The Importance of the Project
Asking experts of any type to project the future is always hazardous. It is especially tricky when the subject is changing as rapidly as the Internet, which as a commercial proposition is only seven years old (dating from the invention of the browser), and is only just beginning to affect the economy.
Nonetheless, policymakers in both the public and private sectors must do their best to peer into the future. At a September 2000 conference sponsored jointly by Brookings, the Berkeley Roundtable for the International Economy, the Department of Commerce, the Internet Policy Institute, and the Organization for Economic Development, academic experts on various sectors of the U.S. economy were asked to project the likely economic impact of the Internet on their particular sectors, with special attention paid to the impact on productivity growth. Those sectors included automobile manufacturing and sales, non-auto manufacturing, higher education and private-sector training, financial services, government, health care, retailing, and trucking.
Productivity growth matters to policymakers and to all participants in the economy. The Federal Reserve must attempt to forecast the future growth of the economy and prices in order to set monetary policy. A key component of the Fed's projection is how fast it thinks the economy can grow—the potential growth rate of productivity and the labor force—without setting off an acceleration of inflation. Similarly, the president and the Congress must estimate the potential growth rate in order to know its "baseline" surplus (the happy circumstance we now enjoy) or deficit (the problem that plagued us in the 1980s and for nearly all of the 1990s). Wage earners should care about productivity growth because it determines the growth of their compensation, and hence their living standards. For firms making investment plans, it is important to project future productivity growth because that will determine the size of the market in the years ahead.
While the anticipated growth in the labor force is relatively easy to project—because future population growth and labor force participation are, for the most part, known—it is no easy thing projecting the growth rate of labor productivity. Productivity growth trends have fluctuated widely in the post-war era, and these fluctuations have been neither anticipated nor well explained. For nearly three decades after World War II, labor productivity grew at roughly 2.5 percent annually, a pace that enabled the standard of living of the average American to double about every 30 years. From 1973 to 1995, annual productivity growth slowed dramatically, to 1.4 percent. Analysts pointed to various culprits for the slowdown, including higher energy prices (an argument that became less convincing when oil prices fell sharply after the early 1980s), the influx of baby boomers into the workforce, and variable rates of inflation that allegedly deterred investment. But none of the explanations proved very convincing. Then, productivity began soaring, averaging growth of about 3 percent a year since 1995. Virtually no one anticipated that outcome, either inside or outside the government.
Just as we do not fully understand why productivity growth slowed down after 1973, the dramatic pickup since 1995 is also something of a mystery. But not completely. Clearly, heavy investment in computer and telecommunications technology in the 1990s—accounting for as much as a half of all plant and equipment investment in recent years—played a significant role in the recent productivity surge. Although the computer and telecommunications revolutions began earlier, they apparently did not have enough impact on business processes, practices and organization to show up in aggregate productivity growth until the second half of the l990s. The macroeconomic conditions of the late l990s—tight labor markets, low inflation, and fierce global competition—also encouraged firms to use new technologies as a way of economizing on labor while surviving in a fiercely competitive marketplace.
What about productivity growth in the future? Official government forecasts anticipate that the recent rapid pace of productivity growth will not continue. Both the Congressional Budget Office and the Office of Management and Budget project productivity growth slowing to an average of about 2 percent by 2010—faster than the dismal period of 1973-95, but much slower than the pace of recent years and somewhat slower than the growth of the previous "Golden Age" of 1948-73. Optimists point out, however, that the official forecasters missed the 1995-2000 productivity boom, which has been heavily driven by the information technology (IT) revolution, and suggest that the good times will continue to roll. They now pin their hopes not only on the continued, productivity-enhancing impact of IT, but on the Internet revolution in particular, which as revolutions go, is in its infancy.
Currently, the annual volume of e-commerce, estimated to be between $100 billion and $200 billion, is too small in relation to the overall size of the economy to have had much impact on productivity growth. But, all that could and, we believe, likely will change, especially as Internet use becomes more prevalent. Given the recent demise of many of the "dot coms" that symbolized the Internet revolution, it is tempting to think otherwise. But the real power of the Internet will be felt in the existing, or "old," economy, which we project will make increasing use of the Internet to deliver benefits to consumers. Isolating the potential impact of the Internet on productivity is important because even a few tenths of a percent impact on the growth rate could represent a significant portion of any permanent surge in productivity that is maintained in the future.
The Economic Potential of the Internet Revolution
The Internet has the potential to increase productivity growth in a variety of distinct, but mutually reinforcing ways, including:
- Significantly reducing the cost of many transactions necessary to produce and distribute goods and services;
- Increasing management efficiency, especially by enabling firms to manage their supply chains more effectively and communicate more easily both within the firm and with customers and partners;
- Increasing competition, making prices more transparent, and broadening markets for buyers and sellers;
- Increasing the effectiveness of marketing and pricing;
- Increasing consumer choice, convenience and satisfaction in a variety of ways.
How are these impacts likely to be manifested in some of the sectors our authors have examined? Let's find out.
The most important attribute of the Internet also may be the most obvious: it can transmit information quickly, conveniently, and inexpensively. Routine transactions, including making payments, processing and transmitting financial information, and maintaining records, can be handled less expensively with web-based technology. Using Internet technology, many firms, especially those in data-intensive industries such as financial services and medical care, can reduce their cost of production.
Patricia Danzon and Michael Furukawa from the Wharton School at the University of Pennsylvania note that the potential for transactions cost savings from transition to the Internet is especially high in the health care sector, because it is so large (14 percent of GDP), so information-intensive, and so dependent on paper records. Moving health insurance claims processing to the Internet would require aggressive efforts to standardize claims formats, but savings could be large. One of the providers of the current system, Electronic Data Interchange (EDI), alleges that it can reduce the cost of processing from $10-15 per paper claim to $2-4 per EDI claim. Web-based processors may be able to deliver the same service for 2-4 cents per claim. Only about 40 percent of doctor's claims are now processed electronically. If health insurance claims processing were shifted to the web, $20 billion a year could potentially be saved and the process could be faster and more convenient.
The Internet also offers great potential in the area of managing medical records, not only for cutting costs, but for improving the quality and effectiveness of care. Assuming that privacy concerns can be adequately addressed, patients and providers would benefit enormously from converting the current medical records, which are mostly paper, into an electronic medical record in standard format. Providers would then be able to access the patient's full medical history quickly and enter their own observations and treatments.
Across the financial services industry, retail banking is a major source of tension. Financial services based on customer-provider relationships tied to geography and the provider's knowledge of the customer contrast with savings offered by on-line markets for standard financial products. The tension is perhaps most evident in brokerage services, where the Internet has precipitated a split between the relationship-dependent services of investment advising and portfolio management and the standardized service of stock trading. Customers benefit from cheaper trading on-line, assuming they do their own research, forego advice, and shop for lower commissions by searching the Internet. Some of the cost reduction is offset by increased advertising and marketing costs as on-line brokers compete with one other.
In the mortgage lending industry, customers are using the Internet to shop for information and compare rates, although only a tiny fraction of mortgages now originate on-line. But a substantial portion of these mortgages may originate on-line over the next few years as consumers grow more comfortable using "digital signatures," now legal under a law signed by President Clinton in June 2000. If this occurs, consumers will save through lower margins in this part of the business, plus lower costs in processing mortgage applications.
Another source of potentially significant transactions costs savings comes, perhaps surprisingly, from the government sector. As Jane Fountain of Harvard University notes in her study, governments at all levels spend substantial resources answering questions such as where to get services, who is eligible for benefits, and what laws and regulations apply in certain situations. The Internet has enormous potential, now only beginning to be realized, for dispensing information to citizens less expensively and more accurately than telephone inquiries. Filing tax returns or applications for permits and licenses on-line, for example, cuts costs both to the government and to the taxpayer.
The Internet and Efficient Management
The use of the Internet as a management tool may have considerable potential for improving efficiency in many sectors of the economy and may cause significant restructuring of those sectors in the process.
Many of the potential efficiency gains come from use of Web-based technology to manage supply chains more effectively and reduce inventory. These savings may show up within the firm, from better scheduling to information-sharing across the company, or in more efficient interaction with other firms in the supply chain.
Cisco Systems, for example, has been a leader in dealing with suppliers on the Web to enhance efficiency of its procurement. In the process, the company has changed the definition of what it means to be a "manufacturer," since it outsources most of its manufacturing operations to other companies in its Internet-based supplier community.
The Internet is also being used effectively in other industries to link partners in joint enterprises across large distances, enabling them to share production schedules and integrate their operations. Charles Fine of MIT and Daniel Raff of the University of Pennsylvania examine the automobile industry, and find numerous potential opportunities for Internet-aided increases in efficiency. They project productivity improvements in product development, procurement and supply, and in various aspects of the manufacturing process itself, and suggest the applicability to automobiles of the "Dell model," under which customers specify exactly what features they want and buy a product that is built to suit their tastes. Fine and Raff suggest that the Internet-driven automobile sector of the future should involve far fewer dealers and salesmen than it does now.
Making Markets More Competitive
One of the major features of the Internet revolution is its potential to make the whole economic system, nationally and internationally, more competitive. If prices of well-specified goods and services are available on-line, buyers can shop for the best deal over a wide geographic area and sellers can reach a larger group of buyers. The Internet could bring many markets closer to the economists' textbook model of perfect competition, characterized by large numbers of buyers and sellers bidding in a market with perfect information. The results should be lower profit margins, more efficient production, and greater consumer satisfaction.
Some analysts have counted the anticipated lower profit margins as productivity gains. However, unless more competition translates into lower costs, squeezing profit margins will not produce more output per unit of labor input—or higher productivity. Instead, lower profit margins by suppliers will help the companies which purchase from them, which in turn, in a competitive market, will pass those savings on to consumers. The net result is a transfer of income from producers to consumers—a process to be welcomed—but not counted as higher productivity.
At the same time, it should not be overlooked that in the longer run, the increasing transparency of prices and the widening reach of markets provided by the Internet is likely to be a continuing global force for greater efficiency. The Internet is a great equalizer, lowering barriers to competition by increasing the chances that a company anywhere in the world that develops a better product can win bids, sell products, and force other competitors to improve their productivity.
Increasing Choice and Convenience
Given the prominence of a few Internet retailers, such as Amazon.com and buy.com, and the exponential growth of retail Internet sales from a tiny base, one might have expected analysts to project significant increases in retail competition and productivity.
Joseph Bailey of the University of Maryland, however, finds retail Internet sales insignificant—about 1 percent of retail sales at the beginning of 2000—and unlikely to surpass 10 percent of the total in the foreseeable future. He points out that Internet and conventional retailers are becoming more like each other, with pure e-tailers acquiring physical warehouses and distribution facilities, and conventional retailers using the Internet as a supplement to sales in stores and through catalogues. He predicts that the "hybrid" retailer will become the dominant type, but does not foresee a significant impact on productivity.
Adding It Up
After examining the accumulating evidence in the eight sectors discussed at the September conference, we have concluded that:
- The potential of the Internet to enhance productivity growth over the next few years is real;
- The greatest impact may not be felt in e-commerce, but rather in a wide range of "old economy" arenas, including health care and government;
- As a result of the Internet, there is considerable scope for management efficiencies in product development, supply chain management and a variety of other aspects of business performance;
- Enhanced competition will also enhance productivity, but these benefits should not be exaggerated. Many of them are likely to show up in improved consumer convenience and expanded choices, rather than in higher productivity and lower prices.
But what does it all this add up to? How much overall productivity improvement is likely to come from the Internet? It is too soon to offer even a serious guess. Moreover, most of the potential savings will be impossible to attribute solely to the Internet. For example, to the extent that the Internet simplifies and speeds up interactions between companies and governments, some of the benefits will make their way into private sector productivity growth, although these benefits will be hard to trace directly to the Internet.
However, a study by Andrew McAfee of Harvard Business School calculates the potential impact of the Internet on the manufacturing sector in particular by drawing on the example of Cisco Systems, which estimates its cost savings in excess of 5 percent. Even making the conservative assumption that Cisco remains reasonably well ahead of the rest of the manufacturing sector, McAfee projects cost savings for manufacturing of between 1 and 2 percent over the next five years, which could translate into annual productivity improvements in this sector alone of 0.2-0.4 percent. By the time they are fully realized, these estimated savings could fall between $50 billion and $100 billion.
It would be natural to expect significant variations across industries in cost savings. Charles Fine and Dan Raff estimate the potential costs savings in the automobile industry alone (exclusive of dealership savings) to be about 11 percent, although they caution that these savings are likely to be realized at one time rather than in continued improvements. Even so, as long as they do not occur simultaneously in all firms, the gradual one-time improvements nonetheless should boost measured productivity growth in that industry during the period of adjustment.
However large the productivity savings turn out to be—and the preliminary estimates suggest the economy-wide cost savings could easily exceed $200 billion annually—there should be no mistake about who the ultimate beneficiaries will be: consumers, not businesses. In our highly and increasingly competitive market, most temporary market advantages, reflected in high profit margins, get eliminated as competition drives prices toward marginal costs (unless firms accumulate and then abuse their monopoly positions).
Caution from the Skeptics
Not all economists who have been studying the impact of the Internet and the IT revolution agree that the Internet promises benefits anywhere close to those that have just been suggested. Indeed, they raise several warning flags and it is important to understand the basis for them.
First, as everyone concedes, e-commerce still accounts for a relatively small part of overall U.S. economic activity. Business-to-consumer (B2C) transactions total about $20 billion annually, or less than one percent of total retail sales. Business-to-business (B2B) transactions are much larger, in the $100 billion range, but they also account for a small fraction of the $6 trillion in private sector economic activity. Even if the sharply increased totals for both B2C to B2B transactions are realized—in excess of $1 trillion for both combined in several years—they still will remain a relatively small part of the overall economic landscape. The critics say it is unrealistic under these circumstances to expect giant gains in either measured productivity or non-measured consumer welfare any time soon.
Second, it is difficult to know the baseline or benchmark against which to measure the impact of the Internet. How can we be certain how productive firms would be in its absence? The short answer is: we cannot. Put differently, even if the Internet makes an identifiable contribution to productivity growth, it is quite possible that this will not produce an acceleration in the growth that is already occurring.
Third, the skeptics can point to some empirical evidence that appears to back up these cautions. For example, using research developed by Brookings scholars Jack Triplett and Barry Bosworth, the September 2000 edition of The Economist compared the intensity of investment in information technology in various industries with their growth rates of "total factor productivity," or TFP, which measures the increase in the ratio of output to the sum of capital and labor inputs. This study found essentially no correlation. Indeed, certain industries where IT investment was especially strong in relation to total output—education and banking, for example—had either low or negative growth in TFP. While this result may be due in part to mismeasurement of these industries' output, the study does caution against assuming large economic gains resulting from the Internet until they are demonstrable.
Northwestern University economist Robert Gordon, one of the world's leading experts on productivity growth, is even more pessimistic. Not only does Gordon claim that the benefits of the IT revolution have been largely confined to the IT sector itself, but he further argues that the Internet ranks relatively low, at thirteenth place, in the pantheon of major innovations of the past century.
While the skeptics provide useful cautions to the exuberant claims of the Internet's economic potential, we are not so pessimistic. The economy has been running near or at its productive capacity throughout the post-1995 period when productivity has grown so rapidly. Indeed, it is at the end of cycles that one normally expects productivity to fall rather than increase (because less productive workers presumably are being drawn into the labor force). The fact that productivity growth has accelerated during the past five years suggests, at least to us, that the improvement has not been cyclical, but rather stems from structural changes in the economy. Still, it remains to be seen whether those structural changes will generate future productivity growth rates similar to those of the recent past.
Realizing The Potential of the Internet
What will determine the extent to which the potential benefits of the Internet—both the quantifiable improvements to productivity and the less quantifiable benefits of convenience and quality improvements—will, in fact, materialize? In part, the answer depends on how rapidly use of the Internet spreads throughout the rest of the population (the "width" of the Internet revolution). Large and medium-sized businesses are likely already using the Internet to some extent, although not necessarily extensively or effectively. The spread of Internet use to small establishments will likely be rapid, especially if governments foster Internet use, by favoring or requiring on-line tax filing, for example.
With respect to individuals, however, the Internet is still is very much at the center of a "digital divide" that isolates low-income families and communities from the benefits of the new economy. As tools to access the Internet, including cellular phones and personal computers, become cheaper, the divide will gradually close. Efforts to equip all schools and many community organizations with the new tools and to teach the skills needed to use them will erase the divide faster.
Increasing the width of the Internet revolution will help ensure that its benefits are widely shared. How great those benefits are depends in part on whether potential productivity gains of the Internet identified at the conference turn out to be as "deep" as they project. Will each of the sectors actually use the Internet more intensively and, in the process, change business practices?
The answers will depend heavily on the adjustment of individuals within these organizations and on changes in corporate culture. If past experience with information technology investments is any guide, this is a serious qualification. The business landscape is littered with many examples of poorly planned IT projects. Fortunately, competition should supply a positive countervailing force, provided that markets are not unduly constrained by anticompetitive practices, and that the antitrust laws are enforced. Firms realizing projected gains will increase market share at the expense of those who are not, thereby speeding the spread of Internet use by companies that are losing potential profits. But the improvements will still take time, and we cannot be confident at this point of the pace at which the process will play out.
What about the impact of broadband hookups to the Internet, including faster cable, telephone (DSL, or Digital Subscriber Line), wireless, and perhaps one day, direct fiber optic connections? Shouldn't the increase in speed accelerate the projected productivity impacts? Surely, broadband has already made life easier for the approximately 3 percent of the population who have it. But to the extent that the benefits of B2C already consist largely of non-quantifiable gains that are not reflected in the national income accounts, the benefits of broadband will be no different in character.
One possible exception may be for the millions of Americans who telecommute, or will in the future. To the extent that workers are more productive at home than in the office, as some anecdotal evidence suggests, business-quality speed at home could help improve measured productivity. But as for businesses themselves, most already have or will soon have faster connections to the Internet. Thus, the diffusion of broadband access to residential customers should not affect the gains they realize through B2B transactions, except indirectly, through the added competitive retail pressure that increased use of the Internet should make possible.
Finally, what about the role of policy in helping to realize the projected benefits of the Internet? In theory, if more people had confidence in the security and privacy of their e-transactions, then e-commerce would grow faster, thereby increasing the benefits from its use. However, the benefits of more intensive retail use of the Internet are likely to be non-quantifiable and thus not evident in the productivity statistics. At the same time, if the current controversy surrounding taxing Internet purchases resulted in forced collection on such transactions, it would slow the growth rate of e-commerce. But it would also help rectify an imbalance between the treatment of off- and on-line commerce, and thus probably would not have a significant material impact on overall productivity.
However, a different story might emerge if new laws, or even better, enhanced technology to secure intellectual property rights for music, videos, books, and software, encouraged more Internet-based distribution of content. This would reduce distribution costs and would improve productivity because the outcome would show up in the national income account data.
The economic impact of the Internet will likely not be as insignificant as the pessimists claim, and not as overwhelming as many cyber-enthusiasts suggest. Still, there are reasons for believing that its effect on the economy will be important. The Internet will produce significant cost savings in many sectors of the economy, resulting in faster productivity growth. It will also produce lower prices for consumers, resulting in faster growth in living standards.
The Internet should also generate a variety of benefits to users, in their capacities as consumers and citizens, that are not easily quantified but nonetheless real: savings in time, added convenience, and products and services tailored specifically to them. In an era when consumers are aware more than ever that "time is money, these benefits to many may be more noticeable and appreciated than the numbers that economists like to count.