|Home | Search | Site Map | Help|
Following almost sixty years of strong growth since 1913, US labour productivity growth slowed significantly to a little over 1 per cent a year from the early 1970s to the early-1990s. The earlier productivity surge from deploying the great inventions of the late 19th century seemed exhausted.
However, since 1995 the US has experienced a sustained return to very rapid productivity growth, rising incomes, low unemployment, and low inflation.
Economic analysts attribute an important role in the acceleration of US productivity growth to the diffusion throughout the economy of computer and related communication technologies. Economic historians identify ICT as a `general purpose technology' with the potential for pervasive, productivity-enhancing applications throughout the entire economy, analogous to the application of electricity at the beginning of the last century - see Box 1.
These developments represented such significant changes in economic trends that US economic commentators, academics and officials speak of a `new economy'.
Most innovations are incremental, involving steady improvements within the framework of existing technologies. But there is a special class of `radical technologies', which could not have evolved through incremental improvements in the technology that they displace (for example, synthetic fabrics or the transistor). Within the category of radical technologies, economic historians have identified a very small group of `general purpose technologies' that have great economic significance because they can be applied to a broad range of sectors within the economy, and have many complementarities with other, existing technologies.
Identification of general purpose technologies is in part subjective, but on one reckoning, there have been only about a dozen general purpose technologies in the history of modern humans (that is, about 40,000 years): the domestication of crops; the domestication of animals; bronze; iron; the water wheel and windmill; the three-masted sailing ship; the printing press; automated textile machinery; the steam engine; electricity; the internal combustion engine; and the computer.
The first industrial revolution, the then-unprecedented burst of high productivity growth in the United Kingdom (UK) from 1760 to 1830, arose in part from the diffusion of two general purpose technologies, the steam engine and automated textile machinery. The second industrial revolution in the US from 1913 to 1972 also arose in part from the diffusion of two more general purpose technologies, electricity and the internal combustion engine, together with other major inventions of the late 19th century. (Economic historians have used the label of `industrial revolution' to refer to any great acceleration of output and productivity growth that is pervasive and economy-wide.)3
Getting the most from general purpose technologies requires extensive corporate re-design, often involving extended trial and error and `learning by doing'. For example, the full benefits from the electrification of factories required a completely new conception of factory layout, of job design, of training to convey relevant skills, and of management of workers and workflow, as epitomised in Henry Ford's production line.4 So slow and demanding was the process of diffusion that the productivity pay-offs from the application of electricity and the internal combustion engine, both general purpose technologies of the late 19th century, were still being felt in the middle of the 20th century.
The computer (used here as shorthand for the related group of ICTs including the Internet and the World Wide Web (WWW)) has been the 20th century's only new general purpose technology.
The US achieved a major lift in economic performance in the second half of the 1990s:5
Changes of just a percentage point or two in single-digit growth rates in productivity and GDP might seem small. But their compounded effects on American real incomes and wealth have already been significant.
`Thus the United States today is some 10-15 per cent richer than mainstream economists would have dared to forecast a decade ago. It has an unemployment rate - a hair more than 4 per cent - that is two percentage points lower than mainstream economists would have dared to forecast a decade ago. And it has a much more favourable short-term inflation-unemployment trade-off than the US economy had a decade ago, when a decline in unemployment below 6 per cent set off increases in inflationary pressures reminiscent of the late 1960s or the late 1970s.'6
The technological key to this strong US performance in the second half of the 1990s has been the rapid rise in ICT investment, together with the intense competition that created the incentive for that investment, and the macroeconomic policy framework that created a favourable investment environment and low interest rates.
This rapid investment in new technology was made possible by the highly flexible nature of product and labour markets in the US. The lack of labour market rigidities in the US, particularly in relation to employment dismissal laws, has allowed existing firms to freely and efficiently substitute labour for capital in areas where there is potential for such decisions to improve the production process. This regulatory framework sends a clear signal that innovation and risk taking is encouraged by reducing unnecessary administrative regulations and instilling positive attitudes towards entrepreneurship. This has been a major factor behind the emergence of the US as a centre for ICT development. In turn, this creates an environment conducive to the creation of innovative startup firms, which contributed to the rapid employment growth seen over the same period.
New technologies that permitted the personal computer (PC), cheaper telecommunications and the Internet have been available for many years. However it was only in the 1990s that diffusion of the Internet and the WWW, accelerating falls in ICT prices and the emergence of user-friendly computer software, has allowed these technologies to interact on an unprecedented scale, and has allowed ordinary people and small businesses to use them productively - see Box 2.
Box 2: The evolution of information and communication technologies
From at least the onset of the 1972 to 1995 productivity slowdown in the US, economists had pondered what became known as the `Solow paradox': the impact of computers could be seen everywhere except in the productivity statistics.7
A timeline of significant commercial computer and ICT developments helps explain the technological basis to the very slow economic process of diffusion of ICT throughout the economy. The technological foundations were laid in the 1950s for commercial computing, and even for business-to-business (B2B) e-commerce (albeit through expensive, closed and inherently anti-competitive proprietary systems, rather than the cheap and open Internet). But the economic diffusion of ICTs in large, valuable networks that could link most businesses and many households, awaited the remarkably recent development of low-cost, high-powered computers, low cost telecommunications, the Internet and the WWW, and - perhaps most importantly of all - software that made it easy for ordinary people to use computers:
1951 UNIVAC I, first commercial mainframe computer goes on sale.
1957 SABRE, first proprietary B2B (originally airline reservations within American Airlines, subsequently made available to other airlines).
1980 First spreadsheet software.
1981 First IBM personal computer.
1983 Transition to TCP/IP protocols create the current Internet.
1984 Apple Macintosh `point and click' interface.
1992 WWW software and protocols launched.
1995 Microsoft Windows 95 makes `point and click' available to PC users.
The melding of all these advances into large networks was critical to their economic value. Economists analyse e-commerce and the Internet in terms of `direct network economies'. A direct network economy arises in two-way communications networks, where each new customer increases the value of the network to all previous customers. Users of the network receive increasing returns in consumption. For example, the English language is a communications network, and every additional speaker of English increases the value of the language to all existing speakers. So are the TCP/IP protocols on which the Internet rests. So are all applications of the Internet, such as B2B exchanges.
The 2001 Annual Report of the Council of Economic Advisers (CEA)8 marshals analysis of the last few years in attributing the pick-up in productivity growth to a combination of mutually reinforcing advances in ICT, business practices and economic policies. This need for competitive pressures and good policies to create a positive investment climate has also been important in the previous diffusion of other general purpose technologies.
For example, the original spread of electric power and light through the US economy did not acquire real momentum until the 1920s, although central generating systems for electric power appeared in New York and London in the 1880s. The diffusion of electric technology did not proceed until the fall in regional utility rates for electricity which followed deregulation in 1914 to 1917 (analogous to the accelerated fall in the prices for computers in the mid-1990s). Large investments in both equipment and new human skills were necessary to reap the ultimate advantages of electrification, so a good overall investment climate and a low cost of capital was necessary. This was provided in the expansionary macroeconomic climate of the 1920s.
As with many new general purpose technologies, there was a long initial period of business `learning by doing' and trial and error in getting the most out of electricity, and the first lessons from isolated successes could not be communicated efficiently to the diffuse range of potential beneficiaries.
Competition has been an important factor in driving economic advances from ICT, both within the ICT-using industries and in the ICT industries themselves:
`... in 1999, some 60 per cent of US spending on computers was sourced from imports, while some 50 per cent of domestically produced computers were exported. International competition has reinforced competition at home. ... Antitrust laws limit corporate conduct that undermines competition and consequently harms consumers. ... Regulatory policies have also promoted competition.'9
US research has extensively explored the extent to which accelerated productivity growth has arisen within the ICT sectors themselves, compared to the productivity gains from the use of ICT in the broader economy. Estimates differ in detail because some are constructed from the production side of the national accounts, some from the consumption side and some from the average of the two. Estimates also differ in the amount of the overall productivity acceleration they estimate to be structural, rather than cyclical. But almost all tell the same broad story. The CEA Report is representative in estimating that, of the 1.6 percentage point acceleration in annual productivity growth:
So the US productivity surge has largely been the result of the productivity-enhancing application of ICT breakthroughs by rapidly restructuring companies, driven by competitive markets and a macroeconomic environment supportive of high investment and low interest rates.
Between March 2000 and April 2001, the ICT-heavy NASDAQ index fell by around two-thirds of its total value. This extraordinary volatility bears the hallmark of previous stock market `bubbles' with a rapid run-up in values beyond any supporting data on earnings potential, followed by an equally rapid collapse without any obvious trigger. Has the ICT boom run its course, and might its contribution to the US productivity acceleration already have passed?
Some bubbles in asset markets arise merely from excesses of credit, of greed or of both. But most have at their base some plausible hope for profit from burgeoning demand for a new product (such as for tulips in 17th century Holland), or profits from resources from a new frontier, or a new politically-guaranteed monopoly (such as the South Sea bubble in 18th century England), or from a new technology. In short, bubbles frequently arise when the fundamentals become difficult to assess - see Box 3.
Box 3: Stock market `bubbles' and previous radical or general purpose technologies
Most major technological innovations have caused a `bubble' in the prices of equities in associated companies.
There are similarities in the recent enthusiasm for ICT stocks to the railway age from the 1840s to the 1890s. In the UK, vast sums were raised on the stock market to finance new railway lines, but `...most railway companies never paid a penny to shareholders, and many went bust, largely because over-investment created excess capacity. The Great Western Railway was for decades the most admired railway company in Britain, yet anyone who had bought shares at its launch in 1835 (at a fraction of their peak in 1845) and held them until 1913 would have seen an annual return of only 5 per cent. Even so, the railways brought huge economic benefits to the economy long after share prices crashed.'10
In the US alone, there were once 5,000 railway firms, almost all of which have now disappeared. The real beneficiaries of the US railway boom were the small firms and farmers who benefited from the opening up of the continent.
There are also similarities to the more recent productivity surge from the 1890s to the 1950s driven by the diffusion of the late-19th century's two general purpose technologies, electricity and the internal combustion engine, together with other radical technologies such as radio, sound recording and movies, aeroplanes and industrial chemistry.
Profits and share prices of the early electricity firms were disappointing. During the electrification of American industry, profits actually fell slightly as a share of GDP, as competition drove manufacturers to pass cost savings through to consumers. A few of the `new economy' firms at the beginning of the last century, producing the technologies of the electric age, prospered and have survived into the computer age (such as the archetypical `General Electric', the only company listed in the Dow Jones Industrial Index today that was also included in the original index in 1896). But references to many other firms and their brand names are now to be found only in the Smithsonian museums. In contrast, the whole-of-economy beneficiaries of the electric age are everywhere.
In the US alone, there were initially 2,000 automobile firms; now there are three, all with ownership links or corporate collaborations beyond US borders. The other firms were all driven out, over the years, through consolidations, while the quality, diversity and cost of motor transport options has improved rapidly.
The more radical or potentially pervasive the new technology, the more difficult it is to reasonably foresee just which firms might profit from it, when, or by how much.
With a radical or a general purpose technology, it is likely that the initial production of the key goods and services embodying the new technology will be undertaken by a specially created firm with no previous history or earnings from any other source. Existing firms in competing areas using earlier technologies typically have a vested interest in maintaining that technology, and are burdened with the cost of servicing the capital tied up in those technologies.
So, frequently the firms exploiting new technologies defy benchmarking by historical record or the usual tools of the equity market analyst. For example, their share price might reasonably be positive - they have some prospect of producing a future stream of earnings - but their actual earnings are zero or negative, so their current price-earnings ratio is undefined, and their effect on the average price-earning ratio for the equities market as a whole is to move it towards new heights, as has been observed of the US market over recent years.11
With the recent slowdown in the US economy, some slowdown in productivity growth has occurred, as usual cyclical `labour hoarding' occurs. However, that should not be interpreted as the end of ICT-driven productivity growth. There seems to be abundant scope for future ICT efficiencies throughout the economy. For example, e-commerce efficiencies have not yet made much of a contribution to productivity growth. On one estimate of the value of business-to-consumer (B2C) and B2B commerce in the US in 1999, and assuming such sales were 10 per cent cheaper than if they had taken place thorough traditional channels, those resource savings may only have accounted for less than 0.1 of a percentage point of the increase in multifactor productivity (MFP) growth in the second half of the 1990s.
As with earlier radical and general purpose technologies, the greatest gains from future ICT applications are likely to flow to the users, not the producers. As noted in the OECD's Report on the Growth Project:
`ICT is important for growth, but having an ICT-producing sector is not a prerequisite. ... Moreover, only a few countries will have the necessary comparative advantages to succeed in ICT output. The key to benefiting from ICT is to focus on policies to foster its use, rather than its production.'12
For tomorrow's Internet-driven ICT applications in particular, it should be remembered that the Internet is an open system which inherently tends to lower entry barriers and intensify competition. That will be good for companies, workers and consumers who can flexibly change their practices to use the new technologies, but it is unlikely to be a path to riches for many of the investors in the production of the technologies themselves.
3 Gordon, R.J. (2000), Does the `New Economy' Measure up to the Great Inventions of the Past? NBER Working Paper No. W7833, p 17.
4 David, Paul, A. (1990), The Dynamo and the Computer: an Historical perspective on the Modern Productivity Paradox, American Economic Review, Vol 80, No. 2, pp 355-361.
5 The Annual Report of the Council of Economic Advisers (CEA), Economic Report of the President, January 2001
6 Delong, J. B. (2000), What Went Right in the 1990s? Sources of American and Prospects for World Economic Growth, Paper to Reserve Bank of Australia 2000 Conference `The Australian Economy in the 1990s', p 13.
7 Solow, R. (1997), We'd Better Watch Out, The New York Times Book Review, 12 July 1987, p 36.
8 CEA (2001).
9 CEA (2001), p 47.
10 Economist, The New Economy: Untangling E-conomics, Survey, 23 September 2000.
11 Shiller, R. J., Irrational Exuberance (2000).
12 OECD (2001), p 9.