Automation and the State
In 1944, MARK I—an electromechanical digital computer with over 760,000 parts and 500 miles of wire—was completed. It was a rather slow machine: addition or subtraction took one-third of a second and computing a logarithm to twenty decimal places took an unconscionable one-and-one-half seconds (as compared with today's speed of several millionths of a second). Pressured by the needs of ballistic science, MARK I soon begat ENIAC, the world's first electronic computer, ENIAC made MARK I as archaic as a water wheel. Yet there were drawbacks even in ENIAC, for its memory capacity was small, and it was necessary to rewire afresh for each new problem. And so, with the help of the work of John von Neumann, ENIAC begat EDVAC, a computer with a larger memory and employing the binary number system EDVAC in turn begat EDSAC, which was able to place both instructions and data into the memory unit. Then came RAYDAC, a second cousin to UNIVAC I; SEAC, BIZMAC, FLAC, MIDAC, and DYSEAC quickly followed, all utilizing the basic EDVAC storage idea, MARK III was born in 1950, while OARAC, a close relation, arrived three years later. Some of the immediate descendants of these machines had no names, only numbers, but after ORDVAC came AVIDAC and after ILLIAC came MANIAC I. Finally, a third generation of computers was born surpassing the older vacuum-tube and solid-state systems with magnetic thin film memories and microminiaturization.
By 1963 the number of computers in use in the United States had reached 20,000, and another 5,000 will be added by 1965. The Bureau of Labor Statistics estimates that this technological advance resulted in the loss of over one million jobs in manufacturing industries alone between 1953 and 1959, with prospects for the future looking even gloomier. There are, however, those who dispute the idea that automation has been destroying jobs, and in quarreling with the Bureau of Labor Statistics' estimate, they point to the 1964 Manpower Report which shows a 4.3 million growth in “nonfarm employment” in the years 1957 to 1963. But hidden in the report was a table on the sources of that growth: direct employment by federal, state, and local governments accounted for 46 per cent; government “procurement” programs for 19 per cent; nonprofit institutions for 16 per cent; part-time jobs generated by private demand for 14 per cent; and full-time jobs created by industry's own efforts for only 5 per cent. Lost in the industrial and technological shuffle each year, then, were 200,000 jobs; and this figure does not include the file clerks and accountants whose positions evaporate every time a computer enters an office.
There is little consolation for the factory worker in the expansion of government employment (which has resulted primarily from the assumption of more tasks), since he does not possess the requisite transferable skills. In any event, according to a recent joint report by the Departments of Labor and Commerce, further growth in the number of federal jobs seems doubtful. To make matters worse, employment is also expected to drop because of technology in eighteen major industries which have up to now enjoyed high volume—automobiles, iron and steel, electric power, etc. In fourteen other industries, moreover—including trade, banking, insurance, transport, electronics, and synthetics—only increased demand can, in the opinion of the report, overcome the effects of spreading labor-saving devices.
The fundamental economic relationships involved are very simple: as productivity (i.e., the power to produce) increases, there must be a concomitant increase in output or jobs are bound to disappear. Technology, in part a child of the state, has given a fillip to productivity, but production has not been increasing fast enough to provide all the jobs we need for an ever growing population. Here are the facts: from 1909 to 1947, the average annual increase in productivity was 2 per cent; from 1947 to 1960, it was about 3 per cent; and from 1960 to 1963, it rose 3.6 per cent each year. With such an annual rise in productivity, and with 60 million people in the private labor force, 2.1 million new jobs would have to be created every year just to keep unemployment from going up—and another 1.5 million a year to take care of the young people just coming into the work force. Now, while output has indeed been on the increase, the pace of the increase has been too sluggish to compensate for the enhanced productivity stemming from automation, mechanization, cybernation—that is, from the new technology.
The answer, then, would appear to be simple—just increase output. That, however, is more easily said than done, for there is no way of bringing about greater output (at least in the private sector) when there is no market, and there is no market when there are no jobs. One possible solution—which is indeed on the verge of adoption by default—is to allow Disraeli's Two Nations to develop once again and let the poor go hang. Another solution—advanced by those who believe that fiscal policy alone can cure the affliction of unemployment, no matter what its etiology may be—is to create additional output by such means as the recent tax cut, which is expected to generate an additional thirty billion dollars worth of demand. But the real likelihood is that not even an expanded demand of this magnitude could do much for footloose miners, unemployed packinghouse workers, displaced auto workers, laid-off railroad workers, and all the others whose skills have suddenly become unnecessary and unwanted. To be sure, the Council of Economic Advisers argued before the Senate Labor Committee last year that there is a “proven capacity for a free labor market to reconcile discrepancies between particular labor supplies and particular labor demand.” Yet in recent years the Gross National Product has grown as much as 6 per cent per annum, and no new jobs have come into being for those who need them most bitterly—the displaced, the young, the unskilled. The industries in which displacement has been occurring are mature ones—autos, steel, food-processing, clothing—and it seems highly improbable that they will expand their sales sufficiently to replace the jobs that have been dispensed with. Automation has accelerated productivity in these industries just enough for them to keep pace with normal market growth while getting along with fewer workers.
Despite the grudging recognition the Council of Economic Advisers sometimes gives to the structural distortions stemming from technology, its position is that added investment and greater consumer demand will do more than anything else to ease the burden placed on those who must find new sources of income. The Council insists, too (citing the imperatives of national defense), that the encouragement of innovation—this means automation as well as ordinary industrial change—is a direct responsibility of government. Hence, research and development—for which government paid two-thirds of the $16 billion spent in 1962-63 and which mostly goes to aircraft, missiles, electronics, chemicals, and machinery—needs to be continued. But what of those who, like the handloom weavers of yore and the coal miners of today, are unable to enjoy “the fruits of technology”? Collective bargaining in meat-packing, printing, autos, steel, and on the docks has not been overly successful in dealing with automation. In fact, such schemes as profit-sharing, cost-savings sharing, severance pay, interplant transfers, or early retirement are of help only to the worker who is still on the inside. They do nothing for the worker who was displaced last year, or the year before, or the year before that—nor for the young worker just out of school and looking for a job.
From the point of view of the state—which wants technology unlimited—many of these schemes are really not “practicable” in that they “cannot provide complete worker protection without unduly slowing the pace of technical advance,” as the CEA has so bluntly put it. It seems that ways must therefore he found to help the victims of material advancement, not because they are people, but because they are potential Luddites. This curious view was faintly hinted at back in 1955 when the Joint Economic Committee expressed an earlier optimism about automation. Several years later, President Kennedy's Labor-Management Committee's report on automation, echoing the concern he had voiced for young people seeking jobs and for displaced older workers, recognized that the net effect of technology has been increasing unemployment and suggested that something might have to be done. (Henry Ford II and Arthur F. Burns, who had been Eisenhower's chief economist, dissented and would not support the majority, for it seemed to them that any sort of government action might be oppressive to free enterprise.) Then, last July, it was bruited about that the President might appoint a high-level Commission on Automation: the administration seemed disturbed that if the impact of automation were not softened, resistance would arise and the introduction of new machines slowed down. (Nothing, however, has yet come of the Commission idea despite a revival of Congressional talk about it: it is currently rumored that Johnson may set up an Interdepartmental Committee, an elegant way of burying any controversial issue.)
There are some who say that precedent for government action was provided by the Full Employment Act of 1946. But this statute does not even offer a rough blueprint. Failing to define what “full employment” goals ought to be in an economy that undergoes perpetual change, the Act, which directs the Council of Economic Advisers merely to study conditions and write reports remains what it was intended to be—a superb example of legislative rhetoric. If one looks at the debates in Congress preceding its passage, one finds much talk on fiscal and monetary policy and how the Federal Reserve System might stimulate investment by manipulating interest rates and member bank reserves and by buying and selling securities in the open market. Unfortunately, the Federal Reserve System is still a quasi-independent agency, and it has not always pursued policies that would benefit the rest of us. The Full Employment Act, therefore, is of little help in an age of automation.
The pressures on government to do something, however, continued, and they finally did boil over in 1961 into the first major piece of legislation to deal with structural unemployment—the Area Redevelopment Act, which authorized the spending of $375 million to stimulate economic activity in depressed areas. Most of the money was earmarked for long-term loans, and some of it for grants to local communities. Since the measure was not intended to deal specifically with the aftermath of automation, only a modest sum was allocated for teaching the unemployed new skills.
By the end of 1963, over $200 million had been committed, presumably enough to create 20,000 new jobs in areas long afflicted by joblessness. Additional funds were refused last summer, however, and the ARA will probably expire in futility in mid-1965. Indeed, the overly modest nature of the program suggests that Congress intended it merely as a gesture. Even so, there was a basic defect in the conception of the Act. Although the idea behind it was to plant seed capital in depressed areas to attract new ventures, the very restriction of the loans to companies that were poor risks at the bank in the first place made it unlikely that much good could be done. For the fact is that within the context of the present social and economic order, a depressed area needs the kind of lift that can only be generated by successful firms; marginal companies continue to stagger no matter how much financial blood is pumped into them.
Senator Paul Douglas, who designed a bill to aid depressed areas as far back as 1955, thought of helping just a few of the hardest hit localities as demonstration projects: he wanted to show what could be really done for areas devastated by time and technology. However, as the ARA bill went through one session of Congress after another, through committee and subcommittee hearings, through drafts, redrafts, and rewrites, a pork-barrel sentiment developed. In response to pressures from mayors and governors all over the country, the ARA's geographic responsibility was widened and its possible effectiveness thereby diluted. As Dr. Sar A. Levitan argues in Federal Aid to Depressed Areas,1 the program, considering its limited size, ought to have focused on declining areas—communities which were losing population and whose social capital (homes, schools, churches, businesses) was being left to waste away. Instead, by the time all the baker's dozens of Congressmen and state and local officials got through, there were enough growing towns in the program to convert it into a travesty. Of course, unemployment in a town whose population is on the upswing is just as serious as unemployment in a town whose young people are running off to other places. But ARA was not designed to cope with both situations: its objective was supposed to be to encourage new industries—new to the community—to move into depressed areas to provide jobs.
Yet even if ARA had stuck to this objective, it is doubtful whether the program as conceived could have worked properly. ARA assumed that bolstering existing social capital and social overhead would make a sick town with decrepit shells of abandoned factories attractive enough to draw in a new auto assembly plant or automated abattoir. This “trickle down” theory may have seemed plausible on paper but, in the event, it proved difficult to convince Ford or Armour to invest in a town on the downgrade. In short, it was evident that contemporary politics and economics had alike shown themselves inadequate to cope with the needs engendered by the new technology.
One might think that the huge volume of defense contracts could be employed to help depressed areas. But the Department of Defense may not by statute give awards to businesses in the poorer areas if a lower cost can be arranged elsewhere. This statute is based on normal accounting procedures. From a social standpoint, however, unemployment surely ought to be included in any calculus of cost, as it is in West Germany and Britain, where efforts are made to locate plants in areas that need jobs; in Sweden, similarly, investment tax credits are utilized to locate factories where they can be most helpful. Unfortunately, such ideas are much too suggestive of economic planning to be fully acceptable to American policy-makers; even Senator Douglas did not appear overly enthusiastic with the limited planning provisions urged by some proponents of the Area Redevelopment Act.
Among the other things the ARA experience has revealed is the urgent need for a viable manpower policy. It is a need that arises not only from the high unemployment rates which persist despite prosperity and from the monthly influx of thousands of teen-agers into the labor market, but from the dislocations created by automation as well. Of course, the employment situation is complicated; disentangling in statistically precise ways the variety of elements involved is almost impossible. While some unemployment may be attributed to lagging effective demand, computer technology and rapid obsolescence of older methods of making goods have destroyed numerous skills while creating other more esoteric occupations which themselves require fewer bodies to fill. Hence no amount of heat emanating from a warmed-up fiscal policy will melt the icebergs of hardcore unemployment: existing measures such as the tax cut are like trying to light a match to Antarctica. But what of the other measures that have been adopted—vocational education, retraining, and programs to improve labor mobility? How have they been faring?
To begin with labor mobility: there are some 1,900 public employment service offices in the nation, operated jointly, as it were, by the Department of Labor and the states. The effectiveness of these offices varies from state to state, for administrators have been apt to think of the unemployment insurance program in which they are involved as primary and of placement, counseling, and information as secondary and not very urgent responsibilities. It took the chiding of President Kennedy and the continual prodding of the Labor Department to get the local offices to do more placement work last year. But for all the chiding and prodding and for all the appropriations—$160 million by the federal government alone in 1963—“new hires” through state employment offices still run around a mere 15 per cent of the total number of new jobs acquired.
In 1962, the United States Employment Service installed an “early warning system” to identify in advance groups of workers who might be threatened by technological displacement or mass layoff. This was an excellent idea, except that not all employers were willing to say what they were planning far enough ahead for the system to do much good: after all, to announce a layoff months in advance might have a harmful effect on a corporation's stock prices. When Studebaker closed its South Bend operations in December 1963, it gave exactly one week's notice; with the sudden shutdown, 7,000 workers were stranded, over 7 per cent of the South Bend work force: the shock to the community was a stunning one. Perhaps voluntary “early warning” should be reinforced by some legislative sanction. Certainly at the very least, employers should be persuaded to register all job vacancies. Legislative action, which hitherto has run afoul of the private employment agency lobby, might begin rather modestly by offering firms an adjustment in unemployment insurance rates in return for registering job vacancies at the USES. The President's Executive Order prohibiting discrimination in government contracts could also be broadened to include the same registration requirement. For as Dr. Levitan has argued, people should not have to pay a fee in order to obtain a job in a firm that does business with the government.
While placement by itself cannot create new jobs, it can reduce unemployment by directing those with proper skills to such openings as do exist. The key term, of course, is “proper skills” and here an effective retraining program is essential. By now, there seems little doubt that large numbers of the unemployed lack the skills that would qualify them to fill whatever jobs are available. In Washington, for example, there were recently several hundred empty jobs in the printing trades with no one around to do the work—and no program in operation to train unemployed individuals for these posts. As for the young, two-thirds of the new entrants into the labor force have no specific training of any kind: even possession of a high-school diploma does not qualify them for work. What is being done to provide all these people with the skills they need if they are ever to obtain jobs?
There has, of course, been a federal program in existence for vocational education since 1917, when the Smith-Hughes Act provided for sharing costs with the states. But in the ensuing forty-six years, federal support of these traditional efforts has only averaged about $1.2 million a year, or 7 per cent of state and local outlays. An additional deficiency is that some two-thirds of the approximately four million youngsters receiving vocational education today are being taught home economics and farming, which is hardly likely to prepare them for the age of automation. Probably the situation will be altered somewhat by the Morse-Perkins Act of last year which more than doubled federal allotments to the states and authorized a jump in federal contributions to an annual $226 million by 1967. Occupations other than cookery and farming have also finally been allowed into the program, and part of the new federal money may be used to build vocational schools, to sponsor research, and to engage in such experiments as work-study programs.
Some training was provided for unemployed persons under the ARA in 1961, but it was not until a year later that a more hopeful start was made. The Manpower Development and Training Act, passed in 1962 after much legislative travail, provides allowances for giving unemployed heads of families up to fifty-two weeks of training. At first, lesser allowances were extended to youths between the ages of nineteen and twenty-two, and the proportion of expenditures for such newcomers to the labor market was limited to 5 per cent of the total to be spent on training. The demand for youth training, however, was so great that the law had to be liberalized in 1963. This is all to the good, and would be even better if the training courses were only blessed with more imagination. Stenography, typing, automobile mechanics, nursing, and welding account for half of the program. An effort has also been made to reach farm workers in New Jersey, Arkansas, Oklahoma, and Texas with a course in farm equipment mechanics. (It is good to note that twenty of the twenty-five New Jersey trainees successfully secured new jobs.) In addition, ambitious courses were attempted to train Indians in New Mexico and Arizona as electronic solderers, but when the graduates went to work in the Los Angeles area for an average of $1.76 an hour, well below union rates, organized labor objected. (The Bureau of Indian Affairs would nevertheless like to see the course repeated, since $1.76 an hour is still a good deal more than can be earned on the reservation.) Other courses—the sort of training projects that are said to be necessary to reach the so-called hard-core unemployed—have upgraded janitorial work into a custodial occupation (the latter requires more skill than swabbing a floor with a mop, for maintenance of a modern building requires a knowledge of appliances, detergents, insecticides, and waxes). Finally, there is a modest on-the-job training program and several special demonstration projects aimed at “functional” illiterates and school dropouts.
Yet training programs are not a solution to unemployment. As one economist has remarked, they are rather a salvaging operation to help a few individuals acquire marketable skills. The fact remains that most of the unemployed could not fill the jobs for which there may be openings even if they got a whole decade's training. Most of those in charge of the training programs acknowledge this, and motivated by an understandable bureaucratic desire for success, they have had to impose high admission standards, select the courses carefully, and concentrate on younger people. Consequently, older persons and women have by and large been overlooked (while 10 per cent of all MDTA programs involve persons over the age of forty-five, more than a quarter of the unemployed fall into this age bracket).
How successful has the training effort been thus far? Labor Department officials say that 88 per cent of those who completed their courses got jobs. This sounds impressive until one translates it back into the actual numbers: in 1963 only 27,500 unemployed persons (out of a total of 65,000 who originally enrolled) received training, and of these only 17,000 were able to put it to use successfully. Obviously, much more needs to be done if training is to have any sort of impact. MDTA officials may believe that they have an unqualified success on their hands, but 27,500 graduates a year will not meet Congress's hope for 400,000 retrained workers within three years. At best, it is a useful start; at worst, a piece of misleading showmanship.
Our economy is changing in ways that suggest a far lesser need for workers than has been the case in the past. The old pyramid, in which a large base of employed supported a narrower apex of non-workers, is revolving about its center, so that soon, perhaps in a decade or two, a relative few will provide the material requirements of the many. To prepare for that contingency—which would indeed make automation meaningful—we should start now to reduce the work force. Of course, retraining is necessary to make square pegs round enough to fit into their proper holes. But in addition, a wage system that functioned well enough and at high enough levels to obviate the need for moonlighting and multiple-jobholders would contribute enormously toward alleviating the strains of automation. A variegated works program, geared to rebuilding the neglected “social overhead” in the public realm—roads, parking space, school buildings, urban transport, housing, and parks—would not only give jobs to several million untutored, but also, over the long haul, provide them with minimal work skills. And for those unlikely ever to work again, of which there must be at least a million or more (aside from the aged), why not simply give them enough money to carry on with some measure of decency?
Interestingly enough, other nations have initiated many of these measures precisely as a response to technology. The Swedish Labor Market Board attempts to locate redundant workers and grants travel subsidies and money allowances to help them move to localities where they may be in demand. Financing for new construction is provided in towns where pockets of unemployment develop. By such devices, unemployment in Sweden has been kept down to 1.5 per cent of the labor force—an extraordinary achievement. Last January the Canadian government introduced legislation to establish similar labor mobility schemes. The Dominion authorities proposed to pay half the cost that any province or industry would incur in moving idle workers and their families to places where jobs might be available. Further, unions and management are encouraged to investigate adjustment plans well in advance of changes stemming from automation.
But the centuries-old habits of American politics intrude on our dreams. An archaic ethic which insists that he who does not work shall not eat; a Congressional structure virtually unchanged since colonial days; a federalism that fails to come to grips with the complexities of regional economies; an apathetic response from the affluent; and the American posture of make-do and sudden crash programs—all these raise almost insuperable barriers to the formulation of a reasonable program to deal with automation and its consequences.
Meanwhile, the technology of our age—fostered by the state, which at the same time has responded with such pathetic inadequacy to the problems it has created—continues to spread with undiminished force, altering traditional work relationships and dispossessing hundreds of thousands of people with every passing year.
* Johns Hopkins University Press, 227 pp., $6.95.