Keynesian Criticisms of the Theory
Keynesian Criticisms of the Theory
There are two standard Keynesian criticisms of the Mises cycle theory. One charge takes the followers of Mises to task for identifying saving and investment. Saving and investment, the Keynesians charge, are two entirely separate processes, performed by two sets of people with little or no link between them; the "classical" identification of saving and investment is therefore illegitimate. Savings "leak" out of the consumption-spending stream; investments pour in from some other phase of spending. The task of government in a depression, according to the Keynesians, is accordingly to stimulate investments and discourage savings, so that total spendings increase.
Savings and investment are indissolubly linked. It is impossible to encourage one and discourage the other. Aside from bank credit, investments can come from no other source than savings (and we have seen what happens when investments are financed by bank credit). Not only consumers save directly, but also consumers in their capacity as independent businessmen or as owners of corporations. But can't savings be "hoarded"? This, however, is an artificial and misleading way of putting the matter. Consider a man's possible allocation of his monetary assets:
He can (1) spend money on consumption; (2) spend on investment; (3) add to cash balance or subtract from previous cash balance. This is the sum of his alternatives. The Keynesians assume, most contrivedly, that he first decides how much to consume or not, calling this "not-consumption" saving, and then decides how much to invest and how much to "leak" into hoards. (This, of course, is neo-Keynesianism rather than pure Keynesian orthodoxy, which banishes hoarding from the living room, while readmitting it by the back door.) This is a highly artificial approach and confirms Sir Dennis Robertson's charge that the Keynesians are incapable of "visualizing more than two margins at once." Clearly, our individual decides at one and the same stroke about allocating his income in the three different channels. Furthermore, he allocates between the various categories on the basis of two embracing utilities: his time preferences decide his allocation between consumption and investment (between spending on present vs. future consumption); his utility of money decides how much he will keep in his cash balance. In order to invest resources in the future, he must restrict his consumption and save funds. This restricting is his savings, and so saving and investment are always equivalent. The two terms may be used almost interchangeably.
These various individual valuations sum up to social time-preference ratios and social demand for money. If people's demand for cash balances increases, we do not call this "savings leaking into hoards"; we simply say that demand for money has increased. In the aggregate, total cash balances can only rise to the extent that the total supply of money rises, since the two are identical. But real cash balances can increase through a rise in the value of the dollar. If the value of the dollar is permitted to rise (prices are permitted to fall) without hindrance, no dislocations will be caused by this increased demand, and depressions will not be aggravated. The Keynesian doctrine artificially assumes that any increase (or decrease) in hoards will be matched by a corresponding fall (or rise) in invested funds. But this is not correct. The demand for money is completely unrelated to the time-preference proportions people might adopt; increased hoarding, therefore, could just as easily come out of reduced consumption as out of reduced investment. In short, the savings-investment-consumption proportions are determined by time preferences of individuals; the spending-cash balance proportion is determined by their demands for money.
The Liquidity "Trap"
The ultimate weapon in the Keynesian arsenal of explanations of depressions is the "liquidity trap." This is not precisely a critique of the Mises theory, but it is the last line of Keynesian defense of their own inflationary "cures" for depression. Keynesians claim that "liquidity preference" (demand for money) may be so persistently high that the rate of interest could not fall low enough to stimulate investment sufficiently to raise the economy out of the depression. This statement assumes that the rate of interest is determined by "liquidity preference" instead of by time preference; and it also assumes again that the link between savings and investment is very tenuous indeed, only tentatively exerting itself through the rate of interest. But, on the contrary, it is not a question of saving and investment each being acted upon by the rate of interest; in fact, saving, investment, and the rate of interest are each and all simultaneously determined by individual time preferences on the market. Liquidity preference has nothing to do with this matter. Keynesians maintain that if the "speculative" demand for cash rises in a depression, this will raise the rate of interest. But this is not at all necessary. Increased hoarding can either come from funds formerly consumed, from funds formerly invested, or from a mixture of both that leaves the old consumption-investment proportion unchanged. Unless time preferences change, the last alternative will be the one adopted. Thus, the rate of interest depends solely on time preference, and not at all on "liquidity preference." In fact, if the increased hoards come mainly out of consumption, an increased demand for money will cause interest rates to fall—because time preferences have fallen.
In their stress on the liquidity trap as a potent factor in aggravating depression and perpetuating unemployment, the Keynesians make much fuss over the alleged fact that people, in a financial crisis, expect a rise in the rate of interest, and will therefore hoard money instead of purchasing bonds and contributing toward lower rates. It is this "speculative hoard" that constitutes the "liquidity trap," and is supposed to indicate the relation between liquidity preference and the interest rate. But the Keynesians are here misled by their superficial treatment of the interest rate as simply the price of loan contracts. The crucial interest rate, as we have indicated, is the natural rate—the "profit spread" on the market. Since loans are simply a form of investment, the rate on loans is but a pale reflection of the natural rate. What, then, does an expectation of rising interest rates really mean? It means that people expect increases in the rate of net return on the market, via wages and other producers' goods prices falling faster than do consumer goods' prices. But this needs no labyrinthine explanation; investors expect falling wages and other factor prices, and they are therefore holding off investing in factors until the fall occurs. But this is old-fashioned "classical" speculation on price changes. This expectation, far from being an upsetting element, actually speeds up the adjustment. Just as all speculation speeds up adjustment to the proper levels, so this expectation hastens the fall in wages and other factor prices, hastening the recovery, and permitting normal prosperity to return that much faster. Far from "speculative" hoarding being a bogy of depression, therefore, it is actually a welcome stimulant to more rapid recovery.
Such intelligent neo-Keynesians as Modigliani concede that only an "infinite" liquidity preference (an unlimited demand for money) will block return to full-employment equilibrium in a free market. But, as we have seen, heavy speculative demand for money speeds the adjustment process. Moreover, the demand for money could never be infinite because people must always continue consuming, on some level, regardless of their expectations. Since people must continue consuming, they must also continue producing, so that there can be adjustment and full employment regardless of the degree of hoarding. The failure to juxtapose hoarding and consuming again stems from the Keynesian neglect of more than two margins at once and their erroneous belief that hoarding only reduces investment, not consumption.
In a brilliant article on Keynesianism and price-wage flexibility, Professor Hutt points out that:
No condition which even distinctly resembles infinite elasticity of demand for money assets has even been recognized, I believe, because general expectations have always envisaged either (a) the attainment in the not too distant future of some definite scale of prices, or (b) so gradual a decline of prices that no cumulative postponement of expenditure has seemed profitable.
But even if such an unlikely demand arose:
If one can seriously imagine [this situation] . . . with the aggregate real value of money assets being inflated, and prices being driven down catastrophically, then one may equally legitimately (and equally extravagantly) imagine continuous price coordination accompanying the emergence of such a position. We can conceive, that is, of prices falling rapidly, keeping pace with expectations of price changes, but never reaching zero, with full utilization of resources persisting all the way.
Wage Rates and Unemployment
Sophisticated Keynesians now admit that the Keynesian theory of "underemployment equilibrium" does not really apply (as was first believed) to the free and unhampered market: that it assumes, in fact, that wage rates are rigid downward. "Classical" economists have always maintained that unemployment is caused precisely by wage rates not being allowed to fall freely; but in the Keynesian system this assumption has been buried in a mass of irrelevant equations. The assumption is there, nevertheless, and it is crucial. The Keynesian prescription for unemployment rests on the persistence of a "money illusion" among workers, i.e., on the belief that while, through unions and government, they will keep money wage rates from falling, they will also accept a fall in real wage rates via higher prices. Governmental inflation, then, is supposed to eliminate unemployment by bringing about such a fall in real wage rates. In these times of ardent concentration on the cost-of-living index, such duplicity is impossible and we need not repeat here the various undesirable consequences of inflation.
It is curious that even economists who subscribe to a general theory of prices balk whenever the theory is logically applied to wages, the prices of labor services. Marginal productivity theory, for example, may be applied strictly to other factors; but, when wages are discussed, we suddenly read about "zones of indeterminacy" and "bargaining." Similarly, most economists would readily admit that keeping the price of any good above the amount that would clear the market will cause unsold surpluses to pile up. Yet, they are reluctant to admit this in the case of labor. If they claim that "labor" is a general good, and therefore that wage cuts will injure general purchasing power, it must first be replied that "general labor" is not sold on the market; that it is certain specific labor that is usually kept artificially high and that this labor will be unemployed. It is true, however, that the wider the extent of the artificially high wage rates, the more likely will mass unemployment be. If, for example, only a few crafts manage, by union or government coercion to boost the wage rate in their fields above the free-market rate, displaced workers will move into a poorer line of work, and find employment there. In that case, the remaining union workers have gained their wage increase at the expense of lower wage rates elsewhere and of a general misallocation of productive factors. The wider the extent of the rigid wages, however, the less opportunity there will be to move and the greater will be the extent and duration of the unemployment.
In a free market, wage rates will tend to adjust themselves so that there is no involuntary unemployment, i.e., so that all those desiring to work can find jobs. Generally, wage rates can only be kept above full-employment rates through coercion by government, unions, or both. Occasionally, however, the high wage rates are maintained by voluntary choice (although the choice is usually ignorant of the consequences) or by coercion supplemented by voluntary choice. It may happen, for example, that either business firms or the workers themselves may become persuaded that maintaining wage rates artificially high is their bounden duty. Such persuasion has actually been at the root of much of the unemployment of our time, and this was particularly true in the 1929 depression. Workers, for example, become persuaded of the great importance of preserving the mystique of the union: of union solidarity in "not crossing a picket line," or not undercutting union wage rates. Unions almost always reinforce this mystique with violence, but there is no gainsaying the breadth of its influence. To the extent that workers, both in and out of the union, feel bound by this mystique, to that extent will they refuse to bid wages downward even when they are unemployed. If they do that, then we must conclude that they are unemployed voluntarily, and that the way to end their unemployment is to convince them that the mystique of the union is morally absurd. However, while these workers are unemployed voluntarily, as a consequence of their devotion to the union, it is highly likely that the workers do not fully realize the consequences of their ideas and actions. The mass of men are generally ignorant of economic truths. It is highly possible that once they discovered that their unemployment was the direct result of their devotion to union solidarity, much of this devotion would quickly wither away.
Both workers and businessmen may become persuaded by the mistaken idea that artificial propping of wage rates is beneficial. This factor played a great role in the 1929 depression. As early as the 1920s, "big" businessmen were swayed by "enlightened" and "progressive" ideas, one of which mistakenly held that American prosperity was caused by the payment of high wages (rates?) instead of the other way round. As if other countries had a lower standard of living because their businessmen stupidly refused to quadruple or quintuple their wage rates! By the time of the depression, then, businessmen were ripe for believing that lowering wage rates would cut "purchasing power" (consumption) and worsen the depression (a doctrine that the Keynesians later appropriated and embellished). To the extent that businessmen become convinced of this economic error, they are responsible for unemployment, but responsible, be it noted, not because they are acting "selfishly" and "greedily" but precisely because they are trying to act "responsibly." Insofar as government reinforces this conviction with cajolery and threat, the government bears the primary guilt for unemployment.
What of the Keynesian argument, however, that a fall in wage rates would not help cure unemployment because it would slash purchasing power and therefore deprive industry of needed demand for its products? This argument can be answered on many levels. In the first place, as prices fall in a depression, real wage rates are not only maintained but increased. If this helps employment by raising purchasing power, why not advocate drastic increases in money wage rates? Suppose the government decreed, for example, a minimum wage law where the minimum was triple the going wage rates? What would happen? Why don't the Keynesians advocate such a measure?
It is clear that the effect of such a decree would be total mass unemployment and a complete stoppage of the wheels of production. Unless . . . unless the money supply were increased to permit employers to pay such sums, but in that case real wage rates have not increased at all! Neither would it be an adequate reply to say that this measure would "go too far" because wage rates are both costs to entrepreneurs and incomes to workers. The point is that the free-market rate is precisely the one that adjusts wages—costs and incomes—to the full-employment position. Any other wage rate distorts the economic situation.
The Keynesian argument confuses wage rates with wage incomes—a common failing of the economic literature, which often talks vaguely of "wages" without specifying rate or income. Actually, wage income equals wage rate multiplied by the amount of time over which the income is earned. If the wage rate is per hour, for example, wage rate will equal total wage income divided by the total number of hours worked. But then the total wage income depends on the number of hours worked as well as on the wage rate. We are contending here that a drop in the wage rate will lead to an increase in the total number employed; if the total man-hours worked increases enough, it can also lead to an increase in the total wage bill, or payrolls. A fall in wage rate, then, does not necessarily lead to a fall in total wage incomes; in fact, it may do the opposite. At the very least, however, it will lead to an absorption of the unemployed, and this is the issue under discussion. As an illustration, suppose that we simplify matters (but not too drastically) and assume a fixed "wages fund" which employers can dispense to workers. Clearly, then, a reduced wage rate will permit the same payroll fund to be spread over a greater number of people. There is no reason to assume that total payroll will fall.
In actuality, there is no fixed fund for wages, but there is rather a
fixed "capital fund" which business pays out to all factors of production.
Ultimately, there is no return to capital goods, since their prices are all
absorbed by wages and land rents (and interest, which, as the price of time,
permeates the economy). Therefore, what business as a whole has at any time
is a fixed fund for wages, rents, and interest. Labor and land are perennial
competitors. Since production functions are not fixed throughout the economy,
a widespread reduction in wage rates would cause business to substitute
labor for land, labor now being relatively more attractive
vis-à-vis land than it was before.
Consequently, aggregate payrolls would not be the same; they
would increase, because of the substitution effect in favor of
labor as against land. The aggregate demand for labor would therefore be
Suppose, however, that the highly improbable "worst" occurs, and the demand for labor turns out to be inelastic, i.e., total payrolls decline as a result of a cut in wage rates. What then? First, such inelasticity could only be due to businesses holding off from investing in labor in expectation that wage rates will fall further. But the way to meet such speculation is to permit wage rates to fall as quickly and rapidly as possible. A quick fall to the free-market rate will demonstrate to businessmen that wage rates have fallen their maximum viable amount. Not only will this not lead businesses to wait further before investing in labor, it will stimulate businesses to hurry and invest before wage rates rise again. The popular tendency to regard speculation as a commanding force in its own right must be avoided; the more astute as forecasters and diviners of the economy the businessmen are, the more they will "speculate," and the more will their speculation spur rather than delay the natural equilibrating forces of the market. For any mistakes in speculation-selling or buying goods or services too fast or too soon-will directly injure the businessmen themselves. Speculation is not self-perpetuating; it depends wholly and ultimately on the underlying forces of natural supply and consumer demand, and it promotes adjustment to those forces. If businessmen overspeculate in inventory of a certain good, for example, the piling up of unsold stock will lead to losses and speedy correction. Similarly, if businessmen wait too long to purchase labor, labor "shortages" will develop and businessmen will quickly bid up wage rates to their "true" free-market rates. Entrepreneurs, we remember, are trained to forecast the market correctly; they only make mass errors when governmental or bank intervention distorts the "signals" of the market and misleads them on the true state of underlying supply and demand. There is no interventionary deception here; on the contrary, we are discussing a return to the free market after a previous intervention has been eliminated.
If a quick fall in wage rates ends and even reverses withholding of the purchase of labor, a slow, sluggish, downward drift of wage rates will aggravate matters, because (a) it will perpetuate wages above free market levels and therefore perpetuate unemployment; and (b) it will stimulate withholding of labor purchases, thereby tending to aggravate the unemployment problem even further.
Second, whether or not such speculation takes place, there is still no reason why unemployment cannot be speedily eliminated. If workers do not hold out for a reserve price because of union pressure or persuasion, unemployment will disappear even if total payroll has declined.
The following diagram will illustrate this process: (see Figure 1). Quantity of Labor is on the horizontal axis; wage rate on the vertical. DlDl is the aggregate demand for Labor; IE is the total stock of labor in the society; that is, the total supply of labor seeking work. The supply of labor is represented by vertical line SlSl rather than by the usual forward-sloping supply curve, because we may abstain from any cutting of hours due to falling wage rates, and more important, because we are investigating the problem of involuntary unemployment rather than voluntary. Those who wish to cut back their hours, or quit working altogether when wage rates fall, can hardly be considered as posing an "unemployment problem" to society, and we can therefore omit them here.
In a free market, the wage rate will be set by the intersection of the labor supply curve SlSl and the demand curve DlDl, or at point E or wage rate 0I. The labor stock IE will be fully employed. Suppose, however, that because of coercion or persuasion, the wage rate is kept rigid so that it does not fall below 0A. The supply of labor curve is now changed: it is now horizontal over AC, then rises vertically upward, CSl. Instead of intersecting the demand for labor at point E the new supply of labor curve intersects it at point B. This equilibrium point now sets the minimum wage rate of 0A, but only employs AB workers, leaving BC unemployed. Clearly, the remedy for the unemployment is to remove the artificial prop keeping the supply of labor curve at AC, and to permit wage rates to fall until full-employment equilibrium is reached.
Now, the critic might ask: suppose there is not only speculation that will speed adjustment, but speculation that overshoots its mark. The "speculative demand for labor" can then be considered to be DsDs, purchasing less labor at every wage rate than the "true" demand curve requires. What happens? Not unemployment, but full employment at a lower wage rate, 0J. Now, as the wage rate falls below underlying market levels, the true demand for labor becomes ever greater than the supply of labor; at the new "equilibrium" wage the gap is equal to GH. The enormous pressure of this true demand leads entrepreneurs to see the gap, and they begin to bid up wage rates to overcome the resulting "shortage of labor." Speculation is self correcting rather than self aggravating, and wages are bid up to the underlying free-market wage 0I.
If speculation presents no problems whatever and even helps matters when wage rates are permitted to fall freely, it accentuates the evils of unemployment as long as wages are maintained above free-market levels. Keeping wage rates up or only permitting them to fall sluggishly and reluctantly in a depression sets up among businessmen the expectation that wage rates must eventually be allowed to fall. Such speculation lowers the aggregate demand curve for labor, say to DsDs. But with the supply curve of labor still maintained horizontally at AC, the equilibrium wage rate is pushed farther to the left at F. and the amount employed reduced to AF, the amount unemployed increased to FC.
Thus, even if total payrolls decline, freely falling wage rates will always bring about a speedy end to involuntary unemployment. The Keynesian linkage of total employment with total monetary demand for products implicitly assumes rigid wage rates downward; it therefore cannot be used to criticize the policy of freely-falling wage rates. But even if full employment is maintained, will not the declining demand further depress business? There are two answers to this. In the first place, what has happened to the existing money supply? We are assuming throughout a given quantity of money existing in the society. This money has not disappeared. Neither, for that matter, has total monetary spending necessarily declined. If total payrolls have declined, something else has gone up: the total retained by entrepreneurs, or by investors, for example. In fact, given the total money supply, the total flow of monetary spending will only decline if the social demand for money has increased. In other words, if "hoarding" has increased. But an increase in hoarding, in total demand for money, is, as we have seen, no social calamity. In response to the needs and uncertainties of depression, people desire to increase their real cash balances, and they can only do so, with a given amount of total cash, by lowering prices. Hoarding, therefore, lowers prices all around, but need exert no depressing effect whatever upon business. Business, as we have pointed out, depends for its profitability on price differentials between factor and selling prices, not upon general price levels. Decrease or increase in total monetary spending is, therefore, irrelevant to the general profitability of business.
Finally, there is the Keynesian argument that wage earners consume a greater proportion of their income than landlords or entrepreneurs, and therefore that a decreased total wage bill is a calamity because consumption will decline and savings increase. In the first place, this is not always accurate. It assumes (1) that the laborers are the relatively "poor" and the nonlaborers the relative "rich," and (2) that the poor consume a greater proportion of their income than the rich. The first assumption is not necessarily correct. The President of General Motors is, after all, a "laborer," and so also is Mickey Mantle; on the other hand, there are a great many poor landlords, farmers, and retailers. Manipulating relations between wage earners and others is a very clumsy and ineffective way of manipulating relations between poor and rich (provided we desire any manipulation at all). The second assumption is often, but not necessarily, true, as we have seen above. As we have also seen, however, the empirical study of Lubell indicates that a redistribution of income between rich and poor may not appreciably affect the social consumption-saving proportions. But suppose that all these objections are waved aside for the moment, and we concede for the sake of argument that a fall in total payroll will shift the social proportion against consumption and in favor of saving. What then? But this is precisely an effect that we should highly prize. For, as we have seen, any shift in social time preferences in favor of saving and against consumption will speed the advent of recovery, and decrease the need for a lengthy period of depression readjustment. Any such shift from consumption to savings will foster recovery. To the extent that this dreaded fall in consumption does result from a cut in wage rates, then, the depression will be cured that much more rapidly.
A final note: The surplus "quantity of labor" caused by artificially high wage rates is a surplus quantity of hours worked. This can mean (1) actual unemployment of workers, and/or (2) reduction in working time for employed workers. If a certain number of labor hours are surplus, workers can be discharged outright, or many more can find their weekly working time reduced and their payroll reduced accordingly. The latter scheme is often advanced during a depression, and is called "spreading the work." Actually, it simply spreads the unemployment. Instead of most workers being fully employed and others unemployed, all become under-employed. Universal adoption of this proposal would render artificial wage maintenance absurd, because no one would be really benefitting from the high wage rates. Of what use are continuing high hourly wage rates if weekly wage rates are lower? The hour-reduction scheme, moreover, perpetuates underemployment. A mass of totally unemployed is liable to press severely on artificial wage rates, and out-compete the employed workers. Securing a greater mass of under-employed prevents such pressure-and this, indeed, is one of the main reasons that unions favor the scheme. In many cases, of course, the plea for shorter hours is accompanied by a call for higher hourly wage rates to "keep weekly take-home pay the same"; this of course is a blatant demand for higher real wage rates, accompanied by reduced production and further unemployment as well.
Reduction of hours to "share the work" will also reduce everyone's real wage rate and the general standard of living, for production will not only be lower but undoubtedly far less efficient, and workers all less productive. This will further widen the gap between the artificially maintained wage rate and the free-market wage rate, and hence further aggravate the unemployment problem.