Automation in Perspective (1965)

From the January 1965 issue of the Socialist Standard

Ever since automation, some ten years ago, became a popular topic for newspaper writers in this country an enormous amount has been printed and spoken about its nature and likely consequences — most of it totally uninformed and misleading. Nine out of ten of the busy journalists have no time to find out for themselves all they can do is to ask the so-called experts, and write up the appropriate sensational stories that their editors suppose the readers want. Few of the “experts” are any better placed when it comes to understanding what effects automation can have on the social system.

They may know a lot about the technical side of the computers and other equipment they manufacture or operate but little or nothing about the economic laws of the world in which they are going to be used — the world of capitalism.

It is only necessary to look at the wildly different forecasts that are made to know that something must be seriously lacking in the sources of information. On one side are the prophets of abundance, prosperity and leisure on the automation road; on the other those who tell us that soon most of the working class will be out of work, displaced by automated, worker-less factories.

A couple of samples. A writer in Sunday Citizen (6 Dec. 1964), Mr. Stanley Baron, after he had talked “to the top brains in Britain; made the forecast that before the end of the century,” in every industrial country, certainly in the West, most of the essential work will be performed by about 20 per cent of the people — chiefly the most intelligent. The rest of us will work only as much as we wish – or as much as society requires.” This would, he wrote, call for new attitudes to work and leisure, new cities, new forms of education — but not a word about the need for a new social system in place of capitalism.

On the same day Mr. Arthur Helliwell, writing in the People after a visit to America, quoted trade union officials who foretell that by 1970 — no need to wait until the end of the century — American unemployment will be up to 15 million, with “the American economy grinding to a complete standstill.” It is not difficult to show that there is something wrong with Mr. Helliwell’s facts, his reasoning, or his arithmetic. He tells of experts who “estimate that robots are gobbling up jobs at the rate of 40,000 to 70,000 a week,” and that between 2 and 2½ million men “are being thrown out of work every year.” Mr. Helliwell believes that automation, which has been going ahead fast in America for many years, destroys jobs wholesale and swells unemployment wholesale, and he quoted current unemployment there is “already 5,000,000.” If he had inquired a little further he would have discovered that American unemployment is not a couple of million more than it was a year ago, but actually rather less, and that the peak of unemployment occurred about thirty years ago when it was variously estimated at between eleven and fourteen millions, long before automation had been heard of. Also that the number of workers actually in employment in manufacturing industries is higher than it was a year ago; that is in precisely the field in which automation has had relatively big impact.

In Britain, where automation has so far been adopted much less than in America, the picture as regards unemployment is much the same, unemployment is lower than it was a year or two years ago, and the number of people actually in employment is at a record level. So, with much automation in U.S.A. and little automation in Britain, the trends of employment and unemployment have been much the same.

What then is the kind of impact automation makes in the field of employment? The answer is that it is the same kind as that resulting from all the mechanical and other labour-displacing developments of the past two hundred years. So while there might be something in favour of the arguments of those who want the name automation to be kept separate from other technical developments, there is no need to differentiate when we are considering the effect on the workers’ jobs.

The fundamental cause for the introduction of machinery in the 19th century or for the introduction of automation now is that the firm which is spending money on it hopes to make more profit with it than without it, either because it reduces the cost of production, or, as sometimes happens, because it enables the manufacturer to get his products on the market ahead of his rivals even though the cost is not reduced.

These developments have almost always had the consequence of destroying traditional jobs and putting the workers out of work, often with tragic consequences for the victims, who see their acquired skills rendered useless, and who are forced to take lower-paid jobs or find they can’t get work at all.

But here is it necessary to distinguish between the disappearance of certain kinds of work and the reduction in the total number of jobs available. Mechanisation, automation, and the displacement of whole industries all have the former consequences, but do they have the latter? Those who shudder about the monster automation are sure that it does. They are sure that automation will progressively and speedily reduce the total amount of employment for the working class. One group in U.S.A., “The Labor Committee for Full Employment,” at a Conference in December 1963, heard an address from Dr. Arthur Carsters in which he declared that it will be possible within the next decade, “to produce all the goods we need in the U.S.A. with two per cent of the working population.”

The object of the Conference was “to create more jobs”: if they fail, the result, according to their way of looking at things, will be that by about 1974, 98 of every 100 workers in USA will be unemployed.

Of course it isn’t true. It is based on a common misunderstanding of the relationship of technical change to increased productivity, a misunderstanding which leads those concerned into an absurd exaggeration of productivity and the extent to which it is increased by mechanical and other developments.

The corrective for this misunderstanding it to recognise, as Marx pointed out, that the amount of labour required to produce a given article is not merely the labour of the worker engaged in the concluding manufacturing or assembly processes but also the labour embodied in the materials and in the necessary transport, the fuel and the wear and tear of machinery etc. If we make the error of looking at only one part of the process, for example the work of the telephone operator, we get a wholly distorted picture. The introduction of an automatic system which reduces or eliminates telephone operators, will seem to mark a very great increase of productivity and reduction in employment; but in relation to the whole labour process in the telephone service, including the increased labour cost for the equipment itself, it will represent a very moderate reduction.

It is astonishing that Sir Leon Bagrit, Chairman of Elliott-Automation and the leading spokesman on automation problems, suffers from this kind of blindness to what ought to be an elementary aspect of his subject matter. In the first of his Reith Lectures on The Age of Automation he led up to his theme about the increase of productivity we can expect in future through automation by telling his audience that “in the United States . . . one man on the land produces more than enough food to feed fifteen men in the cities, and in fact is a surplus of food grown even by this small proportion of American labour force.”

The statement is in fact wildly inaccurate and misleading. Food, whether in U.S.A. or Britain, is not “produced” by men on the land: it is produced through a series of operations involving also the labour of the transport workers, the makers of farm implements and fertilisers, the producers of petrol etc. etc. (Sir Leon is equally in error when he supposes that the existence of a “surplus” in relation to the market is evidence of there being an actual surplus in relation to needs.)

If the rest of Sir Leon Bagrit’s estimates and forecasts have no sounder foundation than his opinions on agricultural productivity they don’t invite much confidence.

Growth of productivity has been and is still quite slow. In Britain, measured by the estimates of total annual production, it was rising at about one per cent a year before 1914, one and a half per cent between the wars and about-two per cent since. The plan to raise it to 4 per cent has so far not succeeded and the figure for the current year is expected to be two per cent again.

It remains to link up unemployment and employment with the effects of increased productivity due to mechanisation and automation. If total production of all kinds remained unchanged year by year, then in ten years 20 per cent of the workers would be unemployed and unemployment would increase year by year beyond that level.

But total output under capitalism does not remain unchanged year after year. Over the long term the general experience of all countries is for total production to go on rising, not merely in line with population but beyond it, but equally it is normal for the growth to be interrupted from time to time by periods of “depression” when production falls or stagnates and when unemployment rises. This would happen even without changes in technical methods of production, new machines, automation and so on, though they can aggravate what would happen anyway. It would therefore be possible for capitalism to produce another depression like that before the war, not because of automation but because production under capitalism is to make profit by selling what is produced in the world’s markets. The remedy is to abolish capitalism, but people like Sir Leon Bagrit have hardly begun to understand what capitalism is (though presumably he and his firm have a clear idea of the necessity to persuade British and other firms to buy their products). One of his foolish notions is that the world is divided into “capitalist” and “communist or socialist” countries and that automation may make it possible to put human values first and thus supersede both systems.

Let us enlighten him by reminding him that on both sides of the supposed dividing line between Russian State capitalism and Western capitalism, machinery and automated equipment is bought for precisely the same means — to reduce costs of production and capture markets. Not two worlds but one capitalist world.

Let us also emphasise that automation will not solve the basic problem of the working class. The automated equipment of the future will, like that of the present, become instruments for the exploitation of the working class in the interest of the owning class. Automation will neither destroy capitalism nor change its nature into something beneficial to humanity.

Edgar Hardcastle