|
If The Boffins Say It’s So Then It Must Be True
Has the productivity of workers increased markedly in recent years? According to official statistics, productivity in Australia and the U.S. increased significantly during the closing years of the twentieth century. In this country, according to the head of general research at the Productivity Commission (quoted in The Australian Financial Review on 30 June 1999), “a strong productivity surge is evident in Australia in the 1990s. New Australian Bureau of Statistics estimates show productivity growth to be running at 2.4 per cent a year compared with an historical average of 1.2 per cent a year. The estimates put productivity growth at a record high – higher than in the Golden Age of growth in the post-war period.”
As a result, according to many economists and central bankers (and most politicians, finance journalists and market participants), a quickening pace of profitable economic activity, steady or falling prices of commodities and consumer goods and low unemployment and interest rates, which were once regarded as mutually-exclusive phenomena, can now co-exist. This co-existence, in turn, allegedly underwrites the admittedly high but nevertheless reasonable (to them) prices of most financial assets. Further, the rise of the Internet and anecdotal information about the impact of computers in the workplace has prompted many (Robert Barro, Alan Blinder, Robert Gordon, Daniel Sichel and Paul Strassman dissent to various degrees) to infer that rapid advances in and extensive applications of information technology both underlie and will continue to bolster this allegedly higher level of total workforce productivity. It is not just that the rapid development of IT and growth of productivity are correlated; it has been asserted that the former has caused the latter. Indeed, during the late 1990s it was common to hear that a once-in-a-lifetime structural shift of economic behaviour driven by productivity-enhancing It – a New Era, if you please – was occurring. This shift was lessening or removing entirely the “speed limits on growth” and thereby releasing Prometheus from his chains. This clarion-call, which would leave in its wake those who would not recognise or adjust to it, has been muted but not silenced by the tech wrecks of 2000-2001.
The measurement of worker productivity is clearly a matter of more than mere academic interest. “New Era” thinking has exerted and continues to exert much influence upon American (and to a lesser extent Australian) policymakers. In the U.S., the Federal Reserve has referred repeatedly to productivity and its growth as a consequence of an acceleration of the pace of technological development. On 15 June 1999, Dr Greenspan stated that “an economy that 20 years ago seemed to have seen its better days is displaying a remarkable run of economic growth that appears to have its roots in ongoing advances in technology … The productivity growth seen in recent years likely represents the benefits of the ongoing diffusion and implementation of a succession of technological advances … Likewise, the innovative breakthroughs of today will continue to bear fruit in the future.” Dr Greenspan’s view is steadfast. On 14 November 2001 he stated that “the long-term outlook for productivity growth, as far as I’m concerned, remains substantially undiminished. It’s one of the reasons why, even though we’re going through a very trying period in the last year or so, the outlook in my judgment looks to be extraordinarily good.”
In Australia, the Productivity Commission has been more cautious (and has pointed to reforms of labour laws, regulations and other micro-economic phenomena as more important causes of the growth in Australian productivity), but there has been no shortage of private sector commentators willing to draw the inference between IT and productivity growth on the one hand and the virtuous circle of low consumer goods price inflation, unemployment and interest rates on the other. In both (and indeed other) countries, market participants have found the notion of a productivity-driven New Era to be irresistibly appealing. It flatters CEOs because it implies that the cost-cuttings and sackings they undertook during the early 1990s have yielded demonstrable and permanent gains. Brokers and advisers find it attractive because it provides a basis for their picks’ often-sky-high prices. Indeed, passages in the AFR, Financial Post and Wall Street Journal during November 2001 indicated that strategists and analysts continue to base their aggressive expectations of future growth of earnings – and the nosebleed multiples they are still prepared to advise that their clients pay for many assets – at least partly upon the assumption of high and growing rates of workforce productivity.
Are workers becoming more productive? The answer to this question has momentous consequences. Leithner & Co. takes an agnostic (and, in today’s chastened but still very confident climate, contrarian) view. Important as it is, and much as we would like to know it, we do we not know and in principle cannot know the answer to this question. This circular reasons from first principles to a disconcerting conclusion: notwithstanding their pretensions to scientific rigour and accuracy, the methods used to measure productivity are at best questionable and at worst invalid. These methods also have an unintended and largely unremarked consequence: in recent years they have implicitly encouraged aggressive growth of credit not backed by savings. The result is that interest rates have been forced below market levels for far too long; and as detailed in other circulars (e.g., Interest Rates, Corporate Debt and the Business Cycle and Is Australia Really a Low-Inflation Country?) from this sin flow a multitude of unpalatable and still largely unrecognised consequences.
Measuring Productivity: The Basic Difficulty
An increase in the rate of productivity, if it is maintained over an extended interval of time, indicates that a worker is able to produce greater and greater amounts of goods and services per hour worked (or, equivalently, that fewer and fewer hours are required in order to produce the same amounts of goods and services). Efficient and effective workers are a necessary condition for a profitable company. Accordingly, lest another employer attract such individuals by offering them higher wages, the more productive the worker the greater his value to an employer and the higher his rate of pay. The greater the number of productive workers and the higher their productivity, the larger the benefits to workers, companies and customers. A productive worker, then, is a well-paid worker; and a well-paid worker enjoys a higher material standard of living.
Productivity is an individual-level phenomenon. Individuals, not industries or nations, produce goods and services. Yet government and other statistics virtually never deploy it at the micro (i.e., individual or workplace) level. Instead they almost invariably almost apply it at an aggregated (i.e., industry, state/provincial or national) level. A nation’s rate of productivity during a given interval of time is typically defined as the ratio of total output to total number of hours worked. In the U.S., for example, the annualised rate of growth of output per worker-hour was approximately 3.5% (and as much as 4.5%) during the 1960s. It fell erratically but consistently throughout the 1970s and reached a nadir of 0.0% during the recession of the early 1980s. It then recovered rapidly to ca. 2.0-2.5% during the mid- 1980s; fell to ca. 0.5% during the recession of 1990-1991; increased to 1.5% or so during the early and mid-1990s and increased during the closing years of the century to a rate of ca. 2.5% per annum.
In order to calculate output per worker-hour one must typically first estimate total output and then estimate the number of hours worked in order to produce that output. Clearly, in order to calculate a rate of productivity one must combine at least two (and in practice far more than two) sources of data. Equally clearly, these data can be combined only on the basis of some common unit of measurement. To add one apple plus one apple is a trivial matter; but to add one apple and one orange makes sense only if we are prepared to regard each item as a piece of fruit. Absent this common unit, output data cannot be combined and an estimate of productivity cannot be calculated. Alas, “the economy as it is commonly and glibly reified, comprises a vast and ever-changing array of raw, intermediate and final commodities, products and services. A moment’s thought makes apparent that we cannot simply add them. There is no unit of measurement, in other words, that is common to lambs raised and bales of wool processed; civil engineering, asset management and IT services undertaken; hair cut, dyed, styled and dried; music performed, recorded and broadcast; and motor cars manufactured, sold and repaired, etc.
A further complicating factor is the constantly changing array of goods and services comprising the modern structure of production. To cite but one example, to what extent is the computer which was state-of-the-art six months ago comparable to its current counterpart? By what criterion does one make this comparison? It is not the vast variety of the goods services emitted by the structure of production that causes this measurement problem; rather, it is the basic incommensurability of any two items on this vast list. Accordingly, the basic difficulty of calculating total output – and hence rates of productivity – is just as apparent on Robinson Crusoe’s island as it is in contemporary Australia. How does Crusoe add the results of one hour spent digging a well and one hour spent harvesting fruit from his orchard? Our preliminary conclusion is starkly contrarian: because total output cannot be meaningfully defined it (and hence productivity) cannot be validly measured.
...continued in Part II
 
Designed & maintained by
Artist Web Design
©1999-2010 All Rights Reserved |
