The Specter of "Unused Capacity"
JULY 01, 1978 by HENRY HAZLITT
Henry Hazlitt, noted economist, author, editor, reviewer and columnist, is well known to readers of the New York Times, Newsweek, The Freeman, Barron’s, Human Events and many others. Best known of his books are Economics in One Lesson, The Failure of the "New Economics," The Foundations of Morality, and What You Should Know About Inflation.
One of the most frequent excuses for inflation is that if a little extra money is printed its effect won’t be to raise prices but only to increase the volume of sales and production; because at the moment the new issues of money are being recommended industry is not working at "full capacity."
In the month, say, that the new dose of inflation is being advocated, the official estimates show that industrial plants are working at only 70 or 80 per cent of capacity. Therefore, when the new inflation puts more money into the hands of consumers, they will use it to buy more goods. Manufacturers will simply increase their production to meet the increased demand, and prices will not rise until after plants are working at "full capacity" and cannot increase output further. At that point the issue of new money can simply be stopped.
The writer mainly responsible for the popularity of this theory is John Maynard Keynes. It is akin to the same writer’s full-employment argument. It is, in fact, part of the same argument; because for Keynes the supreme economic goal, the summum bonum, was the uninterrupted full employment of men and resources. What the cost of achieving this might be in other respects was simply disregarded.
It may be thought that the criticisms that the present writer and others have already made of the "full-employment" goal, and of the argument that inflation is the way to achieve it, must apply equally to the "full-capacity" goal, and therefore need not be repeated. But though the criticisms are of the same general nature, an analysis of the fallacies of the "full-capacity" goal makes it possible to bring out with much greater sharpness some of the naiveties and errors in the "full employment" goal as well.
Varying Views or Definitions of "Full Capacity"
We must begin with a definitional question. What is "full capacity?" The question is seldom raised in popular discussion; but as soon as we examine the problem seriously, we find a wide range in possible definitions. If we think of full capacity from a purely engineering standpoint, then we must think of what could be turned out if factories were operated around the clock, twenty-four hours a day, seven days a week. We would then also have to assume unlimited supplies of labor, with the exact types and mix of skills required, working three or four shifts a week, as well as unlimited supplies of raw materials and other inputs.
A situation like this may be actually possible or desirable in a few industries in wartime or even for a few weeks or months in peacetime; but it would obviously involve mounting problems. Hardly any economist would regard it as an ideal state of affairs.
A second concept of full capacity would envisage maximum output under a "normal" operating schedule—with the customary number of hours per shift and days per week, with downtime for repair and maintenance of machinery. If this concept also assumed high-cost, inefficient facilities brought into production, the resulting output might be defined as the maximum practical capacity. This is the figure commonly used in the official estimates of unused capacity rates. But this figure refers to potential physical capacity rather than to the optimum rate from an economic standpoint. Few companies want to push their output to the maximum practical level. They would prefer to hold it to the level that achieves maximum long-run profits or other objectives. This involves the assumption that they can obtain all the inputs they need at existing costs per unit and that they can sell unlimited quantities of output at existing prices. It also involves the assumption that they will not be forced into continuous use of their comparatively obsolete equipment.
This output level has been called "preferred capacity."¹
A Department of Commerce study found that for all manufacturers, the preferred operating rate during the period from 1965 to 1973 was 94 to 95 per cent, considerably above the actual rates.
There are several periodic estimates published of unused manufacturing capacity rates. The two most widely cited are that of the Bureau of Economic Analysis in the Department of Commerce, and that of the Federal Reserve Board. There are also a few private estimates, notably by McGraw-Hill Publications and by the
All use slightly different methods. The BEA (Bureau of Economic Analysis) obtains its rates by a survey of some 3,000 companies. The respondents generally calculate their utilization rates against maximum practical capacity. It is obvious that each individual answer must itself be an estimate rather than a precisely known figure.
This is one reason why we cannot depend on the accuracy of the index. As Alan E. Shameer, associate economist of the General Electric Company, put it: "We have dozens of different plants, producing everything from jet engines to plastics to coal to washing machines. How can we possibly say with precision that the company is operating at such-and-such a rate of capacity?… It’s a jelly-like concept."2
Gathering the Statistics
If we take the BEA figures of capacity utilization rates for all manufacturers for the eight-year period from December 1965 to December 1973, we find that they ranged from a peak of 87 per cent in June 1966 to a trough of 79 per cent in September 1970. The difference between the peak and trough rates, in other words, was only 8 percentage points. More recent figures tend to show a somewhat wider range. For example, the Federal Reserve Board figure of capacity utilization for all manufacturing in 1974 was 84.2 per cent, and in 1975 it was 73.6 per cent, a difference of 10.6 percentage points within a single year.
The FRB and BEA figures do not today tend to differ widely: the FRB estimate of average capacity utilization in 1976 was 80.1 per cent, and the BEA figure 81.2. But a major effort to improve its past statistics was made recently by the Federal Reserve Board, when it started to take into fuller account, among other things, operations at relatively small companies. The upshot was that factory operations as a whole turned out to be much higher than the Fed originally had supposed. For the 1976 third quarter, for example, the plant-operating figure was boosted sharply to 80.9 per cent from the previous 73.6 per cent. Perhaps further investigation may result in further revision of the figure, up or down. This once more raises the question whether the utilization-rate figure is worth using as a "policy-making tool"—even if we were to grant that government bureaucrats should ever attempt to "fine-tune" the economy.
Averages Hide Variations
When we ask why the cyclical range in the official utilization-rate estimates has not been greater, the main answer is clear. These figures represent the average capacity utilization rate of all plants in all industries. Averages tend always to conceal wide divergence and dispersion. In addition to its overall figure, the BEA gives separate capacity-utilization rates for about a dozen different leading industries. We have seen that the spread between the peak and the trough rates of capacity utilization for all industries from 1965 to 1973 was only 8 percentage points. But the spread in the (non-electrical) machinery industry was 15, in the rubber industry 22, and in the motor-vehicle industry 42 percentage points.
Even here, however, the real disparities between capacity utilization in different plants and factories were largely concealed because the foregoing figures are again the average figures for entire industries, lumping the marginal and the most successful companies together.
To make the real problem clearer: Let us suppose that at the moment the average capacity utilization rate for all manufacturing is 80 per cent. A Keynesian might then say that if we increased the money supply by 20 per cent the result would be stimulating but not inflationary, because this new money would merely supply the purchasing power to buy 20 per cent more goods, and industry already happens to have the idle capacity to turn out that much more goods "without inflation" or unwanted price increases.
But suppose this 80 per cent average figure, though reasonably accurate, conceals a real situation in which the capacity utilization rate in different plants actually ranges from a low of 60 to a high of 100 per cent, with the lowest 11 per cent of plants operating at only 60 per cent, the next 11 per cent segment above that operating at 65 per cent, the third segment at 70 per cent, and so on, with the ninth and highest segment operating at full capacity.
Supposing the Keynesian scheme otherwise operates in accordance with the schemers’ intentions, what would be the result? All factories would be operating, or trying to operate, at a rate 20 per cent higher than before. The half that had been operating at less than 80 per cent could presumably do this, but the half that had already been operating above that rate would be running into bottlenecks and shortages in plant and equipment, not to speak of the problems of all manufacturers in buying additional specialized inputs and hiring additional specialized labor. Prices—and wage rates and other costs—would begin to soar.
(Of course the neat and even distribution of dispersion that my hypothetical figures suggest would not occur. I have assumed it merely to simplify the exposition. But it is important to keep in mind that there is bound to be some such dispersion.)
Our analysis brings out the simplistic and completely unreal nature of the Keynesian assumptions, and of so-called "macroeconomics" in general. This macroeconomics deals almost exclusively in averages and aggregates. In doing so it falsifies causation, and neglects individual processes, individual industries, individual companies, individual prices, and the immense diversity of services and products.
The Keynesian economic heaven is apparently one in which there is constant full employment of men and plants and equipment. Nobody and no machine is temporarily idle because the economy is in transition. The balance and proportions among the thousands of individual industries and products remain constantly the same. No industry is contracting and laying off help because of declining demand for its particular product, and therefore no capital and labor can be released so that other industries can expand. No processes, machines, or plants become obsolete because of new methods or new inventions, and therefore never have to be shut down, idled, or scrapped. Every industry is apparently turning out a homogeneous and unchanging product, and can hire additional workers from a sort of homogeneous labor pool. There is no such thing as a surplus or shortage of specialized skills. Unemployment is solely the result of "insufficient purchasing power," and can be remedied simply by increasing that purchasing power.
Not only could such an economy exist only in some never-never land, but no serious economist could regard it as desirable. It is the result of turning full employment and full utilization of capacity, which are merely means, into the overriding economic end.
Let us turn our attention to a few actual consequences of Keynesian policies that the Keynesians chronically overlook.
They assume that an increased money and credit supply—as long as there is not "full employment" and the economy is not "operating at full capacity"—will not lead to increased wage-rates or increased prices because industry will simply hire previously idle labor and turn out more goods to take care of the increased demand.
This assumption neglects two factors. The first is that average or overall unemployment and average or overall unused capacity are not what count. The percentage of unemployment is different in every industry and locality, and the percentage of unused capacity is different in every plant. When general or special demand increases, shortages will quickly occur at particular localities, of workers with special skills, and bottlenecks will soon develop in individual industries, factories and plants. Capacity is reached when we have fully employed our most scarce resource or complementary productive factor, whether that is an important key industry, or specialized labor, plant, or some raw material. When this situation occurs the price of the scarce factor or factors will begin to soar, and this rise will soon force increases in other wage-rates and prices.
There is a second overlooked factor. Even if the distribution of both unemployed labor and unused capacity were uniform, increased demand would in any case promptly bring a rise of wage-rates and prices. Intelligent speculators (and every businessman and even every consumer must be to some extent a speculator) do not wait until there is an actual shortage of anything before they start bidding up prices; they do this as soon as they foresee the possibility or the probability of a shortage. And the greater the probability seems, the higher they bid. Every successful businessman tends to be successful in proportion to his ability to anticipate a change in conditions—to buy or sell before his competitors or the general public are aware of the coming situation. It is only the Keynesians’ blindness to this everyday fact of business life which leads them to assume and predict that new issues of money will not result in inflation until every man is employed and every factory is going full blast.
Seasonal Fluctuations, and the Need for Reserves
Let us come back now to the specific problem of unused plant capacity. The Keynesians seem to assume that it is both possible and desirable for all plants to work continuously at full capacity. It is neither. The demand for all sorts of products—motor boats, snowplows and lawnmowers, skis and roller skates, overcoats and bathing suits—is seasonal. For that or other reasons, their production tends to be seasonal (even though the output season precedes the selling season). In order that there may be sufficient production at the peak of the season, there must be at least some unused capacity off-season. The unused capacity does not necessarily mean economic waste; it is its availability when needed that counts.
For the same reason, when a firm’s plant has been working at full capacity for more than a short period, it is probably a sign that the firm has missed an economic opportunity. It should have foreseen this situation and expanded its plant or built a new one to meet the increased demand for its product. Producers, in fact, constantly try to do just this. It has long been recognized that in periods of low operating rates industry does not tend to expand, but that as operating rates increase, there is an increase of investment in new plant. Businessmen recognize that they must normally accept some "surplus" capacity in order to be sure they will have enough when they need it. Not only is it unprofitable for them to be fully using their more obsolescent plants and machines, but they should be periodically replacing them with the most modern and efficient equipment. In brief, the most desirable normal situation for the individual plant owner or manager is one in which there is at least some "unused capacity."
Inflation Builds Uncertainty
In a recent penetrating study,3 however, M. Kathryn Eickhoff, vice-president and treasurer of Townsend-Greenspan & Company, pointed out that from 1970 till the date her study was made, increased plant operating rates were not leading to investment in new plant as early as they previously did. The "trigger point" that set off new investment seemed to be moving to higher and higher operating rates. That trigger point in 1977 seemed to be an average capacity utilization rate of approximately 87 per cent. This was ominous, because the highest rate in the preceding recovery was 87.6 per cent during 1973, the year the nation moved into double-digit inflation.
What this meant, among other things, was that increased issues of money and credit were tending to lead to output shortages sooner than previously, and therefore were leading to sharper and higher price rises sooner than previously.
Miss Eickhoff also presented an acute analysis of the reasons why inflation and inflationary expectations increase uncertainty and thereby discourage new investment. The greater the uncertainty in the business outlook, she pointed out, the greater becomes the rate of return required for new investment to compensate for that uncertainty, and the fewer the number of projects which will qualify. Inflation, especially when it is expected to accelerate, always increases business uncertainty. Even if overall profits advance in line with the rate of inflation, no single producer can be sure that his profits will rise to the same extent. That will depend upon how much his costs rise relative to all other prices in the economy, and whether or not he can raise his prices correspondingly. As a result, the dispersion of profits among producers increases as the rate of inflation climbs. This dispersion of profits does far more to discourage investment than the prospect of an overall increase of profits does to induce it. In effect, a much higher rate of future discount is applied to inflation-generated profits than to those resulting from normal business operations.
Thus the inflation that the Keynesians and others advocate in order to stimulate employment, production, and investment ends by discouraging, deterring, and diminishing all three.
1For a thoughtful discussion of these conceptual and definitional problems, see "The Utilization of Manufacturing Capacity, 1965-73" by Marie P. Hertzberg, Alfred I. Jacobs, and Jon E. Trevathan, in the Survey of Current Business, July 1974, pp. 47-57, published by the U.S. Department of Commerce.
2The Wall Street Journal, March 11, 1977.
³”Plant vs. Equipment Considerations in the Capital Goods Outlook." Presentation before