The Commission that has issued this Report was established in August 1964 by an Act of Congress. The United States economy was then already booming. The percentage of total labor force unemployed, while gradually declining, was however still a high 5 percent as against the 4 per cent declared by the Council of Economic Advisers to be normal. Automation was the topic of the day.

The Commission was appointed to exorcise the specter of technological unemployment. Among the fourteen members of the Automation Commission, as it was often called, only two—Howard R. Bowen, President of the University of Iowa, and Robert M. Solow of the Massachusetts Institute of Technology—were economists. Daniel Bell, a sociologist and a social philosopher, well-known to the readers of this Review, and Benjamin Aaron, a specialist on Labor Law at the University of California in Berkeley, completed the small professional braintrust of the group. The other ten places, as is fitting on such occasions, were evenly apportioned among representatives of Management, of Labor, and of the so-called Public Interest. With Walter P. Reuther, Thomas J. Watson, Edwin L. Land, and Anna Rosenberg—to mention just a few of the other prominent persons on it—it is indeed a most impressive panel. And as if this were not enough, in the background looms an Advisory Committee representing the sponsors of the Great Society itself: the Secretaries of Defense, Labor, Agriculture, of Health, Education and Welfare, the Chairmen of the Council of Economic Advisers and of the Atomic Energy Commission, as well as the Director of the Arms Control Agency, and the Administrator of National Aeronautics and Space Administration. A staff of twenty-four researchers directed by Garth L. Mangum did the technical background work. In addition a number of special studies were prepared by outside experts “on most of the major parts of that report.” They make up the 1,787 pages of the six appendices.

In the words of its Chairman, the Commission attempted—in this thin, beautifully printed volume—to answer questions it received from Congress and to offer recommendations as charged by Public Law 88-444. The main question raised in the report concerns the current and prospective role of technological change and its effects on production and employment. The report’s most important recommendations are on the legislative steps designed to share the costs of technological change and to help eliminate or at least mitigate its adverse impact on displaced workers.

THE COMMISSION is much more explicit in prescribing cures than in its diagnosis of actual or potential ills. Were the list of proposed cures not so elaborate, the Report would leave the reader with the impression that all the economic ills that threaten our system can be warded off by a regime of well-managed “fiscal policies,” that is, by well-timed doses of deficits or surpluses, as the case may be, in the Federal Budget. Indeed, without stating it directly, the analytical, diagnostic part of the Report implies that unenlightened fiscal policies are in fact the only ill that could befall us. If fiscal policies are inflexible, the result will be a discrepancy between an ever-increasing “full employment output” and an “aggregate demand” which is growing too slowly. The name of the disease—inflexible fiscal policies—thus indicates by definition the appropriate cure.

The familiar argument runs as follows: At any given state of technology, a given labor force fully employed will be capable of turning out a certain “full-employment Gross National Product.” As technology advances and the productivity of labor increases, even an unchanging total labor force becomes capable of producing a larger and larger Gross National Product; if the labor force itself grows at the same time, the “full-employment Gross National Income” naturally grows even faster. If the total volume of private spending falls short of the level required to purchase—at prevailing prices—this full employment output, the gap can and should be filled by increased government spending.

If, on the other hand—as seems to be the case at the present time—the total effective purchasing power becomes so high that the amount of goods and services demanded, at currently prevailing prices, tends to exceed the “full-employment output,” prices tend to rise. The government can in such a situation redress the balance, either by reducing its expenditures, or by increasing taxes (thus, curtailing the expenditures of the private sector) or by an appropriate combination of both.

Technological progress indeed increases the amount of goods and services that can be produced and should be sold to maintain any given level of employment. In fact, instead of translating the steadily rising productivity of labor entirely into a higher and higher level of average per capita consumption, we have absorbed in the past, and certainly will in the future, a substantial part of that advance by gradually shortening the work week, lengthening annual vacations, delaying the entry of the young into the labor force, and requiring the old to retire earlier. In short, increased productivity of labor has permitted the average American not only to consume a larger amount of goods and services, but also to enjoy more leisure.


All these basic aspects of economic progress have been described many times in the last twenty years. As a matter of fact, in 1929 the Committee on Recent Economic Changes of the Conference on Unemployment (headed by the Secretary of Commerce, Herbert Hoover!) issued a two-volume report on Recent Economic Changes that set an example of thoroughness and clarity that has not yet been surpassed. The new Commission recites the lesson once more with particular emphasis on government spending and tax policies as means of keeping the effective demand high enough to maintain full employment. What is surprising is not what is said, but what is not said in the Report on this particular subject.

The sophisticated, as contrasted to the stripped-down popular version of Keynesian theory shows great concern with the crucial relationships between effective demand and fiscal policies, on the one hand, and prices and wage rates on the other. In particular, it demonstrates the importance of fixing either prices—at least some of the more important prices—or money wage rates by custom, agreement, or by edict. If this is not done, the machinery that connects the fiscal or the monetary action of the government with the level of output and employment might inadvertently slip into a higher gear, a lower gear, or even into reverse. If there is no effective limit on wages and prices, the precision of governmental control over the economic system will, at best, be reduced; at worst, it will be entirely ineffective. In 1962, the Council of Economic Advisers introduced the so-called “price and wage guideposts” in order to take care of this potential weakness of the fiscal policies mechanism. These have been invoked with increasing emphasis in recent years: The President, for example, has issued statements to the effect that wages must not rise more than 3.2 per cent in any industry. But how has this worked? The fact is that it has not worked. Increases granted in many recently negotiated wage contracts have risen far beyond that mark.

HOW CAN ONE PREVENT the benefits of a planned government deficit from being dissipated by an unwelcome rise in wages and prices? How can a budgetary surplus, intended to mitigate inflationary pressures, be prevented from producing rising unemployment? These are questions to which, up to now, the New Economics have found no answers. In this Report, for reasons that are difficult or perhaps all too easy to understand, these questions are not even raised.

The rest of the first part of the Report is devoted to a systematic description of the past and the anticipated effects of technological change on the composition and allocation of our growing labor force. We are told that a larger and larger part of the labor force derives its livelihood from all kinds of service industries, a diminishing proportion from manufacturing, and an even more rapidly diminishing proportion from agriculture. Old industries contract while new ones expand; traditional skills become obsolete, while the growing demand for workers capable of handling and repairing new automatic and semi-automatic equipment continuously outstrips the available supply. However, if the overall demand were kept high enough, the only kind of technological unemployment that might show up would be due—so says the Report—to insufficient mobility of labor, that is, to its unwillingness or inability to shift out of the old, declining industries and occupations, into the expanding new ones.

But is this necessarily true? Common sense seems indeed to support such a view. David Ricardo, Karl Marx, John Stuart Mill, the great nineteenth-century economists, however, disagreed with it. Ricardo first accepted the “common sense” position but later, after having given more thought to the problem, came to the conclusion that the introduction of new improved machinery could indeed have a lasting negative effect on the demand for labor:

I thought that the laboring class would, equally with the other classes, participate in the advantage, from the general cheapness of commodities arising from the use of machinery.

These were my opinions, and they continue unaltered, as far as regards the landlord and the capitalist; but I am convinced that the substitution of machinery for human labor is often very injurious to the interests of the class of laborers. [Third edition (1821) of the Principle]

The great theorist was often ridiculed for this statement, but in principle at least, he was nevertheless entirely right.

THE DEBATE ABOUT “technological unemployment” has always been affected by emotions raised by the human—or the inhuman, as the case may be—implications of the two opposite positions. This may be the reason that the present Report, devoted primarily to the analysis of the effects of technological change on employment, meticulously avoids using the term “technological unemployment.”


To bring out the essentially simple logic of the classical argument, let us turn—for a while—away from people and consider horses. According to the Census, the horse population of the United States rose steadily from time immemorial till 1910. In that year, twenty million horses were presumably fully and productively employed in our fields, on our roads and thoroughfares, not to speak of race tracks and riding stables. Judging by the per capita—that is, per horse head—output of oats, the standard of living of that hoofed labor power grew from year to year. Moreover, the Average Annual Hours Worked by Farm Work Animals—and presumably by those employed away from the farms—declined according to the available estimates, from 1,846 hours in 1890 to 1,638 in 1920.

Then something happened. The United States economy continued to grow, but the employment of horses started to decline, first slowly, then very fast, That something was, of course, the invention and gradual perfection of the passenger car, the truck, and finally, the tractor. We, or rather our fathers, witnessed a spectacular case of technological unemployment, accompanied by the usual string of secondary repercussions, painful shifts, and economic readjustments. The carriage makers and harness-making industries declined and finally died out. The sales of equine consumption goods, that is of oats and forage, which ran formerly into hundreds of millions of dollars annually, shrank to a trickle for lack of “effective demand.” True, race tracks prospered and employed more horses than ever before; but this could not compensate for the technological displacement of the less highly skilled horsepower. The automotive industry picked up where the carriage trade left off and gasoline sales soon outstripped the past records of the forage trade.

A NATIONAL (HORSE) COMMISSION on Technology, Automation, and Economic Progress appointed, say in 1907, when self-propelled carriages made their first appearance, would probably have reported on the past steady growth of total employment and of per capita output, that is, the average output of oats and other feeds per horse. Extrapolating this trend into the future, the staff of that Commission could have computed by how much the aggregate demand would have to be increased in the future to secure satisfactory employment opportunities for the steadily increasing equine labor force. Since tractors had not yet made their appearance (although a few automobiles could already be seen on the roads), the 1907 Commission might even have suggested a future increase in the proportion of the total number of horses employed in rapidly expanding agriculture. What would have happened ten years later—that is, what would have happened if horses and not people had had the vote? Tax cuts and increased governmental outlays would probably have been recommended. At the same time, official statistics would have shown a sharp increase in the “average productivity” of (horse) labor: with the farm output going up and the total number of horses employed on the farm going down, the output—per horse—must obviously be soaring. In accordance with the prevailing doctrine, horses would demand more oats per hour of work, which would, however, only accelerate the loss of the remaining jobs. And no amount of deficit spending could have delayed a sad ending.

The moral, or rather the internal logic, of this little fable should by now be plain: What happened to horses conceivably could, but, of course, under no circumstances would be allowed, to happen to human beings who gain their livelihood by selling their labor in a world in which more and more productive tasks can be performed better and more cheaply by machines. Incidentally, horses who happened to own stocks and bonds or titles to valuable real estate would be prospering long after all draft animals had lost their jobs.

The problem of technological unemployment visualized in this way is not a question of the proper balance between potential “full-employment output” and “aggregate demand.” The issue is rather that of distribution of the National Product between those who work as hired labor and those who are the owners of capital, i.e., of productive plants, land, and other natural resources. It is, in other words, the question of the direct and indirect effects that technological innovations might have on the role played by each of these distinct factors in the process of production. What matters ultimately is how these changing roles will affect their relative bargaining strength in the neverending contest for larger shares in the steadily growing National Income. In a rich democracy like ours, the distribution of incomes becomes less and less dependent on the operation of blind economic forces and more and more on the power to influence political decisions and on votes.

None of this comes up in the Report, which obviously represents the final results of a relentless determination to arrive, no matter at what cost, at a general agreement. Only the brief but often lively “comments” signed by individual members of the Commission and reproduced in small type at the end of some of the chapters bear witness to the fact that the consensus was achieved by the Commission not without serious reservations on the part of the different groups of its members.

It seems to be a peculiar characteristic of the leadership that now guides this country toward the Great Society, that it prefers to exhort rather than to explain. Like an old-fashioned family doctor, it hands out prescriptions without fussing much over the diagnosis. But possibly this only proves that even the New Economics, like old-fashioned medicine, is more a practical art than a science. Or again, this might be a bedside manner calculated to keep the doctor from telling all he knows.

In any case, the carefully considered practical recommendations presented in great detail in Parts II and III, and summarized in Part IV of the Report are imaginative and comprehensive. They cover the full range of economic action, from broadly aimed fiscal and monetary policies to specific measures designed to facilitate the shift of labor from obsolete and declining occupations and industries into those that are new and expanding. Among the more general recommendations particularly close to the heart of this reviewer is the timely call for radical modernization of the federal statistical system, which is now quite incapable of serving the needs of policy makers in our computer age.

ON A STILL HIGHER PLANE, the Commission visualizes “some national body of distinguished private citizens representing diverse interests and constituencies and devoted to a continuing discussion of national goals.” However, in small print on the next page, Mr. Sporn of the American Electric Power Company expresses his “strong disagreement” with this suggestion, while Mr. Beirne, Mr. Hayes, and Mr. Reuther—all representing the more progressive wing of organized labor—regret that this recommendation does not call explicitly for “economic planning or programming.” They do not seem to feel that fiscal and monetary policies alone will do the job. So far as the distribution of income, or more generally, of economic welfare is concerned, they are right.

A separate set of recommendations is designed to help those who, in spite of all measures taken to maintain full employment, still find themselves passing or permanent victims of technological change. Most notable, because most radical, among these proposals is the suggestion that Congress should give serious study to a “negative income tax.” In the same way, as families whose incomes exceed the minimum tax-exemption pay a positive income tax, indigent families whose total income falls short of a certain fixed minimum would be entitled to receive payments from the government.

If taken seriously, the establishment of minimum income payments could by itself counteract any effect that technological advance of the labor-saving kind would have on labor’s share in the National Income. Milton Friedman of the University of Chicago, the originator of the negative income tax idea, and erstwhile economic adviser to Senator Goldwater, favored this method as a means of providing supplemental income to poorer groups without government interference in the manner in which they spend it. By contrast, medical insurance, housing and educational subsidies, and similar measures maintain the standard of living of the poor on a selective basis. They not only correct the distribution of income, but also affect and, in a sense, correct the pattern of living. They provide better education, for example, rather than more color television. Milton Friedman rejects such paternalism. He feels that whoever prefers color television to better health and housing should get it.

The Commission seems to side in its proposals with Professor Galbraith, who in essence says that the individual consumer, rich and poor alike, does not always know what is best for him. Some choices would be better made for him by the government. The Report does not spell this out—except in the fine print of some of the dissenting comments. But this might again be the price the Commission has to pay for arriving at a broad general consensus.

Like Generals who are said to be always well-prepared to fight past wars, the New Economists are still laying detailed plans to combat the Great Depression of the Thirties. Technological unemployment, like an atomic attack, might seem to be “unthinkable” but this should not prevent us from thinking hard about it. The Commission on Automation has skillfully accomplished the difficult feat of supplying some imaginative answers, while abstaining from posing clearly and sharply the most relevant questions.

This Issue

December 15, 1966