The Automatic Factory— Fantasy and Fact
Automation and its sister magic word electronics have inspired extremes of panic on the one side and rapture on the other, both flowing, however, from lack of understanding of what is involved. One man’s nightmare is another man’s dream of bliss. On one extreme it spells a world without jobs, all the workers being unemployed; on the other a world of leisure and abundance without anyone having to work.
There is no evidence to support either fantasy. Automation is an extension of the familiar mechanization of production, a series of complex processes operated and controlled by the machine itself using electronic controls which help “to reduce the amount of routine brainwork in factories, just as mechanical handling on the production line displaces routine labour.” (For those interested in definitions the correspondent of the “Times Radio and Television Supplement,” 19 August, 1955, whose words are quoted above, defines an electronic device as one which depends “at some stage upon the passage of current through a vacuum, gas or semiconductor”).
Much that is being written about automation and its likely consequences is confused in the minds of writers and readers alike. Because it has been possible to operate certain factory processes automatically and with very few workers (as is already true of modern oil refineries) it has been rashly assumed that automation is capable of general application and that it will bring great saving of labour and cost. In general the confusion is caused by failure to separate technical problems and solutions from economic and social ones.
It is assumed that because a mechanical and electronic apparatus is technically possible for the engineers it is commercially acceptable to business men and will be adopted by them. While governments may go to almost unlimited expense to apply a technical invention to armaments, business men have to look at cost, which usually means that they look at saving of labour. (There are exceptions to this. As Marx pointed out, if wages fall below the value of labour power, employers, being interested in saving cost not saving labour, have no incentive to go to the expense of introducing a labour saving machine).
The extent to which automation will be commercially practicable and the speed with which it will be introduced are then largely governed, by considerations of cost and profit.
Factory owners who are invited to spend large sums re-equipping plant for automatic methods want to know whether it will cheapen production; it is not sufficient to tell them that technically the job can be done. But so far little attention has been given to providing the kind of information needed. The Manchester Guardian (20/6/55) criticised the recent Margate Conference on Automation for that reason:—
“A more serious deficiency has been that, with one limited exception, there has been no attempt to assess the economic cost and advantage of automatic processes, for the answers to such questions will determine the rate at which automation can be introduced.”
The “experts” generally take the view that the development of automation in Britain will not be extensive, and will not be rapid even where technically it is possible. A contributory cause of delay is that the factory owner, under present boom conditions, has first to overcome the difficulty of getting the equipment and then of getting workers to instal and operate it; the electronics industry is itself hampered by “shortage of technical manpower.” (Times Supplement).
A pointer in the slowness of industrial change in this country is given by the telephone service; the first automatic exchange was opened more than 40 years ago but still a quarter of the exchanges have not been converted to automatic working. This is one of the industries in which the introduction of automatic working was expected by an earlier generation to accomplish a speedy and revolutionary change, but the overall saving of manpower has proved to be quite small. And such is the amount and cost of the equipment that several hundred thousand would-be subscribers still cannot be provided and have to wait in the queue.
The most serious and misleading error in the approach to the problem of automation is that of assuming, without examining the facts to find out, an enormous increase of productivity. To be told that a machine has been produced which eliminates hand labour from the “end processes” in cigarette making, or coal cutting or telephoning or any other operation, does not tell us anything at all about labour-saving and productivity; all it tells us is that labour has been eliminated from the end process, and that labour has been necessarily increased in the preceding processes of making driving and maintaining the machines. To the extent that, for the same total output, the labour saved at one end exceeds the increase of labour at the other, there will be labour saving and increased productivity. Although from time to time a particular industry may make a big jump in productivity when new machines and methods are introduced, the average increase over a period of years for all industries is of very modest amount. In Britain during this century it has averaged between one per cent. and two per cent. a year.
The general picture of British industry since back in the nineteenth century is of a transference of workers from end-process work (miners, landworkers, and hand workers generally) to the engineering and allied trades which have grown enormously and are growing still. In effect many tens of thousands of land workers, mine workers, cigarette makers, etc., are now employed in the engineering industries making machines for those trades.
If it is said that automation will be markedly different from the older mechanization processes and will really increase overall productivity at an unprecedented rate, past experience calls for caution: it should be remembered that similar claims have been made endlessly and falsely throughout the past century, and that no-one has yet produced evidence that will support the claim that automation can rapidly increase productivity in industry as a whole. (An influential committee is reported now to have been set up in U.S.A. to study the effects of automation on costs).
The illusion that industrial invention and mechanisation, have in the past enormously increased productivity is still very widespread. A Labour Party publication, “Towards World Plenty” (July, 1952) informs its readers that the application of scientific knowledge and the revolution in the methods of production in Western Europe, North America and Australasia “has raised productivity a thousandfold.” This may be described as somewhat more than a slight exaggeration. Productivity of British workers in the past 150 years has not risen a thousand-fold or a hundred-fold or even ten-fold. (In passing one wonders why, if the Labour Party believed that productivity of British workers had already gone up by a thousand-fold, the Labour Government was badgering them in 1945-1951 to increase productivity by the trivial addition of 10 per cent. “to save the country”).
We have, too, an ample test of current announcements of new processes. If they really brought about a dramatic saving of labour this would show itself in a fall of prices even against the general monetary factors that are tending to raise price levels. But though we read of six men who “can produce the electricity for a town” and four men who “can do the major job” in the production of a third of Britain’s crude oil (Daily Mirror 29/6/55) and twenty-one gas workers who do the work of 350 (Daily Sketch 27/6/55), we read almost simultaneously that electricity, oil and gas, are going up in price! Doubtless the innovations in these industries have produced some small overall saving of labour so that there is a small increase of productivity, but not enough to offset other factors leading to higher prices. In the coal industry all the mechanisation has failed to counteract the effect of the coal having to be mined from greater depths, so that average coal output per man per year is lower than it was 70 years ago and there is little prospect of it getting back to the former level of output.
And the new fuel source to replace the expected diminution of supplies of coal and oil, atomic energy, only slightly changes the situation since the labour required for its production and application is expected at best to be only moderately below that required for the old fuels, and that not at once but ten years ahead. At the Geneva conference on nuclear power in August
“the papers presented . . . here show that we must not expect the cost of nuclear power to be cheaper in the next decade than power from coal.” (Times 20/8/55).
The Times goes on to quote Sir John Cockcroft, Director of the Atomic Research Establishment at Harwell, as saying that “there is good reason to believe that in the second decade the cost of nuclear power will fall below that of power from coal and oil.”
It would be wise on past form to accept even this estimate with caution, since it may turn out to be no more accurate than the “expert” forecasts of a few years ago of expected increased productivity and cheapening of coal.
But if the problems of automation have been misunderstood by many observers that does not mean that there are no problems for the workers to face. To an economist, remote from the factory and the farm, it is easy to talk of farmworkers becoming engineers, miners going to work in factories, and clerks becoming bus-workers, but the transference may be impossible, or, where possible, very difficult and costly to the worker. Lord Halsbury, speaking at the Margate conference on Automation, dealt with this:
“A more serious problem, he thought, would concern those workpeople who could not be fitted into the future pattern of industry with its demand for high skill. A farm labourer can be trained as an engineer if he is still young, but can this be done after he is 40?
(Manchester Guardian 18/6/55).
His two further remarks, that the farm labourer’s children will become trained as engineers, and that automation will not spread quickly, may or may not bring comfort to the man whose job is threatened.
What other bearing does automation have on unemployment ? As already mentioned there will be the closing down of certain jobs, and unemployment for the workers who held them. If they can find other work that they can do, in the place where they live or where they can get accommodation, the loss of employment may not last long under present conditions, but they may have to take a lower wage because their old skill may not have much use in another job.
But the mass unemployment of a crisis is a different proposition. This is caused by widespread dislocation and failure of markets and when it occurs it engulfs all industries, whatever their degree of mechanization; mechanization itself will not be the direct cause of a crisis though its development can be a contributory factor in helping to create dislocation of markets. Crises occurred in earlier days when machine production was in its infancy, and the problem of their cause is not much altered by automation.
Which brings us to the most important aspect of all as far as the workers are concerned. Automation will not be applied for the benefit of the workers but for the benefit of the Capitalists. They own the old means of production and will own the new ones. If anyone benefits from its introduction it will be the owners not the workers, except to the extent that some of the latter can struggle successfully to make it so. And the owners will as usual find plenty of university professors and technical experts to prove the inevitability of whatever harmful effects automation has on the workers. We may fittingly conclude with an observation made by one of the experts on automation. Professor Norbert Wiener, of Massachusetts:
“The ordinary worker . . . is just a source of low-level judgment. He can and will be replaced by machines.” (Manchester Guardian Review of Industry 1952. p. 72).
It is up to the workers to exercise their capacity themselves and in their own interest by replacing Capitalism by Socialism—they can expect no salvation from the owners of industry, the politicians and parties that administer Capitalism, or from the technicians and scientists who preside over the new technques of production.
H.