Chapter XXVI “FULL EMPLOYMENT” AS THE GOAL

The “contribution” of Keynes that his disciples most often insist upon as valid and “permanent” is the substitution of “full employment” as the goal of economic activity rather than the “maximum production” of the classical economists.

We shall ask here three main questions about “full employment.” 1. Is it definable? 2. Is it attainable? 3. Is it— at all times and under all conditions—even desirable?

1. Is It Definable?

Let us begin with the question of definition. The man in the street has few misgivings about this. “Full employment” means that “everybody” has a job. It means “jobs for all the people all the time.”

This naive conception runs into immediate difficulties. Early in 1958, for example, the population of the United States was about 173 million. But there were only some 62 million employed. Therefore there must have been 111 million “unemployed”! Yet the official estimate was that there were at that time only 5 million unemployed.

For the government statisticians, the “unemployed” consist only of those in the “labor force” who are not employed. But just how was the line drawn between the 67.5 million who were counted as part of the labor force and the 105.5 million who were not? Here is how the U. S. Bureau of the Census described how it decided:

Monthly estimates of the population of working age [14 years and over] showing the total number employed, the total unemployed, and the number not in the labor force are obtained from a scientifically selected sample of about 35,000 interviewed households in 330 areas throughout the country.

So the estimate of unemployed was in large part based on a sample of only one in every 1,400 households in the country.

My purpose here, however, is not to emphasize the probable error in such estimates, but to call attention to the necessarily arbitrary and in some cases purely subjective standards by which “unemployment” is officially determined.

The Bureau of the Census’s explanation continues: “The unemployed total includes all jobless who were looking for work.” How is the number of such persons estimated? From replies to the interviews. What constitutes realistically looking for work? The interviewers must rely in large part upon the realism of the replies. The labor force is not even a constant percentage of the total (“non-institutional”) population. In July of 1957 it was 60.6 per cent; but in December only 58.1 per cent.

Some paradoxical results emerge. The monthly report for March of 1958, for example, opened as follows: “Employment rose by 300,000 between February and March... while unemployment was unchanged.” How could that happen? The layman would naturally expect that if employment rose 300,000 in March unemployment would have dropped that much. The government statisticians’ answer is that the “labor force” increased by that much.

The “labor force” increases partly by census estimates of the population reaching working age, etc., but also partly by changes in peoples’ decisions. Suppose a man has a good job, with a wife at home and a son and daughter in college. He loses his job, whereupon not only he, but his wife, his son, and his daughter start looking for work. Because one person has lost his job, four persons are now “unemployed.” So “unemployment” goes up faster than employment goes down.

Let’s turn now to the explanations of the Department of Labor:

Effective January 1957, persons on layoff with definite instructions to return to work within 30 days of layoff and persons waiting to start new wage and salary jobs within the following 30 days, are classified as unemployed. Such persons had previously been classified as employed.... The combined total of the groups changing classification has averaged about 200,000 to 300,000 a month in recent years.

So the “unemployed” increased about a quarter of a million simply by a change of definition!

We get into the same kind of problems and arbitrary decisions when we come to the matter of working hours. Obviously there cannot be jobs for “all the people all the time.” We must deduct time for eating, sleeping, rest, and leisure. But how much time? It is customary to think of men being “partly unemployed” when they are laid off for two working days a week. But obviously they are as much unemployed if they work every day for correspondingly reduced hours. In the United States today, the standard full working week is forty hours, or five eight-hour days. This is shorter than the standard working week used to be, and in future it may be shorter still. Obviously the length of the working week that constitutes “full employment” is also a matter of arbitrary and conventional definition.

Let us see whether we can get some help from the academic economists, and first of all, of course, Keynes.

In the General Theory, Keynes gives us two definitions, neither of which seems to have much relation to the other. On page 15 he gives an involved definition of “involuntary” unemployment which, as I have already tried to show (p. 30), is invalid. From this he postulates a state of affairs in the absence of “involuntary” unemployment: “This state of affairs we shall describe as ‘full’ employment, both ‘frictional’ and ‘voluntary’ unemployment being consistent with ‘full’ employment thus defined” (p. 16). In other words, “full” employment is a state in which there can be both “frictional” and “voluntary” unemployment! Full employment is not full.

Let’s start again, this time with the definition on page 303:

We have full employment when output has risen to a level at which the marginal return from a representative unit of the factors of production has fallen to the minimum figure at which a quantity of the factors sufficient to produce this output is available.

I confess I find it difficult to follow this jabberwocky; but I assume it implies that some sort of equilibrium has been reached. One is tempted to ask irreverently: Does this mean that Uncle Oscar has a job?

Let us turn to A. C. Pigou. Professor Pigou is aware of some of the difficulties we encounter when we try to define unemployment:

A man is only unemployed when he is both not employed and also desires to be employed. Moreover, the notion of desiring to be employed must be interpreted in relation to established facts as regards (1) hours of work per day, (2) rates of wage, and (3) a man’s state of health.¹

This definition reveals that many subjective and arbitrary elements enter into the concept of “unemployment.” But we shall see presently that there are many more difficulties than even Pigou’s definition allows for.

After considerable discussion Pigou ends up with the conclusion that “The quantity of unemployment prevailing at any time is equal to the number of would-be wage-earners minus the quantity of labor demanded plus the number of unfilled vacancies.” ²

It is important to notice here that the “number of would-be wage-earners” is not only largely a subjective rather than an objective quantity, but that “the quantity of labor demanded” and “the number of unfilled vacancies” are also largely subjective, rather than objective, because they depend on the changing intentions of employers. If I could get a man to mow my lawn at a certain hourly rate, there would be an “unfilled vacancy” at that rate, but if the professional gardeners available all demand more, I may decide either to mow my lawn myself, or to let it grow.

This principle applies throughout industry. Whether there are “unfilled vacancies” in a given firm may depend not only on the wage-rate at which the vacancies could be filled but whether employees with certain special qualities could be obtained.

In another place Pigou writes:

A contrast is often drawn between situations in which there are more men available for jobs than jobs available for men and situations in which there are more jobs than men. In the former class of situation we have less than full employment, that is unemployment, in the latter more than full employment, that is over-full or, more briefly, over-employment.³

All this looks extremely simple. But the difficulty of statistical quantification, of deciding precisely what the numerical relationship is of “men available” to “jobs available,” is precisely the difficulty of defining not only what is meant by “men available,” but what is meant by a “job,” particularly when it is “unfilled.”

Let us try Sir William Beveridge. In his Full Employment in a Free Society, he defines full employment in his opening pages as “having always more vacant jobs than unemployed men.... The labor market should always be a seller’s market rather than a buyer’s market.”

But this is obviously defining full employment as overemployment. (Incidentally, Sir William’s demand that there must always be more vacant jobs than unemployed men is a demand that labor should be always underpaid. For this condition could only exist if the marginal product of labor were higher than its wage-rate, and labor, though “fully employed,” would therefore be getting less than its full potential income.)

One of the most realistic discussions of the difficulties in the concept of full employment is that of Edwin G. Nourse, formerly Chairman of the President’s Council of Economic Advisers. Commenting upon the declaration of policy in the American Employment Act of 1946, he writes:

The phrase, “those able, willing, and seeking to work,” does not define a labor force for whose optimum utilization the federal government can, in good economic conscience, pledge itself to “utilize all its plans, functions, and resources.” In the absence of objective criteria, the word “able” becomes practically meaningless. Whether a given person is, in a commercial or industrial sense, able to work is a decidedly relative matter. Able to work steadily or only intermittently? At the kinds of work for which demand presently exists, only with other skills, or without any particular skill, aptitude, or even teachability? Able to work as determined by a doctor’s certificate or by a foreman’s report? Under standard shop or office conditions or only with special facilities or treatment? Equally rich in ambiguity is the companion term “willing.” It was inserted as a gesture of reassurance to those who feared the camel of authoritarianism might be getting his nose under the tent of free enterprise. But does it mean willing to work at such jobs as are available or only at the job of one’s dreams? Willing to work on a time schedule dictated by employers’ needs or by workers’ convenience? Seeking is, of necessity, the criterion relied on by the Census Bureau in giving us a monthly estimate of involuntary unemployment. But “wanting” would be a more apt term for our purpose since it is a commonplace in the experience of all who have dealt with the unemployed to find not a few persons who want work—may even need it desperately—and who yet are not actively seeking a job because they have become convinced that the search is hopeless.

The plain fact is that the size of our labor force is statistically determinate only within the limits of quite categorical definitions.⁴

When we speak of full employment, therefore, we would do better to use the term not as the Keynesian zealots use it, and not with any effort at an unattainable mathematical precision, but in a loose, common-sense way to mean merely the absence of substantial or abnormal unemployment.

If it be objected that this is not in fact a definition of full employment (and it is certainly not brimful employment), then I suggest that the term might be dropped entirely, and the term optimum employment used instead. This would have, among other advantages, that of reminding the user as well as his audience that employment is rather a means than an end, and that its optimum size is relative to other conditions or goals.

2. Is It Attainable?

Is “full employment” attainable? Here even those who favor the goal begin to waver. Alvin Hansen, in his definition of full employment in Economic Policy and Full Employment declares that: “In an economy as large as that of the United States, it is probable that at ‘full employment’ there would be at any one time between 2 and 3 million temporarily unemployed.” ⁵ (About 4.5 per cent of the civilian labor force of 1945-47.)

Paul Douglas, commenting on Beveridge’s use of a 3 per cent margin for seasonal and transitional unemployment, declared that such a criterion would be “fatal” in the United States: “To use deficit financing to drive unemployment down below 6 per cent is very dangerous. It will tend to do far more harm through inflation than the good it will do by absorbing those who are unemployed from seasonal and transitional causes.” ⁶

3. Is It Unconditionally Desirable?

So when “full employment” is seriously discussed, it turns out to be less than full employment. And the desirability of “full employment at whatever cost” ⁷ is gravely questioned.

Lionel Robbins,⁸ quoting the Beveridge definition of full employment as “having always more vacant jobs than unemployed men,” points out that: “A state of affairs in which, at current rates of wages, the demand for labor is continually greater than the supply, must be a state of affairs in which, in the absence of special restraints, the level of wage rates, and hence the level of prices, is tending continually to rise.” He goes on to point out that even a full employment policy that tried to guarantee a mere equality of jobs and applicants would have to guarantee the trade unions that “whatever [wage] rates they succeeded in getting, unemployment would not be permitted to emerge.” Professor Robbins concludes that Beveridgian full employment “tends to inflation, reduced adaptability, external disequilibrium and a most drastic curtailment of individual liberty.”

Edwin Nourse, in the article from which I have previously quoted, declares that: “Ideally full employment would be such as promotes continuous maximization of production and real purchasing power for the people.” But this definition recognizes that full employment is desirable, not as an end in itself, but only as a means to much broader ends. Even the “maximization of production” must be understood, not in the sense of the mere piling up of physical things, but in the sense of the maximization of consumer satisfactions. And this includes also, for example, the “production” (or “consumption”?) of more leisure at the cost of less desired (physical) things.

If we are talking not of unavoidable means but of desired ends, then we must recognize that the economic objective of mankind is not more work but less. I hope I may be forgiven for quoting what I have written in another place:

The economic goal of any nation, as of any individual, is to get the greatest results with the least effort. The whole economic progress of mankind has consisted in getting more production with the same labor. It is for this reason that men began putting burdens on the backs of mules instead of on their own; that they went on to invent the wheel and the wagon, the railroad and the motor truck. It is for this reason that men used their ingenuity to develop a hundred thousand labor-saving inventions.

All this is so elementary that one would blush to state it if it were not being constantly forgotten by those who coin and circulate the new slogans. Translated into national terms, this first principle means that our real objective is to maximize production. In doing this, full employment—that is, the absence of involuntary idleness—becomes a necessary by-product. But production is the end, employment merely the means. We cannot continuously have the fullest production without full employment. But we can very easily have full employment without full production.

Primitive tribes are naked, and wretchedly fed and housed, but they do not suffer from unemployment. China and India are incomparably poorer than ourselves, but the main trouble from which they suffer is primitive production methods (which are both a cause and a consequence of a shortage of capital) and not unemployment. Nothing is easier to achieve than full employment, once it is divorced from the goal of full production and taken as an end in itself. Hitler provided full employment with a huge armament program. The war provided full employment for every nation involved. The slave labor in Russia had full employment. Prisons and chain gangs have full employment. Coercion can always provide full employment....

The progress of civilization has meant the reduction of employment, not its increase. It is because we have become increasingly wealthy as a nation that we have been able virtually to eliminate child labor, to remove the necessity of work for many of the aged and to make it unnecessary for millions of women to take jobs.⁹

¹The Theory of Unemployment, (London: Macmillan, 1933), p. 3.

²Ibid., p. 10.

³ A. C. Pigou, Essays in Economics, (London: Macmillan, 1952), p. 108.

⁴ “Ideal and Working Concepts of Full Employment,” American Economic Review, May, 1957, p. 100.

⁵Loc. cit., p. 19n.

⁶Economy in the National Government, (1952), p. 253.

⁷ See Jacob Viner, Quarterly Journal of Economics, August, 1950.

⁸The Economist in the Twentieth Century (London: Macmillan, 1956), “Full Employment as an Objective.”

⁹Economics in One Lesson (New York: Harper, 1946), pp. 68-70.