The Fallacies of the NRA
This year marks the 70th Anniversary of the National Industrial Recovery Act, FDR's planning legislation that created the National Recovery Administration, the NRA. Henry Hazlitt wrote a critique of the NRA for The Nation that resulted in his losing his position as literary editor. This piece for the American Mercury (December 1933, V.30.120, pp. 415–23; Hazlitt succeeded H.L. Mencken as editor) was his vengeance. He argues that the NRA cannot assist recovery but rather will only cartelize industry. Here is Hazlitt's complete bibliography.
"You expect me to tell you anything about America?" said Montague Norman to a British audience in September. "It is two weeks at least since I have been there, and in those two weeks a new situation, almost a new country, has grown up. So continuous and so great are the changes that no one who is not on the spot is up-to-date."
It would be presumptuous and unwise for lesser beings not to follow the example of the governor of the Bank of England. In the few weeks between the time that these words are written and the time the reader's eyes are cast on them another new situation, new policy and New Deal may be upon us. The present article, therefore, is not to be taken as a prediction. It is not even to be taken as a description of what is happening. It is intended primarily as a scrutiny of the assumptions on which the current New Deal—in particular, the N.R.A.—is based, to see whether those assumptions are sound or plausible.
Perhaps the most important of them is that the arbitrary cutting down of hours will force the employment of more men. Obviously, however, it can have this effect only in firms and industries that were previously employing men for a larger number of hours, on the average, than the new codes prescribe as a maximum. Most of these codes call for a forty-hour week. But last June, for example, the month before the first code went into effect, the average number of hours of work a week for each employed worker in industry, according to the figures of the National Industrial Conference Board, was only forty-one, and earlier in the year they had been below thirty. A Depression does more to reduce working hours even than a New Deal. The codes can provide employment for more men only in those industries that either were already working, or would otherwise now be working, more than forty hours a week. Among these is the cotton textile industry, where there has actually been an increase in employment from 356,000 persons in March to 466,000 in September.
If other conditions are such as to bring about a revival—and to some extent they have recently been so—then the N.R.A. programme tends to bring a more rapid reemployment than otherwise. But it must be pointed out that in so far as it does so it is essentially a Share-the-Work movement. The total amount of wages paid out is not necessarily made any greater through a mere reduction in the weekly hours of individual workers. Instead of five men getting paid for forty-eight hours a week each, for example, six men may be employed at forty hours a week each—with each of the original five receiving five-sixths as much as he otherwise might have. In brief, each of the original five men is contributing one-sixth of his possible income for the sake of the extra man. The burden of "creating employment" in this manner merely falls on labor itself.
This is what tends to happen as long as there is a slack in employment to be taken up. In prosperous times, however, when labor might otherwise be fully employed the cutting down of permissible hour from, say, forty-eight a week to forty, would mean that fewer goods were being produced than could be produced, and that everybody was poorer than he needed to be. It is true, of course, that the reduction in output would be somewhat less than the reduction in hours, for with fewer hours, and hence with less fatigue, it would be possible to speed up hourly production. Nevertheless, the fatigue of eight hours a day for six days is not in most occupations very intense; and no one can seriously argue that exactly as large a product could be produced in six six-hour or five eight-hour days.
It may be argued, of course, that there is still a gain in the less arduous toil and in the added leisure that fewer hours mean to labor. This is true; but the leisure is something that must be paid for. A worker who is paid 50 cents an hour would get $24 for a forty-eight-hour week and only $20 for a forty-hour week. With so low an income, $4 may mean a great deal; the worker might conceivably prefer to buy something else with it than eight free hours. He might prefer to buy only two more free hours, or at most four more, and use the other $2 or $3 for more decent meals. For it may be pointed out (if it is still legal to use any of the terms of orthodox economics) that each new hour of leisure the worker gets has a declining marginal significance to him, while each dollar of pay he loses has an increasing marginal importance. Under the N.R.A., however, his new leisure is obligatory.
Another of the leading assumptions behind the N.R.A. is that industry really has the funds to pay out in wages, but has wickedly refused to do so, and that all that is needed is a legal or political compulsion. This overlooks the most obvious facts of the whole Depression. A compilation of the returns of 1,800 of the leading industrial, public-utility, and railroad companies in 1932 showed that the majority, numbering 61% of the total, had deficits aggregating $787,000,000, which offset the net profits, aggregating $682,000,000 reported by the other 39%, leaving a deficit balance for all of them of $105,000,000. The Class I railroads taken alone showed a combined net deficit of approximately $153,000,000 for 1932. In the majority of companies, therefore, the funds did not exist out of which to pay higher wages; even the wages being paid were coming for the most part out of dwindling surpluses. If average wage rates had been forced up in 1932, before there had already been a rise in prices and some improvement in conditions from causes outside of the N.R.A., then the net result would have been to throw many more firms into bankruptcy, and to increase unemployment. In itself, the N.R.A. is an anti-Recovery policy.
One of the chief assumptions on which the N.R.A. rests is, of course, that if wages can somehow be forced up to a higher level there will be an increase of purchasing power, and hence a recovery in trade and production. This assumption involves two major confusions: (I) it confuses a mere shift in purchasing power with a net increase of it, and (2) it confuses wage rates with wage payments. Let us consider these more in detail.
An increase in wage rates should not be interpreted as representing necessarily a net increase in purchasing power, because in so far as it is taken at the expense of what was previously profits, the purchasing power of managers and investors falls off to the same extent as that of labor rises. The result in such a case is not the purchase of a greater total amount of goods but rather a shift in demand from some classes of goods to others—from a small volume of expensive commodities to a larger volume of inexpensive commodities. Such a shift might be highly desirable in itself during certain periods. From 1922 to 1919, for example, the aggregate command over goods exercised by manufacturing labor, as Frederick C. Mills has pointed out, increased at a rate of 3.1% a year while that of "ownership and management" increased by 7.3% a year. In this period of high profits too much of the social income was being turned back into reinvestment in new plant[s]. But a shift from profits to wages can accomplish very little toward revival in a period like the present, when profits are at an extremely low point and new capital investment (except that represented by public works) is close to zero.
Let us turn now to the distinction between wage payments and wage rates. In the last few years the former have fallen to an enormously greater extent than the latter. Thousands of factories report their total payrolls to the United States Bureau of Labor Statistics every month. These payrolls have shrunk during the Depression appallingly. At the low point of this year, in March, the amount paid out in wages was only 33% of that paid out in a similar period in 1926; by September this had risen to only 53% of the 1926 level. Hourly wage rates, on the other hand, have fallen on the average comparatively little during the Depression period. The National Industrial Conference Board has calculated that at the end of 1932 average hourly earnings had fallen 21.2%, compared with a decline in living costs of 22% in the same period.
Leo Wolman believes that the conference board's figures do not cover a sufficiently wide field, and estimates that the decline in wage rates was somewhere between 20 and 30% in that period. Wage rates in certain industries, and for non-union labor, fell of course much more than this; but this fall was offset by much smaller declines, or the absence of any reduction at all, elsewhere. There was no reduction from the peak whatever, for example, in the hourly wages of anthracite miners. In other words, the laborer who remained employed, and for the same number of hours that he had been employed before the Depression, enjoyed on the average very nearly the same purchasing power, in terms of goods, during the Depression as before it. In fact, the "real" weekly earnings, not only of anthracite miners, but of railroad workers, public utility employees, and employees in retail trade, actually increased during the Depression.
The purchasing power that really needs to be restored, therefore, is not that represented by the hourly wage rates of the men already on full time, but that of the millions of men on part time and of the ten million still out of work. To say this, however, is to say that the problem of "restoring purchasing power" is merely incidental to the broader problem of inquiring into the causes of the Depression and in the light of that inquiry developing an intelligent general programme of recovery. This broader problem is too large to be considered parenthetically here.
But one thing we are bound to do is to look at the problem of recovery as the individual manufacturer or business firm sees it. Since the World War we have witnessed the most violent collapse of wholesale commodity prices at least since the period just following the Napoleonic wars. That decline has amounted to about 40% in the last four years alone. Now, the problem of employers of labor, if they hope to remain in business, has been to get their unit costs of production down in something like the same ratio. The leaders of the N.R.A. have persistently ignored or forgotten the glaring truth that labor will not be employed unless there is a profit in employing it. The real problem is to return to a working relationship between prices and costs of production. The increase in wage rates involved by the N.R.A. tends not to solve this problem but to intensify it.
In brief, it attempts to deal with the problem of purchasing power from the wrong end. It assumes that if higher wage rates are paid out to labor, this added money will be spent for more goods and so lead to an increase in production and to higher prices. But we have already seen that even on the most favorable assumptions there may not be any net increase in purchasing power, because employers, managers, and investors will lose the purchasing power that labor gains. It is now necessary to admit also that the immediate increase in costs of production involved by the N.R.A. may do immediate damage by obliging employers to lay off more men and so actually reducing purchasing power.
The immemorial assumption behind all popular legislation is that a law or a political programme is to be judged by its intent rather than by its probable results. No matter how often this assumption is exploded—as it finally was so violently by the effects of the Eighteenth Amendment—it returns to the mass mind with undiminished force. This accounts for the popularity of the minimum wage provisions of the N.R.A. It is naively assumed that all the workers who have previously been getting less than the minimum will have their pay raised by the designated amount and will continue to be employed. The plain truth is, of course, that if no one is to be employed in industry at less than $11 to $15 a week, then no one who is not deemed worth that amount will be employed in industry at all.
Obviously, we cannot make a man worth $14 a week by saying that he shall not be offered and shall not take any less. We may be taking away from him even such miserable earnings as he could make, and we may be taking away from the community whatever product his labor could produce. What do we offer him in return for terminating his earnings and his labor? The brains in charge of the N.R.A. do not appear to have thought that far. Must we offer a man a choice between being worth $14 a week and starving? If we throw him out of work don't we owe him a permanent dole in return? And must not this dole be at least greater than the amount he previously earned—since we have already pronounced this to be below the level of decency? Must not the dole, in fact, logically be at or just below the minimum set? And if it is, what about the terrific temptation we create for malingering?
Even if we overlook this temptation, we should at least attempt some estimate of the probable number of persons the minimum wages are going to throw out of work, and of the enormous drain on the Treasury the payment of doles to them would cause. Up to the moment of writing this I have seen no evidence that anyone in Washington has made such an estimate. Yet average actual weekly earnings in all manufacturing industries in June, according to the figures of the National Industrial Conference Board, were $18.49; in March they were only $14.56. These low average weekly earnings are to some extent, of course, the result of part-time work; but they are enough to show that the code minimums, when compared with actual prevailing wages, are quite high.
The probable answer of some supporters of the N.R.A. to all this may be that wages are not primarily determined by productivity but by bargaining power, and that the new compulsion on employers to pay a minimum wage will offset the worker's disadvantage and take the place of this bargaining power. All one can say to this is that, while there is an element of truth in it, differences in bargaining power can on the average affect wages only within certain relatively narrow limits. Thus, in spite of the fact that Great Britain for years has been much more highly unionized than the United States, the wages even of non-union labor in America have been much higher than the wages for corresponding union labor in Great Britain.
It is true, of course, that in numerous cases the new minimum wage requirements actually will result in the raising of wages of persons formerly employed at a lower wage, and that the difference will be absorbed out of profits; but that is no reason why we can complacently ignore the thousands of cases in which this obviously will not occur. And it is fair to ask those defenders of the N.R.A. who do not believe in the possibility of any of the consequences I have indicated, just why they are so niggardly and stony-hearted. If minimum wages can be imposed with good consequences for all and ill consequences for none, why stop at $11 to $15? Why not a minimum wage of $25 a week? or $50? or $100? Obviously, the higher a minimum wage is placed the greater its unworkability and dangers become. This is not to say that there may not be a point at which the potential gains of a carefully considered minimum wage might outweigh its dangers. But I have so far seen no argument anywhere which attempted to show that the new minimums fixed were not above that point. So far as the members of the Brain Trust are concerned, no realistic discussion of the consequences has ever passed the lips of anyone of them.
One contention we have heard is that the new minimums will practically abolish child labor in industry. This is true, and it is a very great gain. But there has been no attempt to consider secondary consequences even here. It is apparently assumed that all the children laid off will be hereafter well taken care of by their parents and sent to a finishing-school; at all events, no painful sense of responsibility seems to be felt towards them. It is also generally assumed that their places will be taken by an equal number of adults. This is plainly improbable. Where children have been paid, say, $5 a week, the men or women who take their places will be paid $11 to $15 a week. If unit costs of production are not raised it will only be because the adult workers on the average will turn out about three times as many goods each as the children did. But if they do so, then only one-third as many adults will be needed as there were children.
In other cases the new minimums will lead to the discharge of women to put on men, or the laying off of older men, of the feeble and the maimed, to make way for the young and the able-bodied. The substitute workers will be smaller in number to precisely the extent that they are more efficient. In some industries it may not be possible, owing to the conditions within them or the nature of the work, to increase substantially the production of individual workers by getting other labor. In these industries it is probable that there will simply be an increase in the pay of the workers already employed. But this would mean a corresponding increase in costs of production, and in the long run a corresponding increase in the ultimate price of the product.
This might be brought about by forcing the marginal producers out of business, and so creating more unemployment. Or the higher price might lead to a reduction in the volume of the product sold, which would also cause more unemployment. It may be replied that cheap goods which are the product of sweated labor are not worth having, and that would be hard to deny. But it must not be forgotten, either, that when the price of such goods is raised, part of the people who have to pay the increased price are the workers whose pay has been raised; and to that extent their increased pay is illusory.
One of the great advantages of the N.R.A. programme, it is claimed, is that it has enormously accelerated collective bargaining and increased unionization. Now, theoretically, an increase in labor's bargaining power—particularly in the trades that have hitherto not been unionized—is in the highest degree desirable. Yet it must be recognized that in practice the unions have seldom exercised this power intelligently. A skilled general knows both when to advance and when to retreat, and by just about how much. American labor leadership has so far shown few signs of possessing this knowledge.
It could and should have got a good deal more for its members in the 1923–1929 period; and the stronger unions should have accepted more substantial reductions than they did, and accepted them more promptly, once the Depression had got under way. When I say this I mean that they should have done so merely in the interests of a majority of their own membership, for the sake of securing as full a volume of continuous employment as possible. But in general it is the policy of unions to "let well enough alone" in periods of prosperity, when they should be making much more rapid gains, and to try to sit tight during depressions, ignoring—even after the greatest price collapse of a century—the clear need for re-adjustment in costs if industrial activity and a high volume of employment are to be maintained.
The policy of the British trade unions is illuminating in this respect. When the pound returned to its old gold parity in 1925, after having been down to a discount of 34% and after prices and wages had to some extent adjusted themselves to that discount, prices fell again in reflection of the higher value of the pound; but the British labor unions refused to accept a reduction in wages. Such a reduction would have been necessary only in monetary terms and not in terms of goods; it need have gone only as far—at most—as the decline in prices and living costs. But highly unionized British labor refused to budge. Its resistance was made possible not merely by its unionization but by the dole to the unemployed.
The result was relative stagnation in British business, and a heavy and continuing volume of unemployment, all through the period in which the United States was at the height of its New Era. When the more complete unionization of American labor which is now going forward so rapidly has been achieved, we may expect a similar rigidity in wage rates in the United States, and the same difficulty in making adjustments to changed conditions. This difficulty will result in prolonging and intensifying future depressions.
In view of all these potentialities in the new programme, not to speak of many of the effects which are already apparent, it may be wondered why so many business men are so warmly supporting it. Some of them, no doubt, are sincerely patriotic, and sincerely happy to think that the lowest class of labor will be getting more pay; but a more substantial cause of the general rejoicing may be found in that part of the codes governing "fair competition." It is obvious that under the guise of eliminating "unfair practices" and "stabilizing the industry", most of these codes involve a movement in the direction of the cartelization of American business—in other words, in the direction of monopoly.
Most of these agreements to "stabilize" prices involve in effect a raising of prices. It is low prices that are always frowned upon in them, not high prices; prices below a "fair" level, not prices above it. In some codes this may seem merely amusing, as when the corset and brassiere industry, while permitting manufacturers or wholesalers to contribute up to 50% of the net cost of a retailer's advertising space, prohibits them from paying any of the cost of advertising on "corsets, combinations, girdle-corsets, or step-in-corsets which are advertised for retail sale at less than $2, or on brassieres which are advertised for retail sale at less than $1."
The case becomes much more serious, however, when it involves price fixing in a basic industry, for example, that of lumber. Under the code in this industry an agency is set up known as the Lumber Code Authority, Inc., to administer the agreement and to undertake the task of controlling (i.e., restricting) production, and the task of "cost protection." The latter, of course, simply means the fixing of prices. The Authority, it is to be observed, is not permitted to fix maximum but only minimum prices. It must not allow such prices to fall below "the cost of production." Now, anyone who has even a nodding acquaintance with business or manufacturing knows that cost of production is a very elusive thing to determine. It is never the same in two firms in the same business; and even in the same firm no two accountants will agree regarding the actual unit costs of production in that firm. But the lumber code takes no chances. The cost of production below which lumber prices will not be permitted to fall is the "current weighted average cost of production", and in adding up this cost the following items are to be included:
(1) wages; (2) materials and supplies; (3) overhead and administration, including trade association dues and code fees; (4) shipping, including grading and loading; (5) selling…; (6) maintenance; (7) insurance, including compensation and employe[e] insurance…; (8) taxes, including taxes on timber…; (9) interest paid on Indebtedness [representing not only plant[s] actually operating but "capable of operating"]; (10) discounts, claims paid, and losses on trade accounts; (11) depreciation…; (12) raw material [including "standing timber carried in capital account"] and (13) conservation and re-forestation, including costs of protection of timbered and cut-over lands….
When all of these items are thrown in, the lumber industry should be able to present a very impressive figure for cost of production. In other words, it can fix a very substantial minimum price. And as this minimum price is to represent the average cost of production in the industry, very liberally calculated, it is obvious that it will also provide a substantial profit for at least half the units in the industry. The prospective home-owner whose house is going to cost him considerably more than heretofore will be glad to notice that foreign buyers will not have to pay these high prices for American lumber: export sales are explicitly excluded from the minimum-price provisions. In other words, where American lumber interests have to meet foreign competition, they will consent to the indignity and the hardship and horrible injustice of selling below cost of production. Anyway, they can take it out of the American consumer. Like all the other codes, the lumber code contains a pious provision to the effect that "this code shall not be construed, interpreted or applied so as to promote or permit monopolies or monopolistic practices."
It is obvious that under the present N.R.A. programme the American consumer is to become the victim of a series of trades and industries which, in the name of "fair competition", will be in effect monopolies, consisting of units that agree not to make too serious an effort to undersell each other; restricting production, fixing prices,—doing everything, in fact, that monopolies are formed to do. The American consumer will be paying higher prices not only because of higher wage rates but because of these innumerable trade agreements. The increasing rigidity which greater unionization will impose on wages will be supplemented by perhaps an even greater rigidity which these agreements will fasten on wholesale and retail prices. Instead of a relatively flexible system with some power of adjustment to fluid world economic conditions we shall have an inadjustable structure constantly attempting—at the cost of stagnant business and unemployment—to resist these conditions.
I must again make it plain that in this article I am not attempting to predict anything. I am merely trying to examine what the effects of the N.R.A. programme would be if it were the sole important force acting on conditions, and if its provisions were not evaded but strictly complied with. But obviously the actual economic future will be determined by numerous factors. A heavy dose of currency inflation alone, for example, would offset the higher wage scales by raising prices to a point where those wage scales became irrelevant.
Conditions in the United States will be influenced heavily by conditions in other countries; since July, 1932, there has actually been—in spite of our bank crisis in March—a rather steady improvement in world conditions; and the forces that have brought this about may continue to operate. The results of the new programme at the moment of writing are thus ambiguous. Secretary Frances Perkins has estimated that at the beginning of October nearly 2,750,000 more persons were employed than in March. Part of this new employment occurred, of course, before the N.R.A. went into effect; and it is impossible to know just how much of it represents mere "work sharing".
On the other hand, while the New York Times weekly business index, compiled from statistics of freight car loadings, steel-mill activity, electric-power, automobile, and lumber production, and cotton forwardings, rose from 60% of a hypothetical normal in March to nearly 100% in mid-July, it is perhaps significant that it was just at this time, when the first code went into effect, that the index began plunging downward again, by mid-October canceling more than half of the early gain. But what has happened since July cannot be attributed to the N.R.A. alone; the economic future will be determined by a complex of forces of which even that programme will be only one.
Against the good elements in that programme—the virtual eradication of child labor in industry, the help to some underpaid labor, the possible elimination of competitive waste in some special industries where such action is desperately needed, as in the oil fields—the offsets are heavy ones. It cannot by itself give new employment to anything like the extent that its supporters imagine. To the extent that it does so it is a Share-the-work movement, with the chief burden on labor itself. In times like the present, it will result in the substitution of more efficient workers for less efficient ones, but in times that might otherwise be prosperous it would reduce the possible production of goods and the possible demand for goods.
It makes leisure compulsory at too high a cost. It has never really investigated the ability of industry to pay the new wages. It confuses a shift in purchasing power with a net increase of it, wage rates with wage payments. It offers a man a choice between being worth $11 to $15 a week or starving. It accepts no responsibility for those it throws out of work, nor is it even interested in learning how many persons that may be. It encourages a unionization that will probably make adjustment to changing conditions more difficult, and it creates a series of cartels under the aegis of government that will soon have the consumer completely at their mercy.
The anonymous consumer was usually the person that William Graham Sumner was referring to when he mentioned the Forgotten Man. That the President who revived this phrase should be the one responsible for the N.R.A. is an ironic consummation at which old Sumner's ghost must be smiling wanly.
See the biography of Henry Hazlitt.