Automation and Joblessness: Is Retraining the Answer?

Assistant to the director of the Salk Institute for Biological Studies, WILLIAM GLAZIISB has devoted his time to the trade-union movement since graduation from Harvard. For many years associated with the International Longshoremen’s & Warehousemen’s Union, he is now writing a book on the impact of automation on labor.

THERE are few important economic issues in our nation which are not in some way attributed to automation. Whether it be joblessness, apprehension over the increased power and market influence of dominant corporations, major collective-bargaining conflicts, or the indifferent prospects for capital investment and economic growth, automation seems to be lurking somewhere in the background.

Ironically enough, the same lack of balance which ascribes too much blame to automation is equally responsible for expecting too many blessings from this advanced technology. Although a great deal of today’s unemployment is correctly attributed to the displacement of men by machines, these same machines are also confidently expected to create higher-skill job opportunities, which the unemployed can presumably fill after retraining.

The present Administration, beset with a nationwide joblessness which refused to melt away in the 1961 business revival, and armed with soothing projections demonstrating the higherskill occupational requirements of an increasingly automated economy, has made a variety of proposals to facilitate the retraining of technologically displaced labor.

There is no denying the appeal of these proposals to retrain the unemployed. But, as so often is the case with complex economic problems, the simple solution is frequently either naive or dangerously disingenuous.

Since World War II, unemployment in the troughs of each recession has averaged about 7.5 percent of the labor force, but when the recovery peaks were reached, it was discovered that not all of those who lost their jobs during the recession had been re-employed. In the peak year 1953, 2.9 percent of the labor force was jobless; in 1957 the proportion had climbed to 4.3 percent; and in 1960, the most recent peak year, 5.6 percent of the labor force was unable to find work.

Traditionally we have accepted the premise that the primary cause of joblessness is a temporary layoff during a business recession. With the help of unemployment compensation, the worker is tided over until economic conditions improve and he is called back to work. But how can we explain an economic recovery which does not reabsorb the unemployed? And what can be done about it?

One of the more facile explanations of this disturbing phenomenon is that the hard core of the unemployed is composed of individuals who cannot find employment primarily because they are not fitted for available job openings. They have either been technologically displaced or have reached working age without requisite industrial skills and training.

There is widespread agreement among congressmen of both parties that technological change is primarily responsible for displacing unskilled and semiskilled workers and for making old skills obsolete; but simultaneously, so the theory goes, this same technological change is responsible for opening up new jobs, of a higher skill content. The conclusion is neat and plausible: train and retrain the jobless, and fit the unemployed worker to the unfilled job.

Not only are most senators and congressmen sold on retraining as the answer to current hardcore unemployment and the technological displacement of labor, but the general public seems to share the same illusion. According to the Gallup Poll, of all the proposals specified by the President in his second State of the Union message, the proposal to train the unemployed was cited by 67 percent of those replying as one for which they were willing to make sacrifices. This was more than twice the support given any other recommendation.

The seriousness of the problem of persistent, hard-core unemployment cannot be exaggerated. When the National Planning Association studied the problem in early 1961, it found that the chronically unemployed had grown from fewer than 500,000 persons in late 1953 to about 1.5 million at the end of 1956 and to about 2 million at the beginning of 1960. The number has continued to grow since. Technological change, decline in some industries and growth in others, shifts in the geographical location of plants, and changes in consumer demand have caused these many millions of workers to be unemployed and have kept them that way. They are the victims of growth and progress in the American economy.

The consequential shifts in the structure of industry have left behind a growing pool of unskilled and semiskilled workers handicapped by the limits of a grade school education, equipped with years of routine production-work experience, and burdened with families to support. Many are members of minority groups. In addition, there are the young people under twenty-two years of age, who have the highest unemployment rate of any group in the nation today; half of them have still to get their first jobs. They are largely unskilled and untrained for employment.

Meanwhile, the paradox, as the advocates of training and retraining see it, is that all over the nation jobs go unfilled. In Detroit, with 12 to 14 percent of the labor force unemployed, employers are unable to fill openings for electronics technicians, computer operators, and bookkeeper system specialists. As the want ads demonstrate, there are nationwide shortages of draftsmen, technicians, and electronic, hydraulic, and pneumatic repairmen. There seems to be no lack of people on the one hand or unfilled jobs on the other; what appears to be lacking is people with sufficient training and the right skills. The jobless worker, in the wrong place with the wrong skills and aptitudes, has become the fall guy.

IN MOST of the discussions about retraining, it is conveniently forgotten that the purpose of laborsaving technical improvements is to enable the same amount of product to be produced by fewer workers. Other things remaining equal, after a technological innovation has been applied in a factory or an office, some members of the work force will have lost their jobs. Moreover, technological progress as such is not going to reabsorb these workers. Only a sufficient increase in production, either in the plant or elsewhere, will provide alternative employment opportunities for the individuals whose labor and skills have been incorporated into a new piece of machinery. The nub of the problem is how to achieve this sufficiency of production.

Automation not only diminishes the number of workers employed in the factory, it also diminishes skill requirements. If this were not so, if the displacement of unskilled and semiskilled workers by machines necessitated the increased employment of higher-skilled, higher-paid workers, then obviously the whole evolution of production to a higher technological level would be economic foolishness.

Equally naive is the expectation that as employment drops in the firms replacing men with machines, the displaced workers or their counterparts will find employment in building machines. If as many people were now employed in manufacturing the machines as had formerly been used in making the final product, there would be no point in substituting machines for people.

The immediate gains from rising productivity that result from the substitution of machines for men largely take the form of cost savings realized from the displacement of labor; this purpose would be frustrated if the labor savings at one point were contravened by increased labor costs elsewhere in the production cycle.

Of course, cases can be cited demonstrating that advanced technological innovations have come into a factory or office with few, if any, layoffs. For example, each of the dozen or so automation situations studied by the Bureau of Labor Statistics of the U.S. Department of Labor in 1960 and 1961 lends itself to the comforting conclusion that with proper preparation and planning, discharges can be virtually eliminated. But it would be unwise to generalize from these few instances in which technologically displaced workers were reabsorbed by their employers. The uniqueness of these particular experiences is demonstrated by the mounting number of technologically unemployed this same government agency reports each month.

Moreover, we cannot forget that the displacement of men by machines need not occur at the same time or even at the same place that the machine makes its appearance. When an auto manufacturer automates without layoffs, and in the process is able to manufacture a component formerly purchased from a supplier, the discharged employees of the supplier are as much the victims of automation as if they had been on the payroll of the automating firm. Or when workers who would have been hired by an automated firm no longer have job opportunities, they too are the victims of automation, of “silent firing.”

In addition, the increase in the skill content of the work force in an automated factory is as elusive to pin down as the increase in new job opportunities.

As the studies made by Professor James Bright of the Harvard Business School a few years ago show, the effect of automation is more than likely to reduce — or, at least, not to increase — the demand for skills and abilities on the part of the direct production workers. The same tendency holds for the indirect production workers, the men who service and maintain the machines, in contrast to those who operate or supervise them. The complaints about shortages of skilled engineers, technicians, and craftsmen are more likely to be heard from the builders of automatic machinery than from the users, who are profitably replacing human labor and human skills with automatic machines and control devices.

Professor Bright could find “little justification for the popular belief that present labor is employable in automated plants only with extensive retraining, or that there is a major shortage of new skills in the automated factories.” His observations are persuasive in the light of what we know about the diminution of labor and the transfer of skills from men to machines in the evolution of automation.

A social objective, training or retraining employed and unemployed persons is much to be desired. It would improve the employability of workers, open up more attractive and higher-paid job opportunities, and raise the productive level of the entire nation. The debatable issue is the appropriateness of retraining as a remedy for the current chronic unemployment. In the enthusiastic endorsement of the retraining proposals embodied in the Manpower Development and Training Act of 1961, a rather important consideration became obscured: there must be job opportunities in prospect for the trainees, or all such programs become exercises in futility. Retraining a skilled, unemployed West Virginia coal miner as an auto mechanic does not accomplish much if he cannot find a job as an auto mechanic.

Retraining and relocating displaced and unemployed labor, useful and necessary when economic growth and progress are giving rise to many new jobs in new places, become of dubious value when the economy is producing job seekers faster than job opportunities.

In late 1960, after the Oklahoma City plant of Armour and Company had shut down, the tripartite Armour Automation Committee offered to help finance retraining for any of the 431 former production workers who showed promise of benefiting from some form of vocational training. Of the 170 who responded and were tested, 60 were found capable of retraining; the balance were told that the best chance for re-employment would be in manual labor. About one year later, according to the AFL-CIO, 13 had completed the retraining courses and 7 had found employment in their new skills. This experience convinced the union members of the committee that to retrain under these circumstances was simply to raise the educational level of the unemployed.

In 1959, unemployed workers in the state of California who had exhausted their unemployment benefits and were found eligible for extended benefits for an additional thirteen weeks were offered retraining to improve their chances of finding jobs. Some 50,000 unemployed were eligible to take these courses. In the entire state 38 applied for retraining; 26 were approved and took courses.

It is not surprising that unemployed workers are so markedly unenthusiastic about retraining when they have so few reasonable expectations for reemployment, despite the want ads.

The Administration proposal for retraining unemployed workers in vocational schools will cost the federal government approximately SB3 million the first year and reach $238 million the fourth year. This compares with total expenditures on vocational school training in the United States of $238.7 million in 1960, which represented $111 million on the local level, $82.4 on the state level, and $45.3 on the federal level.

The contention of the vocational school administrators, who strongly endorsed the new legislation, that industrial skills are best learned in a vocational school is belied by the fact that three out of every five workers with special vocational skills have received their training on the job and through job progression. They worked up, step by step, to jobs of higher skills; training in most cases has been informal, and instruction and guidance have generally come from a foreman, supervisor, or more highly skilled fellow employee.

TRAINING in new skills is much more practical and is likely to be more successful for both trainee and employer when the program is carried on within the plant and geared to specific job needs. An employer training or retraining his own employees knows whom he is retraining; he trains with specific job openings in view, and the program, no matter how informal, is tailored to the special needs of the plant or department.

On-the-job training is, of course, limited to workers already employed who have another job prospect at the same place. Unemployed persons, unless specifically hired for apprenticeship or training purposes, cannot be benefited. To recognize this is nevertheless no justification for arguing that vocational school training can be substituted with similar beneficial results for trainee and potential employer.

There is a good deal of confusion and exaggeration about just how much skill is included in a skilled or a semiskilled job, and just how much training is required before a worker acquires employable and useful skills. A recent New York state study covering manpower needs and technological change in some of the most diversified industrial, commercial, and business areas in the nation points out that the skills required for the bulk of industrial jobs are relatively simple ones. Most can be taught on a full-time basis in a few days, a few weeks, or a few months. Approximately two thirds of all jobs in New York state at the present time fall into this category. The specific skills required by these jobs are typically taught by employers, on the job. If the jobs are there, the job seekers will learn the skills quickly enough once they are employed.

On the other hand, only about one third of all jobs are accounted for by the skilled cralts, the professions, and technical and managerial occupations. These occupations require complex skills and a wide range of knowledge which can only be gained in a matter of years, not weeks or months. They are far beyond the competence of most of the hard-core unemployed.

The lack of enthusiasm which unemployed workers display toward retraining when there is no assurance of employment is matched by management under these same circumstances. Management has never had any hesitancy about conducting on-the-job training programs when unfilled jobs were holding up production plans. During and immediately after World War II, for example, when American industry was desperately short of labor, many firms enlarged the scope of on-the-job training, not only to upgrade their employees but to equip newly hired unskilled and semiskilled workers for the tasks at hand. However, when business and employment are not expanding, when new employees are not being added, any businessmen would be imprudent not to curtail such programs.

Many large firms, according to reports, have found it advantageous in recent months to expand on-the-job training programs. And many unions, like the Auto Workers, Steelworkers, and Machinists, with large memberships in industries undergoing rapid technological change and labor displacement have made retraining a major collective-bargaining demand. In most cases the cost to the company is minimal, little more than the wages and salaries of instructors and trainees. And as some personnel managers have shrewdly recognized, employer-sponsored retraining programs minimize the usual resistance to change which exists in any work force facing the introduction of labor-displacing machines.

It would be unfair to imply that the Administration has altogether ignored the fact that human adjustments to technological improvements are relatively painless only in growing industries or in a rapidly growing economy with increasing job opportunities. But in the face of an economic recovery which has not reabsorbed the unemployed, the effectiveness of reducing joblessness by retraining has been repeatedly overstated. For, would not re-employment be achieved much more directly by policies which stimulate business activity, and thereby job opportunities? Under buoyant business conditions, retraining is not a critical matter. But, unfortunately, retraining is usually pressed by local and federal authorities when unemployment is high and job openings scarce.

Assuming that the present excessive unemployment has come about because output over the past decade has not been sufficiently high to furnish employment to the young people seeking jobs and to the technologically displaced, it is no problem to determine arithmetically how much of an increase in the gross national product would be necessary to meet these job goals. But there is a dilemma here.

Economic growth will result in more, not less, technological progress and in more, not less, labor displacement. We have the pattern of the booms of the last ten years to remind us of this complication. The recent spurts in economic activity did not eliminate chronic unemployment or unused plant capacity.

When labor productivity has already reached a high level as a result of technological improvement, as is the case in the United States, then any given expansion in economic activity and production will induce a slower increase in job opportunities and total employment than would have been the case before the technological change. If we make increased output, and thereby more jobs, a first priority in the next decade, then we shall begin to find ourselves in the same predicament as the Red Queen — each successive increase in output will create fewer job opportunities. The more advanced we become in technological accomplishment, the faster we have to run to remain in the same place. If we were lower on the ladder of technological progress, it would be easier to maintain full employment.

Economic growth is a complex phenomenon in which the many factors that go into production are continually shifted about and recombined in new and more productive relationships. If it becomes more profitable to employ less human labor at some point, it is certainly fatuous to expect that training and retraining will automatically bring about labor’s re-employment. Realistically, we cannot always expect to maximize output and employment by the same policies. Under certain circumstances these become mutually exclusive goals, whether in an underdeveloped country like India or a highly developed one like the United States.

If the national goal is to minimize technological displacement and unemployment without going back to horse-and-buggy production methods, then a variety of possible policies would appear to be appropriate to achieve optimum employment. For one, despite the objections of the Administration and most business leaders, we must move to a shorter work week or work year, combined, if desired, with multiple-shift operations; this would keep more people employed and would prevent expensive equipment from standing idle. For another, young people should be required to stay in school longer. This requirement, combined with an earlier retirement age, would cut persons off from both ends of the labor force, thereby reducing the number of job seekers.

These policies would bring about a distribution of an increasing part of the fruits of advanced technology, through a higher standard of living made available in the form of less hard work, more leisure, earlier retirements, more education, and similar benefits. These benefits would not necessarily be confined exclusively to employed wage earners.

To suggest that technological change be regulated. directed, and harnessed is to call for a degree of economic planning few Americans are prepared to accept today. Yet how else can we ensure that our rising population of redundant labor will be able to share in an economy where the number of jobs is not growing fast enough?