bills in the 1960s, memory typewriters and airline reservation systems in the 1970s, personal computers, ATMs, and bar-code scanning in the 1980s, all before the internet revolution of the 1990s. The slow growth of productivity evident in the above graph for 1972-96 indicates that the contributions of the first round of computer applications masked an even more severe slowdown in productivity growth that would have occurred otherwise.

  It is easy to average the blue and green post-1972 areas as in Figure 7 and provide a succinct summary of the relative contributions of IR #2 and IR #3 in creating productivity growth. Output per hour in the total economy grew at 2.33 percent per year from 1891 to 1972 and then at 1.55 percent from 1972 to 2012. Thus in the past 40 years output per hour grew at almost exactly 2/3 of the rate of the 80 years prior to 1972. Since the post-1972 interval is half of the earlier 81-year interval, the result is that the pre-1972 years contributed fully three-quarters of the cumulative gain in productivity growth since 1891. 

  It is conventional in growth accounting to sub-divide the observed growth in output per hour into the separate contributions of capital deepening, improvements in capital quality, improvements in labor quality through education and training, and a residual term called “total factor productivity” (TFP), more than once described as “a measure of our ignorance.” Yet all capital deepening and changes in capital quality are not separate sources of growth, but rather a side-effect of innovation. As Evsey Domar famously remarked in 1961, “without technological change, capital accumulation would amount to piling wooden plows on top of wooden plows.” Similarly, changes in capital quality reflect the shift from long-lived structures which exhibit little technical change to shorter-lived equipment, particularly computers, which are short-lived precisely because rapid innovation makes them quickly obsolescent.

  If capital deepening and changes in capital quality are endogenous to innovation, then

  this leaves only two sources of labor productivity growth, namely innovation and improvements in labor quality achieved by higher levels of education.

  These two sources of economic growth, together with Figure 7, form the basis of my forecast of future productivity growth. If we optimistically assume that innovation will proceed at the same rate over the next 40 years as in the last 40, we can start with the realized 1972-2012 growth rate of 1.55 and subtract roughly 0.25 for slower growth of educational attainment, reaching the 1.30 percent future growth in productivity implicit in the 2012 paper and explicit in the growth decomposition summarized in the conclusion below.


  The achievements of the Second Industrial Revolution were set out in the 2012 paper and need only be summarized here. Within three months in the year 1879 three of the most fundamental “general purpose technologies” were invented that spun off scores of inventions that changed the world. Two of these are well known but the third is not. In October, 1879, Thomas Edison created the first working electric light bulb and by 1882 was distributing power by wire to customers in lower Manhattan, a revolution that made possible not only electric light, elevators, high-rise cities, stationary and portable electric power tools, consumer appliances, but also air conditioning that transformed life and work, especially in the American South.

  Between 1890 and 1930 the American household became fully “networked,” replacing the previous isolation by five types of connections – electricity, gas, telephone, running water, and sewer pipes. Running water and sewers in turn contributed not just to the first phase of female liberation but also laid the foundations for the conquest of infant mortality in the first half of the 20th century.

  Two months after Edison’s electric light, Karl Benz achieved the first reliable and workable internal combustion engine, and his initial patent is dated 1879. The third great invention of that year is less known – an Englishman named David Edward Hughes succeeded in sending a wireless signal several hundred meters in London almost two decades before Marconi won his earliest wireless patents (Smil, 2005, p. 241).

  The diffusion process of the epochal inventions of electricity and the internal combustion engine was surprisingly similar. Electricity had little initial impact before 1900 except in showpiece displays like the 1893 Chicago Columbian Exposition and the interior of department stores. But after 1900 the use of electricity took off with such speed that by 1929 virtually all urban dwellings were connected to power. Manhattan became vertical by 1929. The transition to electrification of manufacturing was concentrated in the period 1915-29.[1]

  Similarly, there was a gestation period when inventors pondered how to combine the power of an internal combustion engine with the previously flimsy structure of a horse-drawn carriage and to develop the transmissions, brakes, and other essential components. But as if in a race, the horseless carriage bolted out of the starting gate from zero vehicles in 1900 to such an extent that by 1929 the ratio of motor vehicles to the number of American households had reached 89 percent. By that year 93 percent of Iowa farmers owned a motor vehicle.

  At least three aspects of the Second Industrial Revolution have received less attention than they deserve. First is the multi-dimensional nature of the revolution. Unlike the one-dimension ICT revolution of the past 40 years, the creations of IR #2 spanned multiple dimensions – electricity and its spinoffs; running waters and sewers; motor vehicles and their complementary inventions such as highways, personal travels, and supermarkets; entertainment from the phonograph to the radio and motion pictures; public health and reduced mortality; and a revolution in working conditions that utterly changed life on the job from brutal and short to less physically demanding and uncomfortable work.

  The second surprising aspect is that everything happened all at once. When all these transformations are layered on top of each other, they hardly existed in 1870 yet were nearly complete in urban America by 1929. The third surprising aspect is that economic progress through 1972 mainly consisted of consolidating the incomplete aspects of IR #2 across many subsidiary and complementary inventions, including the spread of consumer appliances through the 1950s, the invention of television to supplement the radio and motion picture, the spread of air conditioning from commercial to residential ubiquity, the interstate highway system during 1958-72 as a complement to the motor car, and finally the development of commercial air transport from its primitive and tiny footprint in 1940 to its creation of rapid business and personal travel with the spread of jet planes which was accomplished by 1972.


  As indicated above, there is no need to predict any faltering or slowdown in the rate of innovation over the next 40 years. My baseline productivity growth forecast (for the total economy) of 1.30 percent per year starts from the realized growth rate over 1972-2012 of 1.55 and subtracts 0.25 for the likely effect of the slower advance of educational attainment.

  The historical record spanning the past four decades is all the demonstration we need that the benefits of the Second Industrial Revolution petered out in the early 1970s and were replaced by the less fruitful benefits of the Third Revolution. Yet from our current perspective as we peer out into the future, the achievements of the past 40 years set a hurdle that is dauntingly high. These achievements that must be matched for importance in the next four decades include

  <> Memory typewriters, the personal computer, word-processing and spreadsheets

  <> Bar-code scanning, ATM banking, cable and satellite TV

  <> Internet, e-mail, web browsing, e-commerce

  <> Google, Amazon, Wikipedia, Linked-In, Facebook

  <> Mobile phones, smart phones, ipads

  <> CDs, DVDs, i-tunes, Netflix, movie streaming

  <> Throughout every sector, paper, typewriters and clunk-clunk mechanical calculators were replaced by flat screens running powerful software.

  What is in store for the next 40 years? The usual stance of economic historians like Joel Mokyr is that the human brain is incapable of forecasting future innovations. He states without doubt that “History is always a bad guide to the future and economic historians should avoid making predictions.”[2] His analysis reasons from instruments to outcomes. As an example, it would have been impossible for Pasteur to discover his germ theory of disease if Joseph A. Lister had not invented the achromatic-lens microscope in the 1820s. Mokyr’s optimism about future technological progress rests partly on the dazzling array of new tools that have arrived recently to create further research advances – “DNA sequencing machines and cell analysis,” “high-powered computers,” and “astronomy, nanochemistery, and genetic engineering.” One of Mokyr’s central tools in facilitating scientific advance is “blindingly fast search tools” so that all of human knowledge is instantly available.

  Mokyr does not follow Erik Brynjolfsson and other optimistic leaders of technological progress down the trail of big data, small robots, and driverless cars. His examples involve fighting infectious diseases, the need of technology to help reduce the environmental pollution caused by excess fertilizer use, and the evocative query whether “can new technology stop [global warming]? He joins Brynjolfsson and others in worrying whether the machines will displace human workers, but he rightly appeals to two centuries of mechanical inventions that replaced workers by machines without causing mass unemployment.

  Mokyr finishes by citing Keynes’ 1931 essay “Economic Possibilities for Our Grandchildren” without any mention that Keynes suggested that technological change would “solve mankind’s economic problem” and (in a section that Mokyr does not quote) allow men to reduce their work load to 15 hours per week. We are still waiting for the 15-hour week for prime-age adult workers.


  My forecast of 1.3 percent annual total-economy productivity growth in the future does not require any forecast beyond suggesting that the past 40 years of 1972-2012 is a more relevant benchmark of feasible productivity growth than the previous 80 years of 1891-1972. Yet most of the observers of the debate about future U. S. economic growth appear to believe that “technological change is accelerating.”[3] A lynchpin in the optimism about future technological change is that the future cannot be forecast, and anyone who claims to forecast a lack of innovation in the future is doomed to the mistakes of past pessimists.

  While my pessimistic growth forecast for the U.S. does not call for the end of innovation, and rather accepts that future productivity growth of 1.3 percent is feasible (about 1.0 percent slower than the realized rate of 1891-1972), nevertheless the common assumption that future innovation is non-forecastable is wrong. There are too many historical precedents of correct predictions made 50 or 100 years in advance. After we review some of the most important long-run predictions of the future of technology going back to the mid-19th century, we will return to today’s forecasts for the next two to four decades.

  An early forecast of the future of technology is contained in Jules Verne’s 1863 manuscript Paris in the Twentieth Century, in which Verne made bold predictions about the Paris a century later in 1960.[4] In that early year, before Edison or before Benz, Verne had already conceived of the basics of the 20th century. He predicted rapid transit cars running on overhead viaducts, motor cars with gas combustion engines, and street lights connected by underground wires.

  Much of IR #2 was not a surprise. Looking ahead in the year 1875, inventors were feverishly working on turning the telegraph into the telephone, trying to find a way to harness electricity coming from batteries as the power source to create electric light, trying to find a way of harnessing the power of petroleum to create a lightweight and powerful internal combustion engine. Once that was achieved, the dream since Icarus of human flight became a matter of time and experimentation.

  Some of the most important sources of human progress over the 1870-1940 period were not new inventions at all. Running water had been achieved by the Romans, but it took political will and financial investment to bring it to every urban dwelling place. The first industrial revolution made possible steam-powered water pumps to distribute water within cities, and IR #2 soon replaced the steam pumps with more economical electricity-driven pumps. A separate system of sewer pipes was not an invention, but implementing it took resources, dedication, and a commitment to using public funds for infrastructure investment.

  A little-known source of technological forecasting was published in November, 1900 in the unlikely source of the Ladies Home Journal (Watkins, 1900). Some of the predictions were laughably wrong and unimportant, such as strawberries the size of baseballs. But there were enough accurate predictions in this page-long three-column article to suggest that much of the future can be known. Some of the more interesting forecasts in this article were:

      “Hot and cold air will be turned on from spigots to regulate the temperature of the air just as we know turn on hot and cold water from spigots to regulate the temperature of the bath.”

      “Ready-cooked meals will be purchased from establishments much like our bakeries of today.”

      “Liquid-air refrigerators will keep large quantities of food fresh for long intervals.”

      “Photographs will be telegraphed from any distance. If there is a battle in China a century hence, photographs of the events will be published in newspapers an hour later.”

      “Automobiles will be cheaper than horses are today. Farmers will own automobile hay-wagons, automobile truck-wagons . . . automobiles will have been substituted for every horse-vehicle now known.

       Persons and things of all types will be brought within focus of cameras connected with screens at opposite ends of circuits, thousands of miles at a span. . . the lips of a remote actor or singer will be heard to offer words or music when seen to move.”

      “Wireless telephone and telegraph circuits will span the world. We will be able to telephone to China just as readily as we can now talk from New York to Brooklyn.”

      “Grand Opera will be telegraphed to private homes, and will sound as harmonious as if heard from a theater seat.”

  Very soon after the Ladies’ Home Journal article, in 1900 H. G. Wells published his Anticipations which reflected on the influence of technological change on “human life and thought.” Compared to the Ladies’ Home Journal, Well’s famous book is narrow in its vision. It is Anglo-centric, concerned with development of mass transit and housing in London and the social problems that are related to that development. The book shows no recognition of how far was America at that point in developing a modern society driven by technology.

  The Jules Verne 1863 and the Ladies Home Journal 1900 visions of future technological progress were true leaps of imagination. Somewhat less challenging were predictions of the future made at the time of the 1939-40 New York Worlds’ Fair. By 1939-40, IR #2 was almost complete in urban America, so it is no surprise that the exhibits at the Fair could predict quite accurately the further complements to IR #2 inventions, such as superhighways and air conditioning.

  What could have been predicted in 1939 about the post-1940 era? Surely the war and the details of its aftermath were a surprise, but for domestic consumer goods the 1939-40 New York World’s Fair provided an accurate preview of the 1950s and 1960s. A future of air-conditioned homes and businesses was no intellectual stretch at the fair, as air conditioning in movie theaters began in 1922 and was nearly ubiquitous by the late 1930s.

  Television was introduced at the fair, and it was easy to predict then that television over the next two decades would follow the American model of commercially-supported radio, with entertainment provided over several large networks spanning the continent. Elaborate scale models at the New York fair showed the future high-rise city straddling multi-lane superhighways where travel on the inner lanes occurred at 100 miles per hour. Exhibits previewed the typical kitchen and its appliances almost exactly as came true for most Americans in the 1950s and 1960s. While commercial aviation was primitive in 1939, still it was easy to forecast from the rapid progress in the size and speed of aircraft just over the 1926-40 period that much larger aircraft could fly much longer distances, and indeed it took only a few years before the DC-6 and DC-7 were spanning the continent and the globe prior to the epochal introduction of the Boeing 707 jet in 1958.

  What was missing at the 1939-40 Worlds’ Fair was any vision of the computer revolution that created IR #3. But Norbert Wiener, a visionary, in a 1949 essay that was ultimately rejected by the New York Times, got a lot of the future of IR #3 right. Among his 1949 predictions were

  These new machines have a great capacity for upsetting the present basis of industry, and of reducing the economic value of the routine factory employee to a point at which he is not worth hiring at any price. . . . if we move in the direction of making machines which learn and whose behavior is modified by experience, we must face the fact that every degree of independence we give the machine is a degree of possible defiance of our wishes. The genie in the bottle will not willingly go back in the bottle, nor have we any reason to expect them to be well-disposed to us. (Markoff, 2013).

  Just as some future inventions have been a surprise, including the entire electronics and digital revolutions, other anticipated inventions never came to pass. The Jetsons’ vertical commuting car/plane never happened, and in fact high fuel costs caused many local helicopter short-haul aviation companies to shut down.[5] Dick Tracy’s wrist radio in cartoon comic strips of the late 1940s finally is coming to fruition 70 years later with Google’s miniaturized smart phone for the wrist, but the marginal utility of a wrist device compared to the existing smart phone sinks into the insignificance of measurement of small things.


  Recall that my future forecasts come true if the headwinds evolve as I suggest, and if future innovation matches the contribution to productivity of the past four decades since 1972. To match in the future the digital revolution in the past will be an enormous achievement, but this match is required for even my pessimistic forecasts to come true. Yet a bit of skepticism is warranted about the next four decades of innovation, because we have already experienced the digital revolution for the past 40 years during which the most fruitful applications of electronics have already occurred.

  Just as the 1939 World’s Fair forecast numerous life-changing inventions and further diffusion of existing inventions like air conditioning and television, so today’s future-looking literature can credibly identify several directions of major advance. Yet when more closely examined, the value to consumers of these hypothetical inventions seem of small scale compared to the inventions that could be accurately forecast at the 1939-40 New York world’s fair, by the Ladies’ Home Journal in 1900, or by an acute observer at the dawn of IR #2 in 1875.

  The future advances that are widely forecast fall into four main categories – medical, big data, robots, and driverless vehicles. Enthusiasts of “big data” sometimes label this category of advance as “artificial intelligence.” It is worth looking briefly at the potential of each of these categories of future innovation in turn.

  Medical and Pharmaceutical Advances. Even though Q&A sessions after my speeches often start with “what about future advances in medicine related to the genome?” it turns out that this category of future advance has already proved to be disappointing. The most important sources of higher life expectancy in the 20th century were achieved in the first half of that century, when life expectancy rose at twice the rate of the last half. This was the interval during which infant mortality was conquered by the discovery of the germ theory of disease, the development of anti-toxins for diphtheria, and the near-elimination of air-and water-distributed diseases through the achievement of constructing urban sanitation infrastructure.

  Many of the current basic tools of modern medicine were developed between 1940 and 1980, including antibiotics, heart procedures, chemotherapy, and radiation. Life expectancy in the U.S. inched ahead in the last half of the 19th century albeit at half of the rate of the first half of the same century. The current status of science in medical treatment and pharmaceutical advance is well described by Jan Vijg (2011, especially his Chapter 4). Progress on physical disease and ailments is progressing faster than on mental disease, so that we can look forward in two or three decades to an exponential rise in the burden of taking care of elderly Americans who are physically alive but in a state of mental dementia. Pharmaceutical research has reached a brick wall of rapidly increasing costs and declining benefits, with a decline in major drugs approved each pair of years over the past decade. At enormous cost drugs are being developed that will treat esoteric types of cancer at costs that no medical insurance system can afford. Vijg is highly critical of the current regime of drug testing in the U.S. as inhibiting risk taking.

  Small Robots and 3-D Printing. Much attention has been paid to small robots in the popular media since “Baxter,” the small $25,000 robot, appeared on the TV program 60 Minutes. The appeal of the Baxter example is that the cost is so cheap. Baxter can perform the functions of a human, as long as it can stand in place, and replace humans in routine assembly-line work. From a broader time-perspective, however, Baxter elicits a big yawn. Industrial robots were introduced by General Motors in 1961. By the mid-1990s, robots were welding auto parts and replacing workers in the lung-killing environment of the auto paint shop.

  Reflections on Baxter lead to skepticism that it/he is a major threat to American jobs outside of routine tasks in manufacturing, which only makes up 10 percent of American employment. For his demonstration at the TED conference in Long Beach in late February, Baxter had to be packed in a suitcase. He could not get his own boarding pass and walk onto the plane. This is the problem with robots – they are both mentally and physically limited to narrow tasks. Their homeland is in manufacturing plants and in wholesale distribution warehouses, where they find the Amazon books and bring them to the packer, who so far remains a human because the robots have not yet figured out how to arrange books in a package.

  This lack of multitasking ability is dismissed by the robot enthusiasts – just wait, it is coming. Soon our robots will not only be able to win at Jeopardy but also will be able to check in your bags at the sky cap station at the airport, thus displacing the skycaps. But the physical tasks that humans can do are unlikely to be replaced in our lifetime by robots. The physical leaps on and off trucks and onto porches exhibited by human UPS drivers create a high hurdle for robot designers.

  What is often forgetten is that we are well into the computer age, and every Home Depot, Wal-Mart, and local supermarket has self-check-out lines that allow you to check out your groceries or paint cans by scanning them through a robot. But except for very small orders it takes longer, and so people still voluntarily wait in line for a human instead of taking the option of the no-wait self-checkout-lane. The same theme – the most obvious uses of robots and computers have already happened – pervades commerce. Airport baggage sorting belts are mechanized, as is most of the process of checking in for a flight. For more than two decades payments have been made by plastic cards instead of by cash or checks as in our parents’ era, but that revolution is over and the marginal utility of paying by swiping a cell phone is at the outer limit of small gains.

  3-D printing is supposed to support a future revolution, but its potential impact is limited. Recent reports suggest that 3-D printing is a small deal best suited for one-on-one customized operations, such as the ability to create a crown in a dentist office instead of having to send out a mold. Recent reports suggest that 3-D printing represents custom production rather than mass production, and thus interpreted retreats from the economies of scale and efficiency of Henry Ford’s 1913 assembly line.

  Big Data. The enthusiasts for future technical progress seem to think that their world operates exponentially whereas the previous eras of economic progress were linear.[6] A growth rate is a growth rate – a 2.0 percent growth rate in per-capita income of the U.S. over 1891-2007 represents an exponential increase of EXP(116*2/100) = 10.2. Yes, real per-capita income over those 116 years increased by a factor of 10. What is more exponential than that?

  What is lost by the enthusiasts for big data is that it is almost all a zero-sum game. The vast majority of big data is being analyzed within large corporations for marketing purposes. If the marketing wizards can use big data to figure out what their customers buy, why they change their purchases from one category to another, or even why they move from merchant to merchant, then that particular corporation may be able to devise a strategy to steal market share from competitors. An excellent and current example involves the large legacy airlines with their data-rich frequent flyer programs. The analysts at these airlines are constantly trolling through their big data to try to understand why they have lost market share in a particular city or with a particular demographic group of travelers.

  The one true aspect of the word “exponential” applied by the enthusiasts to “Big Data” is that the quantity of electronic data has been rising exponentially for decades. But diminishing returns have set in. The sharp slowdown in productivity growth in recent years has overlapped with the introduction of smart phones and ipads, which consume huge amounts of data. The skeptic feels comfortable in predicting that these sources of innovation will disappoint in what counts, their ability to boost output per hour in the American economy.

  Driverless Cars. This category of future progress is demoted to last place because it offers benefits that are so minor compared to the invention of the car itself, or the improvements in safety that have driven a ten-fold improvement in fatalities per vehicle mile since 1950. The most important distinction is between cars and trucks. People are in cars to go from A to B, much of it for essential aspects of living such as commuting or shopping. Thus the people must be inside the driverless car to achieve their objective of getting from point A to point B. What is the revolutionary achievement of consumer surplus by being able to commute without driving? Instead of listening to the current panoply of options, including Bluetooth phone calls, internet-provided music, or home-provided i-tunes music, you can actually look at a computer screen and do your e-mail. This is not a potential revolution even on a par with the invention of the iphone, much less the web or e-commerce, much less the invention of the automobile itself.

  That leaves the alleged future productivity advantage offered by driverless trucks. This is a potentially productivity-driving innovation, albeit within the tiny slice of U.S. employment consisting of truck drivers. However, anyone who opens his or her eyes to what truck drivers do realizes that driving from place to place is only half of it. UPS drivers jump off trucks and deliver packages to businesses and residences, knowing exactly where to put them and whether to drive up driveways or not. Wholesale trucks arrive at supermarkets and do not just stop in the back. The drivers are responsible for loading the cases of Coca-Cola or the stacks of bread loaves onto dollies and placing them manually on the shelves. In fact, it is remarkable in this late phase of the computer revolution that almost all placement of individual product cans, bottles, and tubes on retail shelves is achieved today by humans rather than robots.


  The core contrast between past and future in my growth forecast is between a realized 2.0 percent growth rate per year in real per-capita GDP over 1891-2007 to a future post-2007 growth rate of 1.3 percent per year for productivity, 0.9 for average output per capita, 0.4 for the pre-tax pre-transfer income received by the bottom 99% of the income distribution, and 0.2 for the disposable income of the bottom 99% once the obligation is imposed to fix government debt problems by slower future growth in entitlement transfers and/or faster growth in taxes.

  The forecast of slower U.S. group relies for about 2/3 of its shocking prediction on the impact of the six headwinds, of which four are quantified here in terms of their impact on future growth. What about innovation? The original 2012 paper “Is Growth Over?” had the unfortunate sub-title “faltering innovation”, but this paper recognizes that it is not necessary to forecast that future innovation will be any less important than the record of innovations already achieved in the four decades since 1972.

  Instead, the sense in which innovation has been “faltering” is not any prediction about the future compared to the recent past, but rather is a comment from the raw data about American economic growth in the 80 years before 1972 as contrasted to much less rapid growth in the four decades after 1972. That slowdown which happened so long ago is the best available evidence that the third industrial revolution (IR #3, mainly digital, post-1972) was a mere shadow of the second industrial revolution (IR #2, multi-dimensional, 1891-1972).

  While there is no need to forecast that future innovations will be less important than the past, some skepticism is offered here. When young audiences are confronted with the question “Is Technological Change Accelerating?” they vote en masse “yes.” In a debate in the early summer of 2013, I lost a debate on that question by a margin of 71 to 29. I can only interpret the result as coming from the infatuation of the younger generation of their smart phones and i-pads, together with their ignorance of the sources of past progress that have provided them with the air conditioning and electric plugs that allow them to enjoy their electronic devices.

  After one year of debate about this topic, the most interesting questions remain to be discussed. Clearly Canada, the Nordic nations, and Switzerland suffer less from the headwinds than does the U.S. They are less subject to the interlocking quagmire of poverty, low school achievement, high college costs, an expensive and inefficient medical system, and lack of child care than the U.S.

  How many other nations suffer less from the headwinds? Does it matter to the rest of the world whether U. S. innovation falters or not? The great American inventions of the last two decades have spread across the globe with impressive speed. My final and perhaps most radical thought is that innovation is a free good for the world as a whole, while headwinds are uniquely nation-specific. This leads to the suggestion that over the next 50 years the faltering growth of the American standard of living will be surpassed by one country after another, and there is no need today to choose whether the eventual winners emerge from northern Europe or eastern Asia.



  Economist Robert J. Gordon is one of the USA’s leading experts on the current productivity of the American economy. He is writing a book on the technological revolution that propelled American economic growth between about 1891 and 1972. In August 2012 he published a paper asking the deliberately provocative question “Is U.S. Economic Growth Over?”

Gordon’s purpose was to alert American economists and policy-makers to the possibility that, as they continued to attempt to speed “recovery” from the 2007-2009 Great Recession, they could not count on a return to the high rates of growth to which Americans have become accustomed. His question has indeed provoked much debate among American economists.


  [1] Reasons for the long delay in the impact of electricity in manufacturing are provided in the classic and oft-cited paper by David (2000).

  [2] This and other quotes in this section come from Mokyr (2013).

  [3] In June 2013 the Economist blog ran a debate using Oxford Union rules. A stark proposition “Technological Change is Accelerating,” a set of initial statements, rebuttals, and closing statements, and then the vote. My opponent Andrew McAfee of MIT won the vote by 71 percent to 29 percent.

  [4] Details about the Verne book and its predictions come from Vijg (2011, pp. 35-36).

  [5] When I arrived home in Berkeley from college or graduate school near Boston, I often traveled by the SFO helicopter line which flew frequent schedules from the SFO airport to the Berkeley marina. This company abruptly shut down in 1974 as a result of the first oil shock.

  [6] This is a direct quote from Erik Brynjolfsson in his presentation to the National Association of Business Economists on September 8, 2013.

推荐 5