How "New" Is Today's Economy? ============================= J. Bradford DeLong Professor of Economics, University of California at Berkeley, and Research Associate, National Bureau of Economic Research delong@econ.berkeley.edu http://www.j-bradford-delong.net/ August 18, 1998; Version 2.0; Second Draft 6500 words A. Introduction --------------- At the end of the twentieth century legions of the powerful--politicians, intellectuals, journalists, business leaders, and visionaries--made what can only be called pilgrimages to a spot in California between San Francisco Bay and the Pacific Ocean, some forty miles south of San Francisco: Silicon Valley. In the past the powerful had made similar pilgrimages. In the 1830s and 1840s Alexis de Tocqueville (reprinted 1991), Benjamin Disraeli (reprinted 1993), and Friedrich Engels (1844) were among the pilgrims who journeyed to Manchester, England, to see the shape of the future emerging from the factories (and the smog, and the slums) of the rising British cotton textile industry. In the 1920s the pilgrims traveled to Detroit, to see Henry Ford's assembly lines and the extraordinary upward leap in productivity that turned automobiles from luxuries into commodities within the reach of the bulk of American families. The revolution of mass production of Detroit may have driven an even bigger change in human conditions of life than the original industrial revolution of Manchester: Aldous Huxley (reprinted 1989) wrote in _Brave New World_ of a future in which people dated years from the production of the first Model-T Ford, and in which the key problem facing governments was how to brainwash people so that they would consume enough to keep the economy near full employment. In all three cases people went on pilgrimage to see the future--the industrial future, in Manchester; the mass-production future, in Detroit; and the microelectronic future, in Silicon Valley. The pilgrims invariably came away convinced that new technologies had created a fundamentally different future--a new society, a new culture, and a new economy. But what, exactly, is "new" about this particular "new economy" that awaits us in our future? Each pilgrim to Silicon Valley seems to bring back a slightly different report. Many of the reports are simply wrong. Some see as brand-new phenomena patterns of entrepreneurship and enterprise that are in fact quite old--and do so because they fail to view Silicon Valley from a long enough historical perspective. Some fail to understand the most important fact about economic growth: that ever since the industrial revolution there have always been new industries, new inventions, and new commodities. of the measurement of economic growth. Nevertheless, when all is said and done, there is something to the idea that we live in a "new economy"--just as there was something in the air in the past in Manchester or Detroit. What is new is not the extraordinarily rapid pace of invention and innovation: there is always, or rather for the past two hundred years there has always been, some leading sector or other the pace of invention and innovation has been extraordinarily rapid. What is new is not that our standard of living is growing faster than official statistics suggest: for at least a century our official statistics have been understating rates of economic growth because official statisticians don't have the information to adequately understand leading sectors. What is--or rather might be--substantial in the idea of a "new economy" is that the information age may see the breakdown of those underpinnings of scarcity and control over commodities that supported economists' belief that the market system was a good social mechanism for guiding production and distribution. Economists preach that the Invisible Hand of the market will carry us all to wealth, luxury, and utopia as long as the government is smart enough to protect property rights, enforce contracts, and tweak the system around the edges to deal with "externalities." Such reliance on competitive markets has, by and large, proven a good bet in the past. But it rests on principles--that commodities are excludible (easy to control), rival (scarce and expensive to society to make use of), and transparent (easy to identify)--that may not be true when the information age arrives. That would be a new economy indeed, but one very different from the picture currently being painted. B. Leading Sectors ------------------ The first dimension of "newness" of today's economy--the one that strikes almost all observers immediately, and leads to breathless descriptions of technological revolution-is the sheer speed of technological progress in the semiconductor and the semiconductor-derived industries. _Business Week's_ Stephen Shepard (1997) writes of how information technology and the computer- and network-driven international integration of business are: ...the magic bullet--a way to return to the high-growth, low-inflation conditions of the 1950s and 1960s. Forget 2% real growth. We're talking 3%, or even 4%. Forget double- digit inflation and the natural rate of unemployment... The "undermining of the old order..radical restructuring that is making us more efficient" have left traditional analysts of the economy in the dust: "unable to explain what's going on... wedded to deeply flawed statistics and models." From his post at _Wired_, Executive Editor Kevin Kelly (1997) writes of the new economy as a: ...tectonic upheaval... [driven by] two stellar bangs: the collapsing microcosm of chips and the exploding telecosm of connections....tearing the old laws of wealth apart.... As the size of silicon... shrinks... costs ...become cheap and tiny enough to slip into every... object we make...an ephemeral package... your chair, each book, a new coat, a basketball.... Soon, all manufactured objects, from tennis shoes to hammers to lamp shades to cans of soup, will have embedded in them a tiny sliver of thought... Since the early post-World War II invention of the transistor, the onrush of invention, innovation, and industrial expansion in information technology has been constant and rapid (see Hanson (1982)): transistors, integrated circuits, microprocessors, and now entire-computers-on-a-single-chip and the high-speed worldwide networking of potentially every computer within reach of a regular or a cell phone. Year after year Moore's Law (named for Intel Corporation co-founder Gordon Moore) continues to prove itself: the density of silicon-on-a-single-chip doubles (and thus the cost of silicon circuits to consumers halves) every eighteen months or so (see Saxenian (1996)). Moore's law has been at work since the early 1960s. It will continue until--at least--the middle 2000s. Observers note the enormous fortunes made on the stock market by founders of startup corporations that have never turned a profit: how intenet bookseller Amazon.com is seen by the stock market as worth as much as established bookseller Barnes and Noble. Observers see how last year's high-end products are this year's mass-market products, and will be obsolete and sold for scrap within two more years (see Kelly (1997)). Hence this vision of this "new economy": a vision of never-ending cost reductions driven by technological innovation, "learning curves," "economies of scale," "network externalities," and "tipping points." In the old economy you made money by selling goods for more than they cost: in the new economy you make money be selling goods for less than they cost--and relying on the learning curve to lower your costs next year. In the old economy you boosted your company's stock price by selling valuable goods to customers: in the new economy you boost your company's stock price by giving away your product to customers--and relying on network externalities to boost the price you can charge for what you have to sell next year. In the old economy the first entrant often made big mistakes that followers could learn from and avoid: it was dangerous to be first into a market. In the new economy the first entrant that passes a tipping point gains a nearly unassailable degree of market dominance. And there are pieces of the world that fit this vision of the "new economy" very well. Think of the fortune of Bill Gates. Think of the rapid and steady price declines of silicon chips. Think of the rocket-like rise of new companies that did not exist a decade ago to prominence in the stock market. Think of the rise of companies that did not exist three decades ago--like Intel--to industrial prominence . But, somewhat paradoxically, it is along this marveled-at dimension that our economy today is perhaps the least new. For what this particular set of returning pilgrims to the future are describing are the standard economic dynamics of a "leading sector," of a (relatively narrow) set of industries that happen to be at the center of a particular decade's or a particular generation's technological advance. There have been such leading sectors well over a hundred years. Manchester was the home of the leading sectors of the 1830s. It was the Silicon Valley of its day--and saw the same creation of large fortunes, the plummeting of product prices, and the sudden growth of large enterprises. Every leading sector goes through a similar process. Consider the automobile. The average car purchased in 1906 cost perhaps $52,640 of 1993 values (Raff and Trajtenberg (1997)). By 1910 the price had dropped to $39,860--while the "quality" of the average car had jumped by at least 31 percent. By the time the heroic, entrepreneurial age of the American automobile came to an end during World War I, an average car cost 53 percent less in inflation-adjusted dollars than and was at least twice as much car as in 1906: 4.2 times as much car for the (inflation-adjusted) dollar as a mere decade before. This does not match the pace of innovation found in semiconductors recorded to Moore's Law--which generates at least a 32-fold, not a four-fold, increase in value over a decade. But it is in the same ballpark, especially once one recognizes that the tremendous improvements in semiconductors have not been matched by many of the other components that make up microelectronics products. Thus a citizen of the late nineteenth century would not have had to wait for our age in order to see the "new economy." In order to be astounded by how the best of products got cheaper and more reliable every year, he or she would only have had to go to Detroit, and be astounded by the pace of technological progress in the automobile manufacturing industry. During the 1920s magazines like the _Atlantic_ would publish articles claiming that mass production made not just for greater efficiency or higher productivity but "a better world" and demanded the rapid creation of a "Fordized America" (see Hounshell (1984)). The automobile industry is not alone: other industries have had similar transformations during their times as leading sectors. For example, my wife and I have a small dining room, with a small four-bulb chandelier. But should we go to Monterey for the weekend and leave the dining room light on, by the time we return our dining room will have consumed--in our absence--as much in the way of artificial illumination as an average pre-1850 American household consumed in a year. It cost them about five percent of their household income in candles, tapers, and matches. Because of the technological revolutions that made possible the cheap generation and transmission of electricity, it costs us so little that we cannot see it in our Pacific Gas and Electric bill (Nordhaus, 1997). Silicon pilgrims have made an elementary mistake. They grew up with the goods of the industrial revolution--from automobiles to washing machines, railroads to container ships to airplanes to radios. They remember these goods from their childhood, and see them today in much the same form as they existed thirty or more years ago. So they assume that such "industrial" goods must have always existed, that the pace at which they changed must have always been glacial. But we have had a succession of productivity revolutions in leading sectors since the start of the industrial revolution: that's why it's called the industrial revolution--it kicked off the process of staggered sector-by-sector economic transformation of which Silicon Valley is today's instance. The economy cycles through a number of leading sectors: textiles, transportation, construction, textiles again, watches and jewelry, telegraphs, construction again, telephones, transportation again, household utilities, household appliances, broadcasting, textiles and apparel, and medical care-all before the microelectronics revolution. It will continue to cycle through different leading sectors in the future, long after the pace of technological change in microelectronics has slowed down. So how would pilgrims returning from Silicon Valley change their vision of the future after recognizing that much of what they have seen are the more-or-less standard economic and technological dynamics of a leading sector? I believe that they would change their mind in at least four areas: First, they would recognize that in microelectionics, as in every leading sector, the "heroic" period of rapid technological progress will come to an end. Henry Ford perfects the Model T. Cable and Wireless figures out how to properly insulate submarine telegraph cables. The first easy-to-find antibiotics--penicillin--are all discovered. Moore's Law exhausts itself. Thereafter computers and communications will become a much more mature industry-with different focuses for research and development, different types of firms, and different types of competition. Second, in every leading sector the true productivity revolution comes before the heroic period has come to an end. The first railroads connected key points between which lots of bulky, heavy, expensive materials needed to move. Later railroads provided slightly cheaper substitutes for canals, or added redundant capacity to the system in an environment in which one's ability to make profits often hinged on one's power to threaten competitors with a rate war. The first three TV networks came amazingly close to sating Americans' taste for audiovisual entertainment. The first uses of modern telecommunications and computers--Plain Old Telephone Service, music and news via radio, the first TV networks, Blockbuster Video, scientific and financial calculations, and large database searches--were the highest-value uses. Thus it is unwise to extrapolate the economic value added by semiconductors, computers, and telecommunications far into the future. Later uses will be lower-value uses: if they were higher-value uses, someone would have applied technology to them already. This is a version of the standard economist's argument that the $1,000 bill you think you see on the sidewalk can't really be there: if it were, someone would have already picked it up. But this economist's argument is rarely false: there aren't many $1,000 bills to be found lying on sidewalks. Third, after the heroic age the form of competition changes. During the heroic age, technology alone is the driving force. After the heroic age, firms make money by giving customers exactly what they want. As long as automobile prices were falling and quality rising rapidly, Henry Ford could do very, very well by riding the leading edge of the technology wave: making a leading-edge car and letting the customer choose the color and the options, as long as the options were zero and the color was black. As long as computer prices are falling and quality rising rapidly, Bill Gates can do very, very well even if his computers crash and show their users the Blue Screen of Death twice a day. After the 1920s, however, the Ford Motor Company was overwhelmed by Alfred P. Sloan's General Motors, which figured out how to retain most of Ford's economies of scale while at the same time offering consumers a dizzying variety of brands, models, styles, and colors--a worthwhile undertaking, but not the stuff of economic revolution. Before GM, no one knew what kind of options car buyers really wanted. Today, no one knows what kind of options computer and internet access purchasers really want. Moreover, no one will until a computer and communications counterpart emerges to play the role of twenty-first century GM against Microsoft's role of Ford. That is why so many entrepreneurs and so many venture capitalists today are placing so many high-risk bets, hoping that their one particular idea will turn out to be the twenty-first century equivalent of the Cadillac tail fin: the particular option that is cheap to produce yet in high demand because it fits exactly with what consumers really want. The consensus belief is that the company that will play GM to Microsoft's Ford will succeed by making access--to computing power, to research materials, to the yet-to-be-built high-bandwidth successor to the internet, to *information*--uniquely valuable to *you*. But no one knows exactly how. And many medium-sized fortunes will be lost and some mammoth-sized ones made in the process of figuring out exactly how. Fourth and last, each leading sector does produce a technological revolution. It does leave us with previously unimagined capabilities. The railroad gave us the ability to cross the continent in a week rather than in months. Electric power gave us the ability to light our houses and power our appliances. Metallurgy gave us the ability to build really big and really strong buildings and bridges. Microelectronics has given us intellectual vision orders of magnitude better than eagles. We now have, or soon will have, the ability to find out within minutes the answer to any question as long as the answer is already known by somebody. We now have, or soon will have, the ability to calculate the answer to any problem that we can formulate clearly. One and a half generations ago the economists William Sharpe, Merton Miller, and Harry Markowitz did the work on how a rational investor should diversify his or her asset portfolio that won them the 1990 Nobel Prize in Economics, and at the time all three thought that their work was of theoretical interest only: that the calculations required to implement their formulas were beyond the reach of humanity. But today the computing power to carry out calculations orders of magnitude more complicated than those sketched by Sharpe, Miller, and Markowitz is on every Wall Street desk. But a technological revolution is not an economic revolution. Just because microelectronics revolutionizes our capabilities to process information doesn't mean that it will dominate our economy. The economy, after all, focuses its attention on what is expensive--not on what is cheap. In every leading sector the story has been the same. Once the new product has been squeezed into a relatively low-cost commodity, the economic energy has flowed elsewhere. There is no "economy" of air--air is free--even though our capability to breath cheaply is perhaps the most important thing of all. _Business Week_ does not run special issues on electric lighting and its vast superiority over whale-oil lamps. Thus as our capabilities grow, the salience of our expanded capabilities in the economy--which is, after all, the realm of things that are scarce--does not. C. Growth and Measurement ------------------------- A second (and somewhat overlapping) set of pilgrims have returned from Silicon Valley believing that the economy could grow much faster than it has over the past generation. They believe that the technological revolutions of microelectronics are having a large and positive impact on our standard of living, and could be having a larger and more positive impact if only those managing our economy would recognize the importance and enormous benefits of this postindustrial revolution in telecommunications and computers. This group returns from Silicon Valley to announce that growth in productivity and living standards is (or could be) back to the rapid pace of the first post-World War II generation between 1945 and 1973. The post-1973 productivity slowdown was a profound shock to America. It caused a stock market crash in the mid-1970s. It meant that government promises of future benefits that had been based on assumptions that tax revenues would continue to rise at their pre-1973 rate could not be fulfilled. It made false the basic American assumption that each generation would find itself living significantly better than its parents' generation, with bigger houses, better jobs, and markedly easier lives. But today new economy advocates announce that because of microelectronics and telecommunications the productivity slowdown is over, and that economic growth can return to the pace of the period before 1973. Many good things will follow: real wages and living standards will grow rapidly, the Federal Reserve will be able to expand the money supply more rapidly without generating higher inflation, and the stock market will continue to boom because the high value of today's stock market is perfectly justified by the high profits to be made in the "new economy." Thus Newt Gingrich opines that better technologies should give us an economy that can year after year generate measured economic growth of 4 or 5 percent without rising inflation, instead of the 2.5 percent per year in growth in measured GDP calculated by the tight-fisted central bankers at the Federal Reserve. Stephen Shepard (1997) writes of how information technology is a "transcendant technology" that affects everything: "boosts productivity, reduces costs, cuts inventories, facilitates electronic commerce." The "statistics are simply not capturing what's going on" because of the way the Department of Commerce collects and processes data: "we don't know how to measure output in a high-tech service economy." As best as I can tell, this group of returning pilgrims seems to have failed to recognize the importance of the word "measured" in the phrase "measured economic growth." Their pilgrimage has led them to the belief that true economic growth is higher than measured economic growth. And they are right: there are a large number of statistical and measurement problems built into our official economic statistics that together imply that "true" economic growth is faster than "measured" economic growth by one percentage point per year or so. But what this group of returning pilgrims fails to see that this understatement of growth and improvement of livings standards is a lot older than today's "new economy." Our government's official statistical system was little, if any, better at measuring growth in the network-television-and-superhighways economy of the late 1950s or the automobiles-and-appliances economy of the late 1920s than it is today. For more than fifty years national income accountants at the Commerce Department have known that estimates of gross domestic product grossly underestimate the economic benefits from the inventions and innovations of leading sectors. Yet the national income accountants have continued to follow their established procedures because they lack the information to do a better job, and would rather report numbers that they can count reliably rather than numbers that are based on guesswork. It is likely that the problems of measurement today are somewhat bigger than in the past: Moore's law tracks productivity growth that is even faster than that of previous leading sectors. But they are not of an entirely different order of magnitude So the correct answer seems to have two parts: First, that improvements in living standards and economic growth are faster than standard statistics suggest (at least for the middle class: it is harder to get excited about $29.99 a month unlimited internet access and cellular phones if you are on food stamps). Second, that the productivity slowdown is not over, that growth in real wages and in living standards is not back to its pre-1973 pace, and that the stock market is unlikely to continue to boom--for back before 1973 the American economy possessed faster growth in measured GDP and roughly the same gap between "true" and "measured" growth, because for nearly a century before 1973 the Department of Commerce has had similar problems estimating the economic benefits from previous leading sectors that it has estimating the benefits from telecommunications and computers today. Economic growth is more rapid than official statistics tell us. But the productivity slowdown is still there: still a shock to American expectations about the magnitude of economic progress. And just because measured growth is lower than real growth does not mean that the economy could be growing faster than it is. How would we tell if economic growth was too slow? How would we tell if inappropriate policies were hobbling the rate of economic progress? One guide would be the rate of investment: are bad economic policies stealing capital that should be going to expand the productive capacity of the economy. While you can argue that the budget deficits of the 1980s hobbled economic growth by crowding out investment, the budget deficits of the Reagan and Bush administrations are now gone. If businesses formed long-term ties with their workers, and believed that worker loyalty to the firm should be matched by firm loyalty to the worker (as in today's Japan), it would be hard to figure out if the economy was growing as fast as it could. But we live in an economy in which firms have little if any loyalty to their workers, and have a duty to shareholders to fire workers whenever doing so increases profits. So if productive capacity were growing faster than production, businesses would be firing workers on a large scale: unemployment would be rising as firms used technology to economize on workers. Yet there are no signs of any persistent rise in unemployment. The labor force as a share of the adult population is a record high: it is not that large numbers of discouraged workers are masking massive underemployment. Unemployment remains constant--a good indicator that the economy is growing at the sustainable rate of growth of its productive potential--when measured real GDP grows at some 2.5 percent per year (see Blinder (1997); Krugman (1997)). D. The New Economy ------------------ Nevertheless, this debunking of the "new economy" is not unsatisfactory. Pilgrims return raving of visions that cannot reflect reality. Much political-economic discussion of the new economy does result from basic confusion between measured and true rates of growth. Much of the hype is the same hype that always surrounds the firms in a leading, technologically-dynamic sector--whether Edison General Electric in the 1880s or the Radio Corporation of America in the 1920s. But. The pilgrims did see something. First, it is likely that official statistics understate recent economic growth by more than they have in the past simply because the pace of productivity improvement in telecommunications and microelectronics has been faster than in most--if not all--leading sectors of the past. For an example from telecommunications, look at the creation and spread of network television throughout America beginning in the 1950s. Network television dominated and in some ways still dominates American culture, occupying in its heyday perhaps a fifth of the average American's leisure hours. Something so important in American life should play a large role in the measured economy. But it did not. Calculations of real GDP look at the goods and services that are produced and sold through the market at two points in time. But nobody ever paid a cent to receive network television. So its product received--and still receives--a value of zero in the national income and product accounts. The salaries and profits of the networks, of the production studios, of the actors, of the advertising managers, and of the commercial makers do appear--but they appear as a cost of the production of the goods being advertised, not as an increase in the economic value produced. The growth of broadcast television increased the size of the denominator in productivity calculations, but not the size of the numerator. Each extra one-tenth of one percent of the labor force that shifted to working in the network television industry broadly defined decreased measured productivity and measured GDP per worker by one-tenth of one percent. For an example from network computing, look at the internet: a source of entertainment and information that does not (or does not yet) rival network television, yet shows the same pattern. People pay a toll to telephone companies and to internet service providers in order to access the network. But then the overwhelming bulk of information is free, and is likely to remain free simply because the internet makes information so cheap to distribute and anyone who tries to charge will find their audience going elsewhere. Once again the national income accountants at the Department of Commerce are, when they estimate real GDP, subtracting one-tenth of a percent from American productivity for each one-tenth of one percent of the labor force employed creating and maintaining the world wide web. But there is more substance to claims that we live in a "new economy" than just the observation that Moore's Law is making this particular set of leading sectors larger and faster growing than the average leading sector. There is reason to think that this particular leading sector may have broader consequences in the very long run than other leading sectors in the past. For the first time since the invention of printing in the fifteenth century, the economy's leading sector focuses on information processing and distribution. Now previous leading sectors changed the conditions of life of relatively small, select groups. The automatic loom of the nineteenth century bankrupted those who wove cloth in their homes--the handloom weavers--and transformed the weaving business from one in which the entrepreneur rode from village to village dropping off yarn and collecting cloth to one of factories and steam engines: the dark satanic mills of the nineteenth century. But it left the conditions of life of others largely unchanged (save that clothes suddenly became much, much cheaper). But today's leading sectors appear to be changing the conditions of life of those who use information to direct enterprises (managers), who process information in their jobs (white-collar workers), and who use information to decide what to buy (consumers) (Cairncross (1997)). Moreover, "information" has to be defined very broadly. As Danny Quah of the London School of Economics puts it, when you think of a typical "information good" you should think of a pair of Nikes. The Nikes may not look like an information good: they look like a pair of shoes, made of leather, cloth, and plastic rather than of bits of information. But what does the consumer see when he or she looks at or wears the Nikes? They are shoes, yes, but more important to the consumer--the source of more value, utility, satisfaction--is that the Nikes are a symbolic link to the basketball star Michael Jordan. People have always valued symbolic links to those they admire and respect. Medieval monks would surround the bones of Saint Aldhelm with gold and precious jewels, and display them for the populace to venerate. The touch of the King of England was supposed to cure scrofula. This desire for symbolic links to those we admire and respect is very human, and modern information and telecommunications technologies enable the Nike Corporation to mass-produce such symbolic links to Michael Jordan. Some people read biographies or articles or accounts of games. Others wear shoes. Thus Nike is as much an information-age company as is Microsoft, or Bloomberg, or Amazon. And the fact that this particular set of leading sectors is revolutionizing the production and distribution of information means that the "next economy" may well be genuinely different. For more than two centuries, economists have been preaching that the Invisible Hand of the market will carry us all to wealth, luxury, and utopia as long as a few conditions are met. The government is smart enough to protect property rights, enforce contracts, and tweak the system around the edges to deal with "externalities," in which one person directly benefits (or suffers) from another's actions. And such reliance on competitive markets to produce economic growth and prosperity has, by and large, proven a good bet. But there are three preconditions that must be met if the market's Invisible Hand is to do its job (see DeLong and Froomkin (1997)). The commodity must be excludible, so that its owner can easily and cheaply keep others from using or enjoying it without his or her permission. It must be rival, so that if I am using it now, you cannot be. And it must be transparent: purchasers must know what they are buying. Commodities that take the form of single physical object--hammers, cars, steaks--are rival and excludible by nature. If I am using my hammer, you are not. The process of distribution made transparency straightforward as well: if I am buying this car from this showroom, I can see it, touch it, drive it, and kick it before writing the check. But if a the commodity is not "excludible"--if I the owner cannot block you from using it--then my relationship to you is not the relationship of seller to buyer, but much more that of a participant in a gift exchange: I give you something, and out of gratitude and reciprocity you give something back to me. Think of an economy run like a public radio pledge drive. There is no reason to presume that commodities will get into the hands of those who value them the most. If a commodity is not "rival," then the market will not set its price correctly. If my using it does not keep you from doing so, then there is a sense in which its price should be zero: things should cost money only if they use up society's scarce resources. But at a price of zero there is no revenue to cover the initial cost of production. If a commodity is not transparent, then markets may fail completely. Think of a private individual trying to buy health insurance: the insurance company responds with: "Why do you want this? You must be sick in an expensive way. It raises the price until only the truly and expensively sick want to try to buy insurance, and the market breaks down. Thus if these three conditions are not met, then there is every reason to think that the Invisible Hand will not work well. But the information technology and communications revolutions are eroding these pillars that have supported the Invisible Hand. Words distributed in electronic form (and, with improvements in scanner technology, words distributed in books and magazines as well) are becoming non-excludible. Information goods are by definition non-transparent: if you know what the piece of information is that you are buying, you don't need to buy it. Software is becoming non-transparent as well: when you purchased Microsoft Word or Wordperfect for the first time did you fully realize that you were committing yourself to a long-run path of upgrades and file-format revisions? Computerized words, images, programs are nonrival: the file doesn't know whether it is the second or the two-thousandth copy of itself. How far will this process extend? Will it confine itself to a relatively small set of e-goods, or will it extend to cover the rest of the economy as well. The case of Nike suggests that the definition of "information goods" will turn out to be remarkably broad: a symbolic link to Michael Jordan is certainly a non-rival good that costs next to nothing to produce. On the other hand, ferocious use of intellectual property law has--so far--kept others than Nike from drawing the same sort of symbolic links. But on still another hand, Apple Computer is having no problems constructing symbolic links between its translucent teal computers and Amelia Earhart. It is possible that the ultimate scarce resource will turn out to be people's attention, and that the true source of economic value in the twenty-first or twenty-second century will be (a) providing an experience compelling enough to gain consumers' attention, and then (b) selling that attention to others. That would truly be a new economy. But we cannot now know how it would function. References ---------- Francis Bacon (1605, reprinted 1997), The New Atlantis (New York: Kessinger: 1564592308). Edward Bellamy (1887, reprinted 1989), Looking Backward 2000-1887 (New York: New American Library: 0451524128). Alan Blinder (1997), "The Speed Limit: Fact and Fancy in the Growth Debate," The American Prospect, September-October 1997. Michael Boskin et al. (1997), "The Boskin CPI Commission Report," Journal of Economic Perspectives (). George Boyer (forthcoming), "The Communist Manifesto in Historical Perspective," Journal of Economic Perspectives (). Frances Cairncross (1997), The Death of Distance (Cambridge: Harvard Business School Press: 0875848060). Robert X. Cringely (1996), Accidental Empires (New York: Harper Collins: 0887308554). J. Bradford DeLong (1997), "Old Rules for the New Economy," Rewired 12.09.1997 J. Bradford DeLong and Michael Froomkin (1997), "The Next Economy?" (Cambridge: Kennedy School of Government). Benjamin Disraeli (reprinted 1993), Sybil, or the Two Nations (London: Penguin: 0140431349). Friedrich Engels (1844, reprinted 1968), The Condition of the Working Class in England (Palo Alto: Stanford University Press: 0804706344). Dirk Hanson (1982), The New Alchemists (New York: Little, Brown: 0316343420). Aldous Huxley (reprinted 1989), Brave New World (New York: Harper Collins: 0060809833). David Hounshell (1984), From the American System to Mass Production (Baltimore: Johns Hopkins University Press: 080183158x). Kevin Kelly (1997), "New Rules for the New Economy," Wired 5.09 (September). Kevin Kelly (forthcoming), New Rules for the New Economy. Krishan Kumar (1978), Prophecy and Progress: The Sociology of Industrial and Post-Industrial Societies (Harmondsworth, Penguin Books: 0713911468). Paul Krugman (1997), "Speed Trap: The Fuzzy Logic of the New Economy," Slate Karl Marx and Friedrich Engels (1848, reprinted 1985), The Communist Manifesto (New York: Viking: 0140444785). William Nordhaus (1997), "Do Real-Output and Real Wage Measures Capture Reality?: The History of Lighting Suggests Not," in Timothy Bresnehan and Robert Gordon,eds., The Economics of New Goods (Chicago: University of Chicago Press: 0226074153). Daniel Raff and Manuel Trajtenberg (1997), "Quality Adjusted Prices for the American Automobile Industry," in Timothy Bresnehan and Robert Gordon,eds., The Economics of New Goods (Chicago: University of Chicago Press: 0226074153). Annalee Saxenian (1996), Regional Advantage: Culture and Competition in Silicon Valley and Route 128 (Cambridge: Harvard University Press: 0674753402). Stephen Shepard (1997), "The New Economy: What It Really Means," Business Week, November 17, 1997 Alexis de Tocqueville (reprinted 1991), Democracy in America (New York: New American Library: 0451628012). Hal Varian and Carl Shapiro (forthcoming), Competitive Advantage in the Information Economy (Cambridge: Harvard Business School Press).