In response to:
The Bogey of Automation from the August 26, 1965 issue
To the Editors:
Any popular exploration of a topic as complex as cybernation is bound to be skimpy: to fuse fact and speculation. This includes Daniel Bell’s contribution in the issue of August 26. To wit:
1. Bell is properly cautious about our ability to define and measure technological change. Nevertheless, the attention he gives to productivity measures endows them with an undue aura of authority and utility. According to the “father” of current productivity measures, Solomon Fabricant (in “Measurement of Technological Change,” Manpower Administration, U.S. Department of Labor, 1965):
The problem of measurement has not yet been solved…. There are competing and widely differing measurements of technological change…. I’m afraid that people talk about both the past and the future…with more confidence than is warranted by the available knowledge about technological change….
Another problem is created by irregularity in the rate of change…. It is going to make a difference which 5-year period is picked or whether the trend rate measured over a 5-year period of a 10-year period….
…the same average rate of 2 or 3 or 4 per cent per annum (productivity increase) for the country as a whole may mean a serious problem of adjustment when there is great divergence in the rates of growth in productivity in different industries….
Nobody is preparing current statistics on productivity by individual industries covering a substantial number of industries…
2. Inquiries about the Department of Labor report Bell refers to elicited only expressions of mystification. (All agreed that the recent 36-industry study on technological trends known to them was not prepared as a study of the probable effects of automation and does not yield the kind of data Bell describes.) At any rate, no report conscientiously estimating the future could assert categorically that computers could be only “only in…offices” in certain industries, or that “only 14 of the 36 industries could possibly use computers to control production itself” during the coming decade. (Italics mine.) No available models can adequately, much less precisely, predict the rate at which an industry will cybernate as a response to domestic or foreign competition or the degree to which other unanticipated changes in associated technologies and products—not to mention the interplay of economic, humanitarian, public relations, and other factors—may make cybernation more or less possible and pervasive in a given industry. These assertions are speculations, not facts—and incautious speculations at that.
3. No serious student has argued that “fully automated” processes are the likely near future destiny of most of our industrial activities. What’s more, in the absence of countermeasures, unemployment would be a desperately serious problem long before that stage were partially reached. Much more importantly, it has taken ten years to learn how to fit cybernation to industrial and business practices and vice versa, and to invent the complex, subtle, and flexible programs, programming methods, and hardware. Very few users of computer systems now use them as efficiently or as imaginatively as their potential allows. This accumulating know-how inevitably means more organizations will be able to use these techniques increasingly effectively, and as some do, others will have to follow.
4. The averages, time trend data, and so on, used to demonstrate that cybernation has had little or no effect on the labor force—and by implication, presumably, that it will have little effect—by their very nature wash out the present exceptional examples of cybernation’s impact on the workers. As Secretary of Labor Wirtz said recently, expressing reservations about labor statistics, “A man with one foot on a hot stove and another in a deep freezer—I suppose you could say that he is on the average comfortable.” It would be surprising if the impact of cybernation in its present relatively early stages were felt everywhere or in particular places sufficiently impressively to show up in the crude statistical measures of technological change which are all we have.
5. In The Next Generation, I argue a different, if related, viewpoint than that of the Ad Hoc Committee on the Triple Revolution regarding the effects of cybernation on employment. I deliberately used “disrupt,” not “unemploy,” in order to emphasize both unemployment produced by cybernation, and, for many who are employed, radical changes in the substantial content and status of their jobs. For example, at the American Society for Public Administration’s 1963 Symposium of Automation and Government, Harold Seidman, then Acting Assistant Director for Management and Organization in the U. S. Bureau of the Budget, said:
A great proportion of [government] employees will face the prospect of (1) a mad, increasingly competitive scramble for the relatively fewer jobs in the fields in which they may be qualified or (2) shifting or upgrading their occupation and career objectives numerous times over the years in order to qualify for new opportunities.
6. Possibly we could invent a society which one way or another provides full “employment.” Doubtless we could have a full employment economy with a war and its stiff regulation of consumption, production, and manpower allocations. But can our type of society and economic system have full employment without a war? (Indeed, would we have as much employment as we do now without the cold war weapons industry?) To do so, our society would have to transform its values about what it is willing to pay well for and what it is willing to give up in terms of efficiency in the name of humanitarianism. (For it may well be that if massive efforts in the public sector were to be operated efficiently, they would, by their very scope and scale, encourage their cybernation.) That we seem unready yet to make these transformations is evident in Bell’s lament that “It has been extraordinarily difficult to get the government to undertake programs (involving direct government spending in areas such as housing, urban transport, medical care, and education).” Part of what’s going to be involved in getting the needed shift—if we do get it—in values and styles of doing and paying for work will be recognizing the likely impacts of cybernation on conventional work—not in deprecating its very likely contribution to the “difficult political and economic problems that lie ahead.”
Donald N. Michael
Institute for Policy Studies
November 25, 1965