6. Compensating Reactions to Compensatory Spending*

In the plans for maintaining full employment after the war that flood the country, governmental “compensatory” spending plays a major role. Only a few proposals mention the question of adjustment that was so prominent in the earlier literature on the liquidation of booms. Most planners do not recognize the paramount importance of the wage level for employment. If they see it at all, the importance given wages in sustaining a high effective demand far overshadows the importance given them as a cost factor. To be sure, there is an opposition. To it adjustments remain a major problem, and compensatory spending in an unadjusted economy seems of highly doubtful value. But in view of the overwhelming number and influence of those who favor “spending without adjustment”¹ and the appeal of their plans to laymen, the opposition must be considered “voices in the wilderness.”

To anyone watching the trend of economic thought in this country the situation is not astonishing. It is just an outgrowth of the general acceptance of Keynesianism. For, though some planners² are distinctly out-Keynesing Keynes in a way that Lord Keynes himself would surely reject, the planners’ basic attitude is distinctly an application of Keynes’ General Theory.³ According to this theory, employment can, “as a rule” and “in the general case,” be raised or restored by raising effective demand to a sufficiently high level. Why, then, should there be any need for a painful adjustment process?

In this chapter the author tries to explain what seem to him to be the fallacies not merely of some of Keynes’ implications but also of the general assumption underlying the entire system.⁴

THE ILLUSION EFFECT OF MONETARY MANIPULATIONS IN CREATING EMPLOYMENT

Lord Keynes contends that “as a rule” and in the “general case” an expansion in “effective demand” increases employment as well as prices. “Effective demand spends itself partly in affecting output and partly in affecting price.”⁵ The reason is—to put Keynes’ argument in the simplest form—that additional labor can be used profitably despite its diminished marginal productivity. This, in turn, is because “the decreasing return from applying more labor to a given capital equipment has been offset by the acquiescence of labor in a diminishing real wage.”⁶ And labor acquiesces in a diminishing real wage, because usually “. . . the supply of labor is not a function of real wages. . . .”⁷ For “it is not their [the workers’] practice to withdraw their labor whenever there is a rise in the price of wage goods.”⁸

However, “a point comes at which there is no surplus of labor available at the then existing real wage.”⁹ As soon as this point is reached, the supply price of labor is fixed in accordance with the declining purchasing power of wages. In other words, the supply curve of labor in terms of money wages moves upward with prices which rise because of the expanding effective demand. From this point on no additional unit of labor can be applied profitably. The “crude quantity theory of money” once again functions. “Output does not alter and prices rise in exact proportion to [the quantity of money].”¹⁰

This is doubtless a correct picture of the process of monetary expansion. It is generally agreed that monetary expansion in its first phases increases employment. But from a certain point on, which we may call the “reaction point,” reactions on the side of the productive factors compensating the effect of credit or money expansion will set in, and prevent a further rise in employment or even bring about a decline.¹¹ And this will be true even in the case not mentioned by Keynes—when price increases should have been avoided, diminishing unit costs offsetting the effects of diminishing marginal efficiency of labor. In this case entrepreneurs derive extra profits which labor, from a certain point on, will claim for itself,¹² just as it claims for itself, and with success, the increments technical progress brings to its productivity.

Where does the difference between Keynes and the classical attitude lie? Keynes assumes that the period before the “reaction point,” the period free from “compensating reactions,” is so long and general that it is characteristic of “the economic society in which we actually live,”¹³ and is thus a sufficient basis for a “general theory of employment.” The classicists, on the other hand, consider the “reaction-free” period as usually short and occurring only exceptionally. This apparently unimportant difference in factual assumptions is, as far as I can see, responsible for all the wonders of the Keynesian world so paradoxical to classical thinking.

What prevents labor during the “reaction-free period,” whether short or long, from raising its demands for money wages to correspond with its declining purchasing power and/or the profits resulting from increased sales? After all, men work for food, clothing, etc., not for pieces of paper, even if dollar amounts or other denominations are printed on them. Keynes does not answer the question; he merely states the facts so important to his system. It is a complicated sociological economic problem that is at stake. But one thing seems certain. If, during the “reaction-free period” labor does not insist on money-wage increases, it is not because it wishes to receive lower real wages. It can only be because it does not, or does not immediately or fully, realize what is happening when prices begin an inflationary rise. What works is the phenomenon Professor Irving Fisher described in his famous book The Money Illusion¹⁴ and what we may therefore call the “illusion effect” of monetary manipulations.

THE ILLUSION EFFECT INDUCING INVESTMENT

According to Keynes’ theory, a larger effective demand, leading to a higher level of employment, will depend, given a certain propensity of the community to consume, on the amount of current investments. “The amount of current investment will depend, in turn, on what we shall call the inducement to invest; and the inducement to invest will be found to depend on the relation between the schedule of marginal efficiency of capital and the complex of interest rates on loans of various maturities and risks.”¹⁵ So by lowering the interest rate, investment and effective demand can be increased.

Increasing effective demand through lowered interest rates is what European writers used to call “inflationary credit expansion.” It depends not on any spontaneous decision of the community to save more, but on the will and capacity of the banking system to expand the amount of credit and the quantity of money. According to the classic approach it is a case of monetary manipulation.

A downward monetary manipulation of interest rates can undoubtedly induce an increase in investments and effective demand, especially as “rising prices . . . will redistribute incomes to the advantage of the entrepreneur and to the disadvantage of the rentier,”¹⁶ and as this is equivalent to a further decline in the interest rate or even to a negative interest rate. But just as in the case of lowering wages through monetary manipulation, investments are induced only before the “reaction point” is reached. Then the productive factors whose rewards, although nominally unchanged, have really been lowered will react. As soon as the supply curve of credits is raised in accordance with the decreasing value of money, the investment-inducing effect of interest manipulation disappears. Similar developments which can, incidentally, be observed during every business cycle are the basis of all monetary cycle theories of the Wicksellian type.

That the illusion effect of money at first prevents compensating reactions was demonstrated drastically during the great German inflation. Until about the middle of 1922 the majority of the population, especially the creditors, were not aware of what was happening. They were deceived by the “illusion effect.” Loans were still offered in ample quantities and at low rates. When the creditors were no longer taken in by the money illusion, they raised their demands for interest to fantastic levels, wishing to be compensated for the decreasing purchasing power of their money during the lending period. The Reichsbank, thinking it should not tolerate this healthy compensating reaction, tried to keep the rates down by maintaining a ridiculously low discount rate. This low discount rate was one of the chief reasons for the runaway character that inflation in Germany assumed in 1922.¹⁷

THE ILLUSION EFFECT OF GOVERNMENT SPENDING

If the manipulation of interest rates downward does not induce sufficient investment, and it probably will not, compensatory deficit spending by the government is recommended. “The State which is in a position to calculate the marginal efficiency of capital-goods on long views and on the basis of the general social advantage,” has, as Keynes says, to take “the responsibility for directly organizing investment.”¹⁸

In the case of government spending, too, investment and employment increase because the supply price schedules of the productive factors have, through manipulation, become “really” lower than they seem. Here, too, sooner or later the point is reached where compensating reactions prevent further improvements, or even reverse already achieved improvements. Until then it is the illusion effect of monetary manipulations that prevents compensating reactions.

When the efficiency of a further unit of capital is too small to cover the interest charges, private enterprise refrains from new investments. If governments can invest where private enterprise cannot, it is not because capital is supposed to be more efficient in its hands. (Professor Hansen, for instance, warns against “timidity” and urges consideration of a 50 per cent recovery of principal—not of interest!—as a sufficient return on invested capital.¹⁹) It is because the interest rates charged the government are much lower than the market rate, or, if capital is allowed to be 50 per cent unrecoverable, even negative. A negative interest rate means that the entrepreneur does not have to make payments but receives payments from those who lend the capital. This is just what happens in government deficit spending through shifting of the deficit in one way or another to the community, which becomes liable for the amounts.

What bearing the liability will have on the economy depends on whether and to what extent the members of the community have to redeem it by tax payments. Distinction is made in this respect, in literature, between the liability for interest and for the principal.

Sooner or later the day must come when interest no longer can be paid through issuing new government securities, but has to be paid out of taxes. This happens either when private investment is picking up so that government deficit spending has to be discontinued entirely in order to check an excess of effective demand;²⁰ or it happens, at the latest, when the public refuses to take over the ever-increasing public debt, i.e., when what can be called the saturation point for government securities is reached.

If interest on the public debt has some day to be met by taxation, it will mean a heavy burden on postwar America. For it would come on top of taxation for interest on the war debt, which alone will swallow a substantial part of national income. Nor, incidentally, can the tax burden be minimized by pointing to the growth of the economy which will reduce the ratio of the debt to the national income. The inherent weakness of the “growth argument” is that it assumes the “growth” to be, so to say, natural,²¹ whereas it depends to a large degree on economic conditions and will probably never materialize if employment is made dependent upon government spending for any length of time. It is furthermore forgotten that the enlarged future economy will have its own larger problems, be they war, unemployment, or other. It does not therefore seem permissible to mortgage the growth.

The “only-the-interest” argument in contemporary literature on government spending is used to such an extent that the impression is created that only one-fortieth part of every deficit (2½ per cent of the capital) is the real burden, whereas the principal can be forgotten as a gift.²² Obviously, when this argument is used, the beneficial effects of spending are made to compare very favorably with the ensuing burden.

However, the “only-the-interest” argument is tenable only if spending can be discontinued because private spending has picked up. The argument is not tenable in the case of spending continued indefinitely in order to counteract a chronic tendency to unemployment.

A curve representing the amounts of bonds or money accumulated by the public does not mount steadily to a saturation point, to run horizontally thereafter. Curves representing data which depend on psychological factors—such as confidence in business prospects, in the value of the currency, and in the future level of prices—never remain on a plateau but rise and fall according to the laws of action and reaction. In other words, once deflationary tendencies are relieved through inflationary tendencies, it becomes highly probable that—through the unloading of previously accumulated money and government securities—deflation turns into inflation, not into stabilization.

What must be done to prevent runaway inflation at such a time? It is not sufficient simply to stop further deficit spending. Certain amounts of existing public debt become due, and these amounts will be larger the smaller the portion of the debt which has been consolidated to longer terms. Suddenly, what seemed a gift for eternity is transformed into a real loan. The bill must at last be paid. In addition to taxation covering all public expenditures, business will then face the burden of high interest rates. For, in order to stem the demand of those who want to profit from the dwindling purchasing power of the currency and therefore borrow from the banks, interest rates will have to be raised. This also hurts legitimate enterprise. In short, a very severe deflation crisis will occur.

The result of all this is that interest on the public debt must be met by taxation. The principal, or at least parts of it, will also some day become a tax burden if government spending is permanent. If this is so, government deficits mean for the economy that net wages, net profits, and net interest received by the factors of production are really lower than their gross earnings. For from gross earnings the amounts should be deducted that will have to be paid as taxes in the future and thus represent a mortgage on present income. Owing to the illusion effect, this is not at once realized.

There is no miracle in government spending. The fundamental fact remains that investment and employment increase only when the supply schedules of the contributors are lowered in real terms. The difference is merely that government spending effects the reduction indirectly and unobtrusively, as does every inflation and monetary manipulation in general.

Incidentally, our analysis shows not only the mechanism at work when one tries to overcome by government investments the alleged “secular stagnation” of the economy, caused by insufficient profitability of new capital investments; it also shows that insufficient opportunities of capital investment can never be the reason for lasting unemployment—even if the “maturity” of the economy were proved.²³ For if investment becomes possible where the prices of the productive factors are lowered, excessive factor prices and not the low “efficiency” of capital are the secular cause of unemployment. Low efficiency of capital can explain why no new unit of capital is applicable to a given amount of labor, but not why no new unit of labor is applicable to a given capital. Even if the capital structure cannot be deepened, labor can be employed and savings utilized at the prevailing depth of the capital structure, if only labor is not more expensive than corresponds to its marginal efficiency. This has been recently restated with great clarity by Professor Pigou.²⁴

When employment is created by means of governmental deficit spending, the day will come when people realize that the real rates of earnings have been reduced and they will demand higher rates. Labor will not be satisfied with the prevailing wage level, less capital will be offered at the prevailing interest rate, and less entrepreneurial activity at the prevailing profit rate. All supply price schedules will move upward. Which of these upward movements will be the strongest depends upon whether labor, capital, or entrepreneurial earnings are expected to be taxed most heavily. The consequences for the structure of the economy are well known; in any case, a further increase in employment will not be possible. And if the government tries to compensate for the compensating reactions by spending still more, again still higher taxes will be anticipated, and so on in a vicious spiral.²⁵ All this will happen at the latest when the first taxes to meet the larger government obligations are to be levied.

THE “GENERAL CASE”

The fundamental difference between the classical and the Keynesian employment theory is one of factual assumption, not of theoretical analysis. Lord Keynes assumes that the state of monetary illusion is a normal state; that money wage, interest, and profit demands are normally not altered when the rates of earning no longer represent the same real value. The classicists assume as normal that the money illusion is always and immediately seen through and the supply schedules accordingly revised upwards, because people are interested only in their real, not their nominal income. These writers therefore contend that what they call monetary “falsifications” and even swindle do not really change the amount of employment but only the value of the currency. Thus, the whole question of whether Keynes’ theory and its practical consequences are acceptable boils down to this: does the “illusion effect” of monetary manipulation work so long and so regularly that it can rightly be used as the basis of a general theory of employment?

The world Keynes paints is not the real world. To realize this fully one has merely to compare his remarks with any newspaper report about the bargaining policy of labor. Generally and as a rule, whether we like it or not, the “money veil” nowadays is seen through most thoroughly and clearly. Wage demands do not remain unadjusted, for any length of time, to the sinking purchasing power of money, i.e., to higher living costs. They are demands for “real” wages. As a matter of fact, money wages have not lagged behind prices during the last decade. On the contrary, they have run ahead of prices; labor has succeeded in raising its standard of living because its wages have risen with the increase in its productivity.

Our conclusion is that the case Lord Keynes regards as the “general case” is in reality a special case, valid only under special conditions and for a certain time. His theory is a special theory of employment for the case when the money illusion works.²⁶

In the “general case,” the equilibrium which the economy attains through monetary manipulations is not real, definite, and stable, as Keynes claims, but at best transitory and dynamic. It yields to a real and stable equilibrium as soon as the supply schedules of the participants in the economic process are adjusted to the changes brought about through the manipulation.

CYCLICAL VERSUS STABILIZED UNEMPLOYMENT

In one special case Keynes’ scheme works: in the special case of the recovery phase of the business cycle after the liquidation of the preceding boom. For here, indeed, ideal conditions prevail for the money illusion. Here the demands for interest are still influenced by the memory of the low profits on capital during the depression. Here reductions of real wages are not watched closely and, if recognized, not followed immediately by reactions because real wages have only recently risen through the deflation of prices. And the burden of government spending is not yet taken into account; first, because at the beginning the amounts spent are not substantial; and secondly, because the decision as to which class will have to foot the bill is deferred and everyone gambles on the hope that it is the other fellow who will have to pay.

Consequently, monetary manipulations will be effective in shortening the transition period from a cyclical depression to recovery. Lowering interest rates below the prevailing market rates and governmental deficit spending are defensible, even advisable at this juncture.²⁷ But all this is nothing more than the discount policy, open-market policy, and fiscal policy recommended as a means of mitigating cyclical movements, long before Keynes, by almost every monetary business-cycle theorist.

Now there is no doubt that Keynes’ employment theory was conceived during and under the impression of such a cyclical prerecovery and recovery period. This alone can explain his factual assumptions which are typical for such periods but entirely atypical for other periods. On the other hand, Lord Keynes certainly does not intend his theory to be merely a theory of fluctuations in employment during business cycles; these are treated as a special case toward the end of his work. He deals with the establishment of stable equilibria with larger employment, as distinct from the increase of employment during the dynamic process of the cycle. He means his theory to be, chiefly, a theory of noncyclical and thus stabilized, or—to use the European expression—structural employment and unemployment;²⁸ in short, a general theory. And it is just and only as a general (not as a business-cycle) theory that it is original, challenging, and different from the classical. And it is at this point that there arises a phenomenon that is tragic for economic theory and dangerous for practical economic policy: what is really a theory of cyclical unemployment is formulated as a theory of structural unemployment. And once formulated, it leads its own life, detached from its premises, and becomes the basis and justification for policies concerning situations for which it is not valid, such as unemployment caused by wages which are structurally too high.

To this case Keynes’ scheme is not applicable.²⁹ In other words, neither lowering interest rates nor government compensatory spending is effective when unemployment prevails at a price level that is neither boom-inflated nor depression-deflated. The reason is simply that in this case the illusion effect does not work for any length of time and that the reaction period is therefore very short.

If wages have become structurally too high, it is merely because labor had considered them really too low. Obviously in such a situation rising living costs through monetary manipulation will immediately lead to compensating and (if we may judge by experience) even to overcompensating reactions.

If interest rates are lowered in the case of structural unemployment, creditors revise their interest demands upwards. For there is not the slightest reason why creditors should tolerate a redistribution of income to their disadvantage through inflationary credit expansion for any length of time.

Government spending to compensate cyclical unemployment can be stopped as soon as the special factors making for cyclical depression, especially those of a psychological nature, are checked. Spending to compensate structural unemployment has to go on indefinitely. Otherwise the level of effective demand would again be reduced and a deflationary process started, because private enterprise does not invest at the prevailing marginal efficiency of capital.

If government spending goes on indefinitely and therefore represents an ever-increasing burden on the community, the day must eventually come when it outlasts and outgrows the illusion effect, which is, by its very nature, transitory and limited. Compensatory reactions are inevitable.

It seems to be the tragedy of economic science that psychological phenomena like the money illusion become obsolete when they are discovered. If Lord Keynes has discovered the mechanism of lowering real wages through monetary manipulations, he has at the same time destroyed the working of the mechanism by drawing attention to it.

Only in connection with a policy aiming to adjust, rather than compensate, structural maladjustments will government expenditure be useful in the postwar period. Contrary to a widely accepted opinion, there exists no automatic and mechanical parallelism of spending and creation of employment. Nor do “unexhausted resources,” as such, guarantee that employment, and not prices, will rise in the wake of spending. If this is not recognized in time, postwar planning, far from bringing about full employment, will delay it by creating the illusion that maladjustments need never be corrected.

^* Appeared first in the American Economic Review, March 1945.

¹ For a good survey of postwar full employment plans, see Albert Halasi, “Survey of Recent American Literature on Postwar Security,” International Postwar Problems, Vol. I, 1943, pp. 120-38.

²Cf. Abba P. Lerner, “Functional Finance and the Federal Debt,” Social Research, Vol. 10 (1943), pp. 38-51.

³ John Maynard Keynes, The General Theory of Employment, Interest, and Money, New York, 1936.

⁴ As I take the position that in my Volkswirtschaftliche Theorie des Bankkredits (1st ed., Tübingen, 1920) I advanced a “credit expansion theory of employment” very similar to that of Keynes, I have added in the following notes after the citations from Keynes’ General Theory the numbers of the pages of my Volkswirtschaftliche Theorie on which the corresponding ideas are expressed.

⁵ Keynes, op. cit., pp. 13, 3, 285, and (in a slightly different wording) p. 296 (Hahn, op. cit., pp. 135, 146, 140, 141, 149, footnote).

⁶ Keynes, op. cit., p. 289, and in a different wording, p. 284.

⁷Ibid., p. 8.

⁸Ibid., p. 9.

⁹Ibid., p. 289.

¹⁰Loc. cit.

¹¹ The experience of the last phase of the German inflation illustrates this point. From about the middle of 1922 on, wages were made sliding according to the sinking purchasing power of money (Gleitloehne). As a result, employment no longer rose, but even declined. Inflation spent itself in price rises. For it is one question whether goods already fabricated are purchased at higher prices, and another whether new goods are fabricated. The latter depends on production being more profitable, i.e., wages and other costs not rising so fast as prices; a fact quite obvious though often forgotten in the wake of the spending enthusiasm of our time.

¹² Sumner H. Slichter, “Labor after the War,” in Harris, Postwar Economic Problems, New York, 1943, pp. 241-62: “Union wage policy will tend to keep the prospect for profits unfavorable, because unions will press for wage increases despite the continuation of price controls” (p. 245).

¹³ Keynes, op. cit., p. 3.

¹⁴ Irving Fisher, The Money Illusion, New York, 1928.

¹⁵ Keynes, op. cit., pp. 27-28. (Hahn, op. cit., 1st ed., pp. 132, 137.)

¹⁶ Keynes, op. cit., p. 290. (Hahn, op. cit., 1st ed., p. 137.)

¹⁷Cf. L. Albert Hahn, Geld und Kredit (Tübingen, 1924 and 1929) and Unsere Währungslage im Lichte der Geldtheorie (Frankfurt a. M., 1924).

¹⁸ Keynes, op. cit., p. 164. (Hahn, Volkswirtschaftliche Theorie, 1st ed., p. 151.)

¹⁹ Alvin H. Hansen, “The Postwar Economy” in Seymour E. Harris, Postwar Economic Problems, New York, 1943, p. 23.

²⁰Cf. Lerner, op. cit., p. 43.

²¹Cf. Joseph Stagg Lawrence, Empire Trust Letter, No. 6, p. 5, New York, Empire Trust Company, 1944, for a good refutation of the “growth argument.”

²²Cf. Harris, in Harris, op. cit., pp. 172 ff.; also Alvin H. Hansen and Guy Greer, “The Federal Debt and the Future,” Harper’s Magazine, 1942: “The internal debt of a government need never be paid” (p. 492).

²³Cf. Joseph A. Schumpeter, review of Harold J. Laski’s Reflections on the Revolution of Our Time, in American Economic Review, March 1944, p. 163.

²⁴ A. C. Pigou demonstrates in “The Classical Stationary State,” Economic Journal, 1943, pp. 343-51, how investments which no longer bore interest became profitable again when the value of money increased, after workers had been forced to accept lower wages (pp. 349-50). He comes to the conclusion: “I have been concerned to show that in given conditions of technique and so on, if wage earners follow a competitive wage policy, the economic system must move ultimately to a full employment stationary state, which is the essential thesis of the classicals. There can be no question at all that in this event the equilibrium that is attained is stable” (p. 351).

²⁵Cf. Sumner H. Slichter, in Harris, op. cit.: “The fears which encourage the hoarding of cash may be partly fears of higher taxes, i.e., fears aroused by the deficit itself” (p. 250).

²⁶ There exists, in addition to the incorrectness of his factual assumptions, a methodological reason why Keynes’ theory cannot be considered a satisfactory analysis of a stable equilibrium but only of frictional maladjustments: in an equilibrium analysis it is inadmissible to assume that some of the data, the prices of goods, yield to inflation whereas the others, the wages, interests, profits remain rigid. Either everything or nothing must be considered as flexible. In the first case the quantity theory is valid; in the latter case we have a sort of regulated economy in which not economic but price- and wage-fixing laws reign over the market.

²⁷ I consider the refusal of the Brüning government to follow a reflationary policy in 1931 the most important cause of the victory of the Nazi party.

²⁸ For the distinction between structural and cyclical unemployment, cf. L. Albert Hahn, 1st Arbeitslosigkeit unvermeidlich?, Berlin, 1930. The reader will find in this booklet a summary of the views on unemployment expressed in Europe during a discussion which strikingly resembles the one going on at the present time in this country.

²⁹ Accordingly, in the third edition of my Volkswirtschaftliche Theorie des Bankkredits, the Interest Theory of Unemployment was developed as a cyclical theory.