This sampling of collected testimony is in no sense definitive, but is offered to illustrate some of the dimensions of the manpower problems we face even in expanding labor markets. The absence of reported manpower bottlenecks in the automated areas of company operations represents a sobering consideration when skill upgrading is so frequently identified with the automation process. Needless to say, the benefits of automation that involve increased leisure and improved living standards can set in motion new activities with their own separate mix of manpower and skill requirements. Such activities have not been the focal point for this study.

This analysis suggests, too, that the complete solution to structural unemployment is not likely to be realized by measures that would upgrade labor skills. The nature of unemployment is sufficiently severe to require bold and expanded pressures on both the supply and the demand side of the market. Labor relocation and skill upgrading may well be a vital necessary step for labor reabsorption, but it is hardly a sufficient step. The bromide offered in the 1920's by Foster & Catchings may have some relevance to the structural employment problem of our decade: "A willing buyer doesn't have to wait long: a willing seller may have to wait forever." Unless skill upgrading programs are general, they may do more to redistribute the incidence than to remedy the cause of labor displacement; they may draw a tighter line around those willing sellers of labor service who have already been waiting some time for a job.

[From Harvard Business Review, July-August, 1958]

DOES AUTOMATION RAISE SKILL REQUIREMENTS? This question is too important for unfounded assumptions and unsupported arguments. Critical analysis provides some surprising

answers

(By James R. Bright)

Does the complexity of automatic machinery require extensive retraining of the labor force?

Will the average unskilled worker become unemployable?

Will maintenance needs and technical difficulties cause a radical change in the composition of the labor force?

On what basis should the productivity gains of automation be distributed?

How will the changes in job content caused by automation affect our present wage policies and job evaluations?

Much of the furor over automation has been concerned with its impact on the work force, and hence on society. The above questions represent the key areas of controversy raised by the central assumption that automation increases skill requirements.

It seems immediately obvious and apparently logical that automatic machinery does result in higher work force skills, but is this really correct? I hold that it is not-and that, despite a number of excep tions here and there, the opposite is a truer picture of reality. It follows that a good many managements stand to make some serious errors in wage policy, employment standards, personnel recruitment.

labor contract negotiations, union relationships, and training program planning because of misunderstanding the impact of automation in this regard. It is even possible that faulty social legislation may be enacted because of lack of critical consideration of the true facts.

One purpose of this article is to offer a hypothesis as to how automaticity affects human contribution to production tasks. This hypothesis helps to explain many labor, skill, and training results of automation that are quite at odds with common claims and casual assumptions.

TRUTH AND CONSEQUENCES

The popular train of thought is so persuasive and simple that almost all of us tend to accept, or at least not to challenge strongly, the upgrading-through-automation conclusions. They are found again and again in the testimony before the congressional hearings on automation, in many union publications, in popular writings and speeches, and in numerous professional statements by managers, engineers, economists, social scientists, and students of management.1 While there are many shades of opinion in this school of thought, one will search far and wide in current automation literature to find any dissenting views. In simplest terms, the reasoning runs like this:

Automation results in machinery of a more automatic nature, directed by highly automatic controls. Both the machinery and the control devices are more complex and sophisticated in their action, and must be carefully integrated to achieve performance requirements. Therefore, the employees manning, servicing, and directing this equipment need a higher degree of understanding. They will require additional training, higher types of skills, and even new levels of education. Thus the job content of individual factory tasks generally will require more skill.

The factory payroll will therefore have to include more skilled and fewer unskilled persons. Furthermore, the effect of automatic machinery will be to eliminate many workers and to put relatively highly skilled operators in their places. Meanwhile, the maintenance force will have to be expanded, at least as a proportion of the work force. Obviously, the factory work force will be upgraded by these effects. A more highly skilled work force is inevitable and essential. What effect on labor?

From this conclusion two general points of view have evolved. Many managers, manufacturers of machinery, engineers, and automation enthusiasts have held that upgrading is a blessing of automation. It will relieve labor of drudgery and of monotonous, repetitive work. The superior levels of education and training required to man the automated plant deserve and will command higher prestige and pay for the worker. Automation should be welcomed because it will upgrade labor into higher caliber, more dignified, satisfying, and valuable social and economic positions.

See Automation and Technological Change, hearings before the Subcommittee on Economic Stabilization to the Joint Committee on the Economic Report, 84th Cong., 1st sess., pp. 106, 273, etc.

But see Arthur M. Turner, "A Research Views Human Adjustment to Automation," Advanced Management, May 1956, p. 21.

The other point of view-so vehemently expressed by many labor leaders, but also supported by some social scientists, popular writers, and even a few politicians originates from the same premises and concludes: "Exactly so." Spokesmen then proceed to a logical and alarming end: not only will the average worker be displaced by the higher productivity of automatic equipment; he will be barred from. the plant because he lacks the education, training, and skill necessary to hold one of the automated jobs. The automated plant thus becomes a technological lockout for the common man. Extensive retraining efforts and legislation to soften the blow to labor are, therefore, an urgent social necessity.

Are such sweeping generalizations on firm ground? On what basic premises do they rest? There are several underlying assumptions that seem to be interwoven:

(1) Automatic machinery demands higher degrees of skill and or training on the part of the worker than the old equipment did.

(2) Automatic machinery requires more maintenance attention and/or higher types of maintenance skills than the old machinery did. (3) More engineers and technicians are required to design, build. install, and operate highly automatic production machinery.

(4) These changes are occurring in significant percentages and over short intervals.

(5) The present work force cannot meet the new demands, at least without elaborate retraining.

(6) Conditions in the company are such that a net displacement of unskilled workers in favor of skilled workers will occur at the time the automation is installed.

Popular claims disputed

During the several years which I spent in field research on managerial problems in so-called automated plants and in exploring automation with industrialists, Government personnel, social scientists. and other researchers, I did not find that the upgrading effect had occurred to anywhere near the extent that is often assumed.3

On the contrary, there was more evidence that automation had reduced the skill requirements of the operating work force, and occasionally of the entire factory force including the maintenance organization.

I found frequent instances in which management's stated belief that automation had required a higher caliber of work force skill was refuted when the facts were explored. Other managers admitted, and on several occasions emphasized, that they had made substantial errors in assigning high wage rates to some automated jobs. These wage rates not only proved to be out of line with the difficulty of the task but were unfair in contrast to the wages of employees working with conventional machinery. As will be cited later, the training time for some key jobs, was reduced to a mere fraction of the former figure.

Here, then, was a series of results which directly opposed common automation claims. They certainly challenged the truth, or at least the universal applicability, of the assumption that the automated factory requires a more highly skilled work force than the conventional one. This is not to deny that there were examples of skill increases required

For the results of this research, see James R. Bright, "Automation and Management" (Boston, Division of Research, Harvard Business School, 1958).

by automation. But it did seem that their importance was exaggerated.

A significant conclusion which developed out of this study is that automaticity (or automation) does not inevitably mean lack of opportunity for the unskilled worker. On the contrary, automated machinery tends to require less operator skill after certain levels of mechanization are achieved. It seems that the average worker will master different jobs more quickly and easily in the use of highly automatic machinery. Many so-called key skilled jobs, currently requiring long experience and training, will be reduced to easily learned, machinetending jobs. To understand how this can be, let us examine the demands automatic machinery truly makes on the skill and training of the worker.

THE "SKILLED WORKER"

Perhaps the best way to start is by asking: What kinds of contributions does the worker make to production tasks? Or, better still, for what contributions is he paid? As a general list of things for which he receives compensation we might include:

1. Physical effort. The expenditure of energy through bodily movement, either to manipulate materials and tools or to control the environment.

2. Mental effort. The use of mental powers to sense and analyze job requirements, and to direct action accordingly. This also implies attention and concentration.

3. Manipulative skill. The employment of a specialized physical dexterity.

4. General skill.-Understanding and ability in a task not too susceptible to learning through mechanical rote or formal analysis. This is similar to competence in an art.

5. Education. The knowledge of and competence in the use of a formally organized body of theory and fact apparently required by

the task.

6. Experience. The ability, comprehension, and judgment that have grown out of practice at the task.

7. Exposure to hazards. The extraordinary conditions in which the operator's own safety is jeopardized to some degree.

8. Undesirable job conditions. The unpleasant environment conditions or work arrangements that require special consideration.

9. Responsibility.-The extent to which the worker controls the safety, quality, or productivity of the activity with respect to persons, equipment, and/or materials.

make judgments that have a significant effect on successful per10. Decisionmaking.-The extent to which the worker must or can

formance.

11. Influence on productivity.-The extent to which the worker inreases productivity above an expected norm by the contribution of

12. Seniority.-The mechanistic measurement of service that presumably reflects greater contribution by the worker through some of

the factors above, and perhaps also a reward for loyalty.

valid: that is, that there are different physical and mental activities While this list might be challenged in detail, its concept seems

which the worker may contribute to a production task. That there is room for disagreement on these activities or my suggested definitions also shows that the phrase, "skilled worker," has a highly subjective connotation. Certainly these words are not used consistently by many of us, or uniformly by industry.

It is immediately apparent that not all of these demands or contributions are equally important in a given task, nor are they of a constant relative importance from one production task to another. In even a simple activity the worker's contribution can vary depending on the equipment used. Consider, for instance, the physical effort of a construction worker using wheelbarrow and shovel versus the skill, mental effort, and experience required to use a bulldozer on the same job. Therefore, to understand how automation affects the work force skills we must consider how each of the demands in any given job is affected by increasing degrees of mechanization and automatic control. How is job content altered by the increasing sophistication of a mechanism's performance?

QUALITIES OF MECHANIZATION

Mechanization is not an equivalent thing in every production system. One production line is "more mechanized" than another. Wherein lies the difference? Part of the explanation is that mechanization has at least three fundamental qualities or dimensions:

Span: The extent to which mechanization spreads across a sequence of production events.

Level: The degree of mechanical accomplishment by which a given production action is performed (thus reflecting, in part, the fact that automatic control leads to increasing sophistication in the response of the machinery to environmental conditions).

Penetration: The extent to which secondary and tertiary production tasks, such as lubrication, adjustment, and repair, are mechanized. Levels of accomplishment

The concept of level, which is the aspect that has the most impact on the operator, is based on the assumption that there are different degrees of mechanical accomplishment in machinery. We can sense this by asking ourselves a question: In what way does machinery supplement man's muscles, his mental processes, his judgment, and his degree of control?

We can examine the characteristics of mechanical performance by analyzing how tools refine and supplement man's abilities. This analysis can be arranged and related in a chart, as in exhibit I, which on careful study begins to clarify the relationship of contributions by the worker and the machine as automaticity increases. A distinct evolution is apparent in these levels:

First, there is the substitution of mechanical power for manual effort, which takes some burden from the worker (after level 2). Then, as increasing degrees of fixed control yield the desired machine action, the worker does less and less guidance of the tool (levels) 5-8).

As the ability to measure is added to the machine, a portion of the control decision information is mechanically obtained for the operator (after level 8).