Science, Technology, and Government

In all of the problems discussed above, the charge has been that free market activity was deficient in some form of scientific research of development. In the question of automation, the charge is really the reverse: that technological improvement might become so great as to threaten dire consequences, particularly unemployment.

Now the spectre of “technological unemployment” has been with us at least since the early days of the Industrial Revolution, when benighted workers smashed machines which came to create jobs for them and raise their standards of living immeasurably above the subsistence level. Despite all manner of refutation, it recurs continually, the latest manifestation being the fashionable view that the current chronic unemployment during a recovery is caused by “too much” increase in productivity (when it is really caused by excessive union wage rates). It is about time that this absurd notion of technological unemployment be laid to rest once and for all. Who was displaced by the steam shovel? How many millions of ditch diggers are now out of work because of it? Where are the billions of unemployed that are supposed to have been caused by the replacement of the human pack animal by the wagon and the truck? Where are they, if the doctrine of technological unemployment is correct? Where are the millions of unemployed resulting from the Industrial Revolution—when the truth is the other way round, that thousands of beggars had nothing to do until the Industrial Revolution rescued them!

Actually, a technological improvement in an industry has the following result: if the demand for the product is elastic (and approximately half of the products have an elastic demand), then the lower prices, and lower costs, of the product will stimulate increased demand and increased production, expanding employment in the industry. If the demand is inelastic, then the improvement will cause less resources to be devoted to the industry, and lower employment; but since prices have declined, the consumers take the funds that they had formerly spent on this industry and spend them elsewhere, thus generating more employment in the other industries. One of the “other industries” that will be expanded will be the industry of making the new machines or new products. Thus: there is no technological unemployment remaining. Automation will have the same effect as any technological improvement, expanding employment in some industries, contracting them in others—but leaving no residue of technological unemployment.55

Discussing the problem of technological unemployment, the Earl of Halsbury writes that he knows of no instance where technological progress has caused prolonged unemployment, or, indeed, where technological regression has caused unemployment!56

More specifically on automation, it is expected to increase the demand for skilled workers in industry, and decrease demand for the unskilled, who can shift (thus continuing recent pro-automation trends) into the service trades, which cannot be automated. Halsbury estimated that practically no unemployment, even temporarily, need be involved in such shifting, since there is a 2 percent “natural” turnover in industry per annum, due to retirement of old and recruiting of young workers, and that the redeployment of labor caused by automation will not be nearly as heavy at this rate. The retirement-recruitment process will therefore be a good buffer against even temporary unemployment. Argyle adds that there is even greater room for mobility, for in addition to this process, about 10 percent of workers leave per annum for other reasons and that these too will buffer against forced unemployment.57

Many of the semi-skilled, and even the unskilled, workers will be upgraded from routine, assembly-line type jobs into better paying, more skilled and varied work. It is largely the routine work that will be eliminated. In many instances, automation will not even decrease the workers in the specific jobs affected. Thus, Halsbury estimates that computerized accountancy, which will permit cheaper and more economic calculation of payrolls, and faster inventory and stock control, will also open up and partially solve a range of new problems, which firms couldn’t even have thought of tackling before: such as “production scheduling.” As a result, he predicts that as many accountants will need to be employed a generation hence as now, except that they will need more skills than they require now.

Automation will be largely applicable, and certainly only economically applicable, in the mass production industries, such as manufacturing, electrical goods, office machinery. It will be feasible for small-scale firms (the new “numerical control”) as well as large in these areas. There will still be plenty of room, however, for homemade goods, crafts, services of persons, etc. And Woollard warns against wild overestimation of what automation in manufacturing will amount to:

if by the term “automatic factory” one is tempted to think of a plant in which the materials are loaded at the beginning of the week, then everyone goes home to play golf expecting to come in on Saturday morning to find the work loading itself on trucks for dispatch, the automatic factory is just a pipe-dream. I doubt very much whether we shall ever see anything of the sort.58

In addition, such industries as transportation and retailing do not seem to be adapted to automation. And Spencer estimates that office automation, while requiring considerable retraining and upgrading of office staff, will not lead to any overall reduction in clerical labor. Office needs for labor have been steadily increasing, due to increased complexity of industry, and the effect of computers will be to stop or slow down this growth, rather than actually unemploy any large number of clerical staff; it will reduce considerably the drudgery of present clerical work.59

Rational optimism about the employment effects of automation has been well expressed by H.R. Nicholas, one of Britain’s most prominent trade union leaders. Nicholas points out that automation creates employment, that our present-day technology has been a boon, rather than a handicap, to employment. Nicholas points out that the numbers employed in our presently most automated industries, such as petroleum, have expanded rather than contracted, because of the prosperity of the industry. There has been more work for tankers, railroads, trucks, etc., to move oil, for shipyards to build these tankers, for managerial, sales, maintenance help in the industry: none of whom will be displaced by automation.60

One point about automation that should not be overlooked: “it will greatly improve the safety of industrial work, many of the unsafe jobs (such as handling atomic, fissionable materials) being automatically accomplished.”61

Let us, therefore, put aside the old Luddite (machine-wrecking) bogeyman of technological unemployment, and hail modern developments of automation for what it is and will be: a superb method of greatly increasing the standards of living and the leisure hours, of all of us. We can therefore, hail the Douglas Subcommittee when it reported as follows:

One highly gratifying thing which appeared throughout the hearings was the evidence that all elements in the American economy accept and welcome progress, change, and increasing productivity. This flexibility of mind and temperament has been a conspicuous characteristic of American industry for generations in well-known contrast to that of many other countries. Not a single witness raised a voice in opposition to automation and advancing technology. This was true of the representatives of organized labor as well as of those who spoke from the side of management. ... Labor, of course, recognizes that automatic machinery lessons the drudgery for the individual worker and contributes greatly to the welfare and standard of living of all.62

• 55This, indeed, is the effect of any change in the economy, whether of consumer wants, of natural resources, of climate, or technology: employment in some firms and industries will be expanded, and in others will be contracted.
• 56The Earl of Halsbury, “Introduction,” in E.M. Hugh-Jones, ed., The Push-Button World (University of Oklahoma Press, 1956).
• 57Michael Argyle, “Social Aspects of Automation,” in ibid., p. 113.
• 58Frank G. Woollard, “Automation in Engineering Production,” in ibid., p. 38.
• 59W.R. Spencer, “Administrative Applications of Automation,” in ibid., p. 107.
• 60H.R. Nicholas, “The Trade Union Approach to Automation,” in ibid.
• 61See Automation and Technological Change, Report of the Subcommittee on Economic Stabilization to the Joint Committee of the Economic Report (Washington, D.C.: 1955), p. 6.
• 62Ibid., pp. 4–5.