The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
ECONOMIC ANALYSIS OF GUARANTEED WAGES Bulletin N o. 9 0 7 UNITED STATES DEPARTMENT OF LA BO R B U R E A U O F L A B O R S T A T IS T IC S LETTER OF TRANSM ITTAL U n it e d S tates D B epar tm en t u r e a u of L of L abo r a b o r S , t a t is t ic s , Washington, D. C., June 9,1947 T h e S e c r etar y of L a b o r : I have the honor to transmit herewith an economic analysis of guaranteed wages, which was prepared by Professors Alvin H. Hansen and Paul A. Samuelson, who were commissioned by the Bureau of Labor Statistics to make this analysis as part of its collaboration with the Guaranteed Wage Study of the Advisory Board of the Office of War Mobilization and Reconversion. E w a n C l a g u e , Commissioner. Hon. L. B. S C H W E L L E N B A C H , Secretary oj Labor. ECONOMIC ANALYSIS O F GUARANTEED W AGES Bulletin N o. 9 0 7 A reprint of Appendix F from Guaranteed Wages: Report to the President by the Advisory Board, Murray W . Latimer, Research Director PREFACE Upon the initiative of the Advisory Committee of the Office of War Mobilization and Reconversion in the study of guaranteed wage plans, the Bureau of Labor Statistics commissioned Professors Alvin H. Hansen and Paul A. Samuelson, of Harvard Uni versity, to prepare an economic analysis of guaranteed wages. This report was designed to supplement the description of guaranteed wage plans (Appendix C of the OW M R report and also reprinted separately as a Bureau bulletin), prepared by the Bureau for the Office of War Mobilization and Reconversion, and the exploration of the practical possibilities and problems in the extension of guaranteed wages, prepared by the Guar anteed Wage Study staff of the Office of War Mobilization and Reconversion. It appears in the Final Report of the Advisory Board of the Office of War M obilization and Reconversion as Appendix F. The present report is an economic analysis of the potential effects of guaranteed wage plans on the economy and the relation of guaranteed wages, if widely adopted, to economic security, business cycles, and the pattern of resource uses. In accordance with arrangements described above, the Bureau of Labor Statistics also commissioned Professors J. M . Clark, of Columbia University, and Edward S. Mason and Sumner H. Slichter, of Harvard University, to comment on the analysis by Professors Hansen and Samuelson. These comments are included in the report. The views expressed in all cases are those of the authors. i CONTENTS A ppendix F.— E conomic A nalysis of G uaranteed W ages Page I. Introduction______________________________________________________________________ Brief history of guaranteed wage proposals_______________________________________ Origin of present study__________________________________________________________ Description of guaranteed wage plans------------------------------------------------------------------Some preliminary observations----------------------------------------------------------------------------Proposed outline of analysis--------------------------------------------------------------------------------II. Analysis of the econom ic burden of idleness------------------------------------- --------------------Social versus private costs of idleness------------------------------------------- ----------- ----------The econom ic costs of idleness------------------------------ ------------------------------- --------------The cost of instability to the worker_____________________________________________ The cost of instability to industry------------------- ----------- ---------------------------------------III. The guaranteed wage in a broad security program-----------------------------------------------Different patterns o f job security________________________________________________ Guaranteed wages and stabilized employm ent....................................... ......................__ A limited guaranteed wage----------------------------------------------------------------------------------Three roads to job and income security__________________________________________ IV . Guaranteed wages, over-all effective demand, and the business cycle______________ Introduction and brief summary-------------------------------------------------------------------------Jobs and markets________________________________________________________________ W ill guaranteed wages stabilize consumption?------------------------------------------------------Business cycle stabilization and production efficiencies resulting from guaranteed wages_________________________________________________________________________ Tw o methods of finance--------------------------------------------------------------------------------------Pay-as-you-go finance-----------------------------------------------------------------------------------------Reserve financing________________________________________________________________ V. Guaranteed wages and efficient use of resources----------------------------------------------------The merit-rating controversy____________________________________________________ Business cycle versus other idleness---------------------------------------------------------------------Conditions of chronic unemployment-------------------------------------------------------------------Optimal pricing under high employm ent conditions_______________________________ Dovetailing and large m onopolies________________________________________________ Pooling labor reserves through a well organized market___________________________ Stabilization and differential pricing--------------------------------------------------------------------V I. Policy recommendations----------------------------- ------------ - ............- ------------------------------ 1 1 2 3 5 7 8 8 10 11 15 18 18 19 20 20 22 22 24 25 Addendum : Depression policy and the guaranteed wage......... ........... ................................. The fear of depression-------------------------------------------------Econom ic aspects of Government aid------------------------------------------------------------------Governmental grant-in-aid based on profitlevelsof firm s__________________________ Governmental financial aid based on level ofgrossnational product_______________ A more precise formula for each proposal-------------------------------------------------------- ----Various aspects of the Government aid program__________________________________ Subaddendum: Comments on the Hansen-Samuelson report--------------------------------------Comments by J. M . Clark_______________________________________________________ Comments by Edward S. M ason--------------------------------------------------------Comments by Sumner H. Slichter.......... ...................... Final comments by Alvin H. Hansen and Paul A. Samuelson...................................... 46 46 46 47 48 49 50 53 53 55 58 61 n 26 27 28 32 36 36 37 38 38 39 40 40 42 APPENDIX F—ECONOMIC ANALYSIS OF GUARANTEED W AGES I. INTRODUCTION W orld W ar I I brought fu ll employment to the American people. The com ing o f peace holds forth the promise o f a high level o f job oppor tunities and production demand fo r some time to come. But no one can today forget how in the dozen years before Pearl Harbor the Am eri can economy was plagued with widespread un employment and idle machine capacity—with an attendant loss o f money income to all classes o f society, and a reduction in human living stand ards so calamitous as to threaten the foundations o f our democratic way o f life. It is now abun dantly clear that involuntary idleness cost the American people the equivalent o f some $300,000,000,000 o f real income in the decade prior to the war. The tragedy, suffering, and frustration o f the depressed 1930’s w ill not, and need not, be per mitted to recur. It is not too much to say, therefore, that sta bility at high levels o f employment and produc tion w ill continue to be the number one domestic economic goal o f all the community—business, labor, agriculture, and the Government. Paradoxically, the United States, one o f the youngest and most vigorous o f all advanced na tional economies, seems throughout much o f its recorded history to have experienced among the highest rates o f unemployment o f all the nations o f the world. In the great depression follow ing 1929 our percentage o f unemployment exceeded that o f all other nations, not even being surpassed by the German Weimar Republic. In the recov ery years just preceding W orld W ar I I we barely regained the 1929 level o f real output, having wasted in unemployment a decade o f potential improvements in our standard o f living. Even before W orld W ar I, the percentage range o f unemployment in this country seems to have run about double that o f a country like Britain. In addition to having greater amplitude o f cyclical fluctuations (and somewhat more frequent business cycles), the American economy has also been char acterized by far greater seasonal swings. Besides, there is some evidence that the range of cyclical variation has been getting worse rather than bet ter in the course o f recent decades. No single line o f attack can be relied upon in the battle against depression. In addition to a fu ll employment program, with its many-sided and continuing policies, the Government must be pre pared with its social security and welfare measures to ameliorate the human burdens o f unemploy ment. Nor can business, agriculture and labor stand by, passively relegating to the public au thorities all responsibility and action designed to promote over-all stability in jobs and production. One device aimed at m itigating the instability o f production and pay rolls is the policy of guar anteed wages. It is the purpose o f this report to give an economic analysis o f proposals which have been put forward under this general heading. B R IE F H IST O R Y OF GU ARAN TEED W A G E PRO PO SALS The programs initiated by private enterprise to moderate irregularities in production and em ployment actually antedate by well over a quarter o f a century the explicit acceptance by Govern ment (in the Employment A ct o f 1946) o f its re sponsibility to combat depressions. Particularly after the sharp 1921 post-W orld W ar I recession in pay rolls and production, various forward-look ing corporations began to institute voluntary wage guarantee plans o f one sort or another. Others, without explicitly guaranteeing wages or employ ment, made strong efforts to smooth out the pattern o f their operations so as to provide more con tinuous annual employment. Some o f these plans may have involved overambitious altruism. Others may have represented ill-conceived paternalism, covertly directed against the organization o f unions and the development o f collective bargaining. In any case, most of these 1 schemes were short-lived, many foundering on the rock o f the great post-1929 depression. Does this mean that they were failures? Not necessarily. In an era when governmental unemployment in surance had not yet been enacted, these plans often aimed at softening the burden o f unemployment for at least a limited time. They served as a first wall o f defense against depression; the fact that these bulwarks were breached does not mean that they were not o f considerable value. Also, in all fairness it should be said that the enactment of govem m entally-provided unemployment compen sation reduced the pressing need for the continua tion o f some o f the plans. W hile the great depression o f the early 1930’s spelled the end o f many o f the earlier guaranteed wage plans, it at the same time emphasized in the minds o f everyone the great need for job se curity. And during the recovery years o f the last decade, an increasing number o f additional com panies initiated new guaranteed wage plans. Organized labor—in part stimulated by the late Franklin D. Roosevelt—begin to develop an in terest in wage guarantees during the years just prior to W orld W ar II. A t the present time, many guaranteed wage programs are embodied in man agement-labor collective bargaining contracts. Many unions in industries not covered by such contracts have indicated an intention to explore the possibilities o f this device. 2. Analysis o f methods and possibilities o f reg ularizing production and stabilizing employment. 3. Inquiry on cost o f various types o f wage guaranties. 4. Analysis o f economic effect o f guaranteed wages. 5. Examination o f relation o f guaranteed wage plans to other economic measures also intended to stabilize or increase the national income. In connection prim arily with the last two o f these points the authors o f this report have been asked to make a general theoretical analysis o f the economics o f guaranteed wages and its relations to other policies. In doing so, we have not consid ered it fruitful to attempt a review o f the volum inous literature on the subject, nor to present a well-rounded factual and quantitative picture o f the status and prospects o f guaranteed wages. This we can best leave to others. Early in our in vestigations it became clear that the greatest serv ice we might hope to render lay in bringing under analysis and focusing attention on certain relatively neglected—but crucial—aspects o f the problem. Perhaps most important o f these is the need for a critical examination o f the repercussions o f guar anteed wages upon the level o f aggregate demand and upon the business cycle. H ardly less impor tant is an examination o f the economic principles and patterns o f pricing which society needs to keep in view in order to achieve the most efficient O RIG IN OF PRESEN T STU D Y use o f our human and capital resources, the reduc tion o f wastes o f unstable production, and the In 1944, at the peak o f the war effort, several equitable distribution o f the burdens o f unavoid labor unions presented a request for a wage guar able idleness. These two m otifs—business cycle antee to the W ar Labor Board. The W ar Labor and aggregate demand aspects o f guaranteed Board ruled against the establishment at that time wages, and efficient, equitable use o f our national o f so important a new precedent in American labor resources—run through all our analysis, not only relations. On the occasion o f their refusal to grant in their relationships to each other, but also the United Steelworkers’ request, the board mem against the background o f other public and private bers indicated their interest in the proposal, and economic policies. recommended that the whole problem be given W e have not considered it our primary respon careful study. Early in 1945, President Roosevelt sibility to spell out in any detail a specific program requested that the Advisory Board o f the Office o f action. W e have been prim arily concerned with o f W ar Mobilization and Reconversion make a general economic analysis. In chapter V I we do study o f guaranteed wage plans as a means o f reg put forward certain recommendations, which we ularizing production and stabilizing employment. believe represent policies which are economically In making its investigation into the whole prob sound in the light o f our analysis. Needless to say, lem o f guaranteed wages, the OW M R proposed: 1. Examination o f specific experience with guar these recommendations—and indeed all o f our judgments and findings—are put forward only as anteed wage plans. 2 our own opinions as individual citizens and econo mists. They do not necessarily represent the opin ions o f the Advisory Board o f the Office o f W ar Mobilization and Reconversion, or the Washing ton Research Staff under the direction o f Murray W . Latimer. In part our recommendations over lap or are similar to those in Mr. Latimer’s final report.1 It is a pleasure to acknowledge here the stimulus o f discussion with him and his staff, even though we were not always able to agree fully upon final formulation o f principle and emphasis. D E SC R IPTIO N OF GU ARAN TEED W A G E PLAN S Here at the beginning it is well to ask exactly what is meant by “ guaranteed wages.” No simple answer can be given because o f the extreme variety o f plans that one company or another has intro duced under this general heading in the last quar ter o f a century. A t the one extreme, we observe what is little more than a pious declaration o f the company’s intention to make efforts to stabilize its production so as to provide steady jobs and in comes. A t the other extreme, a company may have signed a legal collective bargaining contract with a union agreeing to guarantee 45 or even 52 weeks o f work per year to almost all o f its employees. Even the three best-known plans associated in the public’s mind with the guaranteed wage—those o f the Hormel, the Nunn-Bush, and the Procter & Gamble companies— differ materially in form and substance. Because there are so many descrip tions o f the various guaranteed wage plans avail able, we do not propose to discuss here (except very briefly) the different features involved in such plans, preferring to refer the reader interested in details and in particular concerns’ plans to the voluminous literature on the subject.2 By a guaranteed wage plan, we shall not neces sarily mean an annual wage plan. Conceivably, 1 Guaranteed W ages, Report to the Advisory Board, Office o f W a r Mobilization and Reconversion, January 31, 1947. See pp. 1 -4 1 1 of this volume. * The most comprehensive study prepared to date is the Latimer report. The National Industrial Conference Board, the National Association of Manufacturers, and Brookings Institution either have issued or will shortly issue studies on various aspects o f wage guarantees; and Joseph Snider’s book, The Guarantee of W ork and W ages (Andover, 1 9 4 7 ), is ju st out. See also the bib* * liography prepared by the TJ. S. Department of Labor, April 1945, entitled “ The Guaranteed Annual W age and Other Proposals for Steadying the Worker’ s Incom e; Selected References,” and the popular book by J. Chernick and G. Hellickson, Guaranteed An nual W ages (University o f Minnesota Press, 1 9 4 5 ). the employer’s commitment to maintain workers’ incomes might be fo r longer durations than a year—fo r example, over the whole o f the business cycle. More often guaranteed wage plans are set up with a calendar year as the fundamental unit o f time. But not infrequently, a firm may wish to guarantee employment fo r only 6 months or only for the whole o f a season. Our analysis w ill apply also to such plans, but we exclude from our consideration programs concerned with insur ing only a fu ll week’s work or any period less than, say, a quarter o f a year. The form o f the guarantee may vary consider ably: Some firms may insure a certain number o f weeks’ employment at unspecified job assignments and wages; or more often, there may be a guar antee o f a certain number o f specified pay checks— even if the concern should find that it has no duties for the workers to perform . Usually, if a firm has bound itself to continue a worker’s income, it w ill cast about fo r some work fo r him to do, even i f it is o f a low -productivity character or o f a currently unprofitable and unremunerative sort. But where even this is lacking, the corporation agrees to maintain the incomes o f the workers who are included under the guarantee. The guar anteed wage becomes in this case the precise equiv alent o f 100-percent unemployment compensation, maintained for the duration o f the guarantee. Often the plan provides fo r the maintenance o f income payments from week to week at a steadier rate than hours o f employment, the overtime o f some periods being balanced against the hours o f underemployment through other weeks o f the year.3 Under the Fair Labor Standards A ct, any weekly hours o f work in excess o f 40 are usually to be paid fo r at a one-and-a-half overtime rate; but that act did provide fo r a partial relaxation o f such penalty overtime rates in cases where a collective bargaining contract is in force guaran teeing certain minimum hours o f total yearly employment. Not infrequently, guaranteed wage plans are associated with various wage-incentive or profitsharing programs. In addition to the steadily maintained periodic wage payments, there may 8 Some companies w ill lend their workers money in periods of underemployment, payable in subsequent periods o f brisk em ployment. Such wage-advance schemes are related to, but dis tinct from , the guaranteed-wage plans discussed in this report. 3 also be supplementary bonus or wage payments at longer intervals o f time. The total remunera tion o f the workers may be geared to their pro ductivity, to corporate profitability, or to the value o f production and sales. So much for the form o f the plan, whether it (a) guarantees employment, or (&) guarantees wage income, or (o) involves profit-sharing and various “ flexible-wage” features. Equally im portant are the other various features and lim ita tions o f the guarantee, such as the follow ing: (1) Lim itation o f coverage. W hat workers are covered ? Are only those who have been em ployed 2 years, 1 year, or 6 months covered? Or are only workers o f certain skills and departments covered? (2) Lim itation o f weeks per year. A re a full years’ work and income (2,080 hours per year, 52 weeks at 40 hours per week) guaranteed? Or are only, say, 45 weeks o f work guaranteed? (3) Lim itation on duration o f guarantee. W hat is the duration and legal character o f the guarantee? One prominent plan guarantees 48 weeks per year o f work to all o f its employees o f more than 2 years’ standing, but the guarantee is a unilateral one which the company reserves the right to discontinue at any time it may wish to do so. Other guarantees may run from the first o f the year to the last, and may be embodied in a collective bargaining contract with a union. Still others may be binding upon the company, but terminable after it has given a year’s or 6 months’ notice. The possible costliness, benefits, and other quan titative economic effects o f guaranteed wages all depend very much upon the type o f guarantees which are adopted—upon their degree o f cover age, the hours guaranteed, and the duration o f the guarantee. T o avoid a tedious catalog o f the various plans that are conceivable or now in force we shall find it convenient to outline a few possible guaranteed wage plans, in terms o f which our exposition may be facilitated. This does not im ply any approval o f the particular details embodied in these examples. They are set out simply as hypothetical cases: E x a m p l e 1: F or each calendar or fiscal year, the company agrees—in a legally or morally binding fashion—to provide some number o f weeks, say 45, o f full-tim e pay (or its equivalent 4 spread over the year) to a specified group o f workers, selected by seniority or length o f previ ous service. The contract or agreement may, or may not, be renewed fo r the follow ing year and its coverage and hours o f guarantee may be altered, provided notice to that effect is given at least 3 months prior to the end o f the contract year. In this case, if the corporation encountered a sudden slump very early in the year it might be obliged to incur the costs o f maintaining income for a period o f almost a year. On the other hand, if the decline in its sales came late in the year, the company might be able to reduce its pay roll in very short order. It would run the risk o f having to maintain, on the average, substantial wage payments for something like 6 months after a decline in its revenues. E x a m p l e 2 : This is the same as example 1 ex cept that the contract or agreement does not run from the beginning o f one fiscal year to the begin ning o f the next. Instead, the guarantee runs on indefinitely—until the company gives 12 months’ notice o f its intention to discontinue the guarantee or change the extent o f its coverage. One could imagine a plan identical to example 2 except that the period o f termination notice is reduced to, say 6 months. E x a m p l e 3: This could be like either o f the above two plans, but with the extra limitation that the employer’s wage liability to workers laid off should never exceed in any year, say, 10 percent o f the guaranteed pay roll. In most years, this would, in fact, provide 100-percent income security to all workers. But in a very bad year, when the number o f idle workers became very high, then the employer’s liability would have an upper lim it. In this event, the laid-off workers would have to share in some manner the amount available for wage payments within the lim itation o f the 10 percent liability. Obviously all sorts o f variations could be at tached to each o f these illustrative cases. A ll three examples have this much in common: They en hance the job and income security o f workers for at least a limited time into the future; and they in volve the employer in the risk that, fo r a limited time at least, he may be obliged to maintain his wage outpayments even if his sales revenue falls off sharply. However, the employer has not ob ligated himself to incur unlimited losses fo r an unlimited period; after the period specified under the terms o f the contract, his guarantee terminates; and, o f course, after these same periods o f time, the worker may find him self jobless and without a current pay check. One point requires notice and emphasis at this place. W hile the letter o f the guarantee may specify that the employer’s liability ends after the date o f termination, management and unions both are well aware that there w ill be considerable social pressure upon the employer to renew his guarantee and to maintain employment longer than specified in the contract. This explains why many financially strong companies, who do not doubt their own ability to finance guaranteed wages for the first year o f a serious depression, are nevertheless reluctant to offer even a limited guar antee. Not without cause, they fear that such a lim ited commitment may put them on the spot after the period o f guarantee is over. They realize that even those companies which guarantee noth ing are under more or less powerful internal and external pressure to minimize lay-offs; but they also realize that such pressures are not brought to as sharp a focus as would be the case under a guar anteed wage plan. In the follow ing discussion we shall, on occasion, have to go beyond the legalis tic letter o f the guarantee commitment to the fur ther economic implications. SOM E P R E L IM IN A R Y O BSERVATION S Before discussing in detail the outline o f our economic analysis in the follow ing chapters, we shall do well to clear the air with a brief state ment o f certain fundamental preliminary conclu sions. First, the guaranteed wage is not a cure-all for the problem o f business cycle unemployment— nor could it ever approach near to being one. Exaggerated claims, overstating its importance and advantages, do more harm than good. A t the very beginning, we wish to state for the record— frankly and explicitly—that if it ever had to come to a choice between guaranteed wages and other important depression and anticyclical public and private policies, such as our system o f unemploy ment insurance,4 our social security programs, or fiscal and monetary policy, then it would be guar anteed wages that would have to be rejected. W e do not wish to be misunderstood. As w ill become abundantly clear from the pages that follow , we do not see any necessary conflict between guaran teed wages and other needed Government policies. On the contrary, it is to be hoped that guaranteed wages can be integrated with unemployment in surance so as to strengthen income security. Also, it is our view that guaranteed wages can be financed so as to contribute something toward moderating the excesses o f boom and slump in the business cycle. Second, although the primary motivation fo r guaranteed wages comes from wage-earners’ fear o f business cycle unemployment, our studies lead us to the follow ing somewhat paradoxical conclu sion. In a world o f high effective demand and pur chasing power, where there is no problem o f chronic or depression unemployment—in such a world, guaranteed wages m ight make its greatest contribution: by stabilizing seasonal and other short rhythms in production and by equitably allocating the economic burden o f the residual amount o f unavoidable idleness and unemploy ment. This conclusion emphasizes all the more the importance o f a fu ll employment program, within the pattern o f which the guaranteed wage can play a significant role. Under the present sys tem o f lay-off on short notice, the employer finds himself with insufficient financial incentive to un dertake the costly efforts which are necessary fo r thoroughgoing stabilization. W hat are truly overhead costs to society often appear to the em ployer as avoidable variable costs, with the result that enlightened competition does not lead to the economically most efficient use o f resources. The view that the single individual firm, even a large one, can by good intentions and clever poli cies substantially contribute to solve the problem o f the business cycle, we are compelled by economic analysis and all the available empirical data to 4 There are, to be sure, many technical imperfections in our present unemployment insurance sy ste m : inadequacies of inte gration and uniformity between states, insufficient benefits in relationship to reserve accumulations, perverse cyclical aspects of merit or experience rating, etc. These should be remedied by legislation and practice. They should not be used as an excuse to attack unemployment insurance, or as an argument for sub stituting guaranteed wages or any other alternatives in place of unemployment insurance. 5 discard in the beginning. No one enterprise is secure enough or powerful enough to withstand and m odify the great swings o f general business activity. In our judgment, there is scarcely more validity in the hope that gradually, by more and more firms voluntarily extending the scope o f their stabilization activities, we shall approach finally to a condition o f continuing high employment in come and production. As we shall show later, there is every good reason for looking with ap proval upon a redoubling o f the efforts o f manage ment, on its own initiative, toward ironing out its employment and production fluctuations. And when a considerable number o f firms have carried through bold, imaginative, and well-con ceived stabilization programs, considerable prog ress w ill have been made toward reducing the costs o f economic idleness. But by far the great est progress that w ill have been made is likely to be in connection with unemployment arising from seasonal and other rhythms o f business activity o f shorter duration than that o f the business cycle. From a dispassionate appraisal o f the efficacy o f guaranteed wages, the device would appear to be, by itself alone, a weak reed upon which to rely in an attack on such widespread unemploy ment as our economy has encountered in recent years. Indeed, it would be nothing less than tragic if, by management initiative or govern mental and trade-union pressure, an ambitious program o f guaranteed wages was undertaken without at the same time instituting other more important programs o f government and business designed to promote high and stable levels o f pro duction, income and employment. W hile it is conceivable that an ambitious pro gram o f universal wage guarantees would diminish the intensity o f minor fluctuations, such as occurred in 1924 and 1927, there is danger that it might be financed in a manner that would in tensify such deep and cumulative recessions as that o f 1929. The relationship between guar anteed wages and the sharp recessions o f the 1920-21 or 1937-38 type are by no means easy to appraise. I f other governmental measures were not instituted to cope with depressions, sober quantitative appraisal o f the relationship o f pay rolls to corporate net worth shows that insolvency and bankruptcy could well be the price o f any 6 guaranteed wage program which was not care fu lly limited in its coverage and duration. Busi ness, even large business, is not the master o f its fate where the business cycle is concerned. On the other hand, we cannot be content to ac cept the chaotic status quo. F or if we look at, say, the automobile industry, which very prop erly stands out in the public’s mind as typifying the very essence o f our industrial society, the picture we see is far from reassuring. Auto mobile production is an efficient undertaking, pay ing higher wages than the rest o f the w orld and yet through greater productivity able to undersell all other countries on a free-trade basis. It is and has been a relatively profitable industry. An important part o f its total output is under the control o f three giant corporations, each o f which has presented typical chapters in American in dustrial life : the self-contained growth o f Henry Ford’s great enterprise, the evolution after W orld W ar I o f the vast General Motors Corporation, and the post-1929 emergence to fu ll stature o f the Chrysler organization. In addition the recent war has placed the remaining smaller independent producers in a comparatively favorable financial position. But if we look more closely we see that it is an industry that lives dangerously. Stocks o f fin ished cars and o f parts and raw materials are literally measured in terms o f days and weeks o f production rather than in terms o f months and years. The continuance o f production hinges on a precarious scheduling o f thousands o f compo nents produced in hundreds o f separate plants. This elaborate division o f labor and interdepen dence, which we hardly notice in ordinary times, makes for efficiency but at the cost o f vulner ability—as was revealed in the postwar disorgani zation o f the industry. Despite the fact that the use o f automobile ser vices, as measured by passenger rides or gasoline consumption, fluctuates only moderately over the year and even over the ups-and-downs o f the business cycle, the flow o f automobile production proceeds in much more fitful jerks and starts, as dictated by the vagaries o f consumer and dealer ordering and the march o f the business cycle. In consequence, the employment record o f the in dustry—the job experience o f the workers attached to the industry—can almost be termed a shocking one. Even in relatively good times workers often spend considerable time waiting to be called back to their jobs. In depression the production o f autos and trucks may decline to a quarter o f the prosperity levels. T o say that the situation is a shocking one is not to imply that any one group o f individuals or or ganization is to blame or that the record o f this industry is worse than that o f many others. D iff erent companies have at various times made de termined efforts to iron out their fluctuations in production. The whole industry together has attempted to reduce seasonal variations by means o f rescheduling the time o f introducing new models and o f holding auto shows. The results have been disappointingly meager. Clearly the historically developed institutional structure o f the industry and o f our unstable American econ omy must be blamed fo r the unsatisfactory charac ter o f the industry’s performance with respect to production stability. Under the existing set-up no one company has the financial incentive or abil ity to provide steady jobs and income. F or a cure to the fluctuations in auto production over the business cycle, we must properly look else where. But it is not too much to hope for, and to ask o f management and labor, that measures be taken and practices be altered to moderate the incidence o f irregular and seasonal employment and o f economic idleness. This may require many far-sweeping changes in the organization o f the industry: increases in inventories and storage facilities, limitation on undue coddling o f con sumers’ desires with respect to modification o f style and model changes, perhaps acceptance o f lower hourly wage rates in exchange fo r higher annual earnings. Conceivably, such a program o f thorough-going seasonal stabilization might even raise costs and prices to the consumer. None o f these measures would be easy and none would be costless. But in a real over-all sense the costs are less than those o f idleness under the present regime o f instability, where the burden falls upon those least able to bear it, the workers. Is it too much to hope that in the not too distant future all parts o f the industry w ill look back and wonder that the present day wastes o f idleness were even for an instant tolerated? W e believe not. But in taking this viewpoint, we do not mean 744705—47-----2 to imply that such a happy state o f affairs w ill come into existence sim ply by forcing the indus try, through Government legislation or collective bargaining pressure, speedily to guarantee wages for substantially all employees. Unless a general, many-sided attack is made upon the problem o f business cycle stabilization, together with a stepby-step industry-wide attack upon seasonal in stability o f production, guaranteed wage plans are likely to be so costly as to be doomed to failure. It is this dilemma—that the present day deeplyrooted wastes o f instability are economically intolerable but cannot simply be legislated out o f -existence—with which this report is concerned. PEO PO SED O U TLIN E OF A N A L Y SIS O f the five divisions o f the study o f guaranteed wages mentioned earlier, our report is naturally concerned with the final tw o: an analysis o f the economic effects o f guaranteed wages, and an ex amination o f such plans in relation to other eco nomic measures designed to stabilize or increase the national income. Such an analysis is a neces sary supplement to a study o f specific experience in different concerns and different industries. The whole is more than the sum o f its parts. Particular firms may have success with the guar anteed wage, in part because they are the only ones offering workers its advantages. The total eco nomic impact o f such plans might be negligible, or if overly ambitious or too “ soundly” financed, the device might even prove harmful to over-all sta bility. I f wage guarantees discourage investment commitments and if firms are induced to stabilize their employment at low levels in order to min imize lay-offs, the problem o f reaching and main taining fu ll employment could conceivably be aggravated. Or if guaranteed wages had the effect o f attaching workers to a single firm in subsidized idleness or inefficient production, then under cer tain circumstances national output and produc tivity might suffer. Under the present wage contract, do the appar ent money costs o f hiring or laying off labor truly reflect the overhead social costs to society o f main taining the workers’ standard o f living? What should the relationship be between prices and wages in order that economic resources may be used most efficiently? 7 Perhaps more important still is the question o f how guaranteed wages—on a limited or on a wide scale—affect the over-all level o f aggregate demand. W ould the mere act o f guaranteeing employment create attitudes and behavior patterns sufficient to realize its stated objectives? W e all know in a rough way that jobs make markets, and that markets for goods make jobs. W ould the widespread commitment to maintain jobs auto matically create the purchasing power necessary to make good such a commitment? Or to put the matter in another way, would the balance o f ag gregate demand in relation to our productive po tential be im proved; and would the basic factors causing fluctuations in investment—factors which underlie the business cycle—be compensated for? Each one o f these questions requires detailed over-all economic analysis. Each one leads in evitably to a consideration o f other economic poli cies aimed at promoting economic security and stabilizing business activity. In the chapter immediately follow ing this in troduction, we undertake to analyze the true eco nomic cost o f different kinds o f idleness: to inquire how its amount is to be measured, to inves tigate how its burden is distributed between wages and property-returns, both in and outside o f the industry concerned, and to examine the adequacy o f incentives toward regularization. Then in chapter IH the interrelationships, com parisons, and contrasts between guaranteed wage programs and other public programs such as un employment insurance and fu ll employment pro grams are discussed. Chapter IV presents an economic analysis o f the effect o f guaranteed wages on the level o f and fluctuations in aggregate demand. The results are found to differ depending upon whether the guar anteed wage program impinges on a society which is highly cyclical and fluctuating, or is reasonably stable at either a low underemployment or a high fu ll employment level. In this analysis the ef fects o f private reserve financing in contrast with the “ pay-as-you-go” method o f financing are examined. W hile the problem o f unemployment was quan titatively most pressing before the war, it cannot be permitted to overshadow the ever-present task o f seeing to it that resources shall not only be used, but used efficiently so as to contribute toward the maximum satisfaction o f individual and social economic wants and needs. Chapter V is con cerned with the effects o f guaranteeing wages upon the pattern o f resource use and upon the structure o f prices, wages, and costs. Here again it is neces sary to distinguish among the different kinds o f economic societies into which the guaranteed wage plans are introduced. In chapter V I, we make a number o f policy rec ommendations growing out o f our analysis. Such guaranteed wage plans as may be introduced should be set up only after the possible dangers and limitations have been thoroughly canvassed. When this has been done, we believe that, under safeguarded procedures, the guaranteed wage can be made to contribute to a better and more effi ciently functioning economic order, and that in ducements such as those recommended in chapter V I looking toward more widespread experimenta tion and expansion are justifiable. II. ANALYSIS OF THE ECONOMIC BURDEN OF IDLENESS In this chapter we shall undertake to analyze in some detail the precise nature o f the economic costs o f unemployment and idleness. In the course o f our discussion it w ill become clear that the prob lem o f irregular production is more than the prob lem o f general cyclical unemployment, although the quantitative magnitude o f the latter has been such in recent years as to dominate our thinking and attitudes. T o the degree that we are able to avoid serious breakdowns in the years ahead, problems (“ labor m obility” fo r example) which seemed—and were—unimportant against the background o f depression unemployment, w ill 8 take on a new significance. Thus we are concerned not only with depression unemployment, but with the problem as a whole, including daily, weekly, yearly, and longer-term irregularities and fluctua tions in business activity, many o f them avoidable. SO C IA L VER SU S P R IV A T E COSTS OF ID LEN ESS But first let us note that the present system o f hiring, employing, and paying labor is only one o f many conceivable alternatives; and one which be came widespread even in our own economy only in the last century. Before industrialization had impersonalized our economic relations, it was probably the rule more often than not for a crafts man’s helper to be hired by the year or season. Often, barring discharge fo r personal incompe tence, the worker could even count upon some measure o f job security and income maintenance even through hard times, the employer and em ployee tightening their belts together. Even today in agriculture and some branches o f small industry, the same relationship often pre vails. And throughout all industry, large and small, executives and white collar workers are in good part assured o f employment over all o f the year and through most o f the business cycle. It is perhaps not too hard to imagine an economic system in which the wage contract with all em ployees was more nearly like that with executives and white collar workers than like the present prevailing mode o f wage payment and tenure. Let us, however, consider the prevailing form o f the wage contract and analyze its economic e f fects for whatever they may be. An industry with high seasonality may hire its workers for 7 or 8 months, leaving them idle during the off-season. Similarly, a concern sensitive to cyclical fluctua tions may take on many new workers during the upswing o f the boom, and conserve its resources during the depression by curtailing its wage payments. Still, as was pointed out so clearly a score o f years ago by P rof. J. M. Clark in his masterly Studies in the Economics o f Overhead C ost:6 In a more general sense, however, there is a minimum o f maintenance o f the laborer’s health and working capac ity which must be borne by someone, whether the laborer works or n ot: that is, if it is not borne, if the maintenance is not forthcom ing, the community suffers a loss through the deterioration o f its working power which is at least equivalent to the cost o f maintaining the laborer. Thus the burden is there in any case: it cannot be avoided. From this point o f view it appears that a large part o f the cost originally counted as wages represents an over head cost which the laborer is responsible for covering as best he can, just as the employer is responsible for covering the overhead costs on account o f capital. How ever, if the laborer fails to cover it the community does not escape the burden, and it is ultimately borne by in dustry in the shape o f reduced productive power and damaged morale. And thus it comes back to the employer in any case. 8 John Maurice Clark, Studies in the Economics o f Overhead Costs (University of Chicago Press, Chicago, 1 9 2 3 ). * * * The reason why the expenses o f production, some o f them, normally vary in proportion to output is simply because the terms o f the wage contract are drawn in that way. The employer leaves the wage earner to care for his own overhead and the terms o f the contract are not scientificaUy adjusted as, for instance, the contract fo r electric current is sometimes adjusted to the over head costs o f the ultimate producer. It may be that we shall find that our general system o f wage payment is thoroughly unscientific and that a more scientific system may operate to improve the steadiness o f employment in much the same way in which scientific rate systems have been used to increase the regularity o f use o f electric power plants (p. 16). The fundamental discrepancy between true social costs o f unemployment and apparent pecu niary private costs is o f course accentuated by the form o f the present wage contract. W ithout as yet going into the merits o f such a proposal, let us suppose that by voluntary action or collective bargaining arrangements, firms were induced to provide income or work for their employees throughout all the year or all the business cycle. Then not only would the burden o f the worker’s maintenance be spread differently between the workers and the employers in the industry in question, and among the other workers and em ployers who make up the community at large, but in addition the use which the employer would make o f labor might be different. Moreover his production or sales decisions might very probably be appreciably affected. Today in a depression, a firm w ill for the most part not hire a man unless in some sense it expects to receive in value a return commensurate with his fu ll wage. N ot so does it act with respect to the machinery it owns. I f, because o f slack times, a piece o f machinery can produce only a fraction o f the value fo r which it was originally purchased, the firm w ill not thereby leave it in idleness. On the contrary, so long as the machinery can produce anything above its costs o f operation (including o f course any extra deterioration engendered by use) it w ill pay the firm to utilize the machinery.6 •T o some degree, therefore, we might expect some differences in entrepreneurial behavior of firms in such rare industries as shoe manufacturing, where much of the equipment is rented on leases of relatively short duration. Even here the changed form o f the contract, from ownership to leasing, has important reper cussions on how the various parties share the total social eco nomic burden. A n important qualification to the theoretical notion that firms do in fact disregard all truly fixed or overhead costs is indicated later. Often, in imperfectly competitive mar kets, they use full-cost considerations in their pricing— for ra tional and irrational reasons. 9 T H E ECONOMIC COSTS OF ID LEN ESS I f an industrial worker has employment only 8 months a year and must lay off for 4, the true cost or economic burden to society is the amount o f production which he is prevented from produc ing during the 4 months o f idleness. Generally speaking, we should expect this to be roughly equivalent in value to the wage income that the worker loses by unemployment plus an additional wastage o f capital and nonhuman resources as a result o f the instability o f production. Similarly, if a longshoreman works only a few days a week, the important costs to society are the foregone value o f production and income on the days o f en forced idleness. Our earlier quotation from J. M. Clark gives an understatement o f these true social costs, because the emphasis there is placed upon the necessary maintenance expenditure o f the worker during periods o f idleness. It is not improbable that an industrial worker capable o f producing $3,000 worth o f output per year might, by practising stringent economies, be able to maintain himself and his fam ily with their health, skills, and (more doubtfully) their morale intact on an expenditure o f $2,000 per year. Nevertheless, the true cost o f a year’s idleness o f such a man is still $3,000 and not $2,000. And even if science were able to devise ways o f costlessly putting men on ice, to hiber nate during slack times like bears during the win ter, the costs o f idleness would still be the same.7 W e feel it necessary to expound at some length these almost trite, fundamental economic princi ples because only in this way can a number o f cur rent misconceptions and distorted emphases be avoided. Thus, it is quite false to think that the social cost o f instability would disappear com pletely if the employer were made to pay the work ers’ salary during periods o f lay-off. Actually, this might be a highly desirable policy, because* * This is all the more evident i f we apply the argument to the less controversial and emotional problem o f a nonhuman re source like land. I f a 100-acre field neither gains nor loses in productivity by being le ft fallow for a season, can we say that there are no costs involved in its idleness just because no outlay on maintenance is required? The answer is “ N o.” The potential crop of that year is lost forever. I f the land actually goes to weeds and deteriorates through nonuse, the costs are that much greater. I f there has been a real change in society’ s desire or tech nology so that the land or worker now has a lower potential economic productivity, then the waste due to idleness is, of course, equal only to his new potential economic productivity. 10 the worker may be least able to bear the cost o f idleness; and because society may deem it desir able fo r the employer to shoulder part o f the cost in lower business profits, or wish the burden to be shifted in part forward to consumers and the gen eral public in higher prices, or wish the burden to be shifted backward in reduced wage rates fo r all workers (including those who continue to hold steady job s). But if the workers’ incomes were maintained, though they are idle, the social wast age o f productive resources would still be borne by society as a whole. The total social cost is not changed in real prod uct terms by different modes o f sharing the bur den o f idleness. The social cost for the com munity as a whole w ill remain unless greater sta bilization o f jobs and production is achieved. It is not enough merely to transfer the burden from Peter to Paul. (Perhaps we should not say “ merely” since it does make a great deal o f eco nomic, political, and social difference just how equitably or inequitably economic burdens are shared between the Peters and the Pauls o f d if ferent economic levels.) W e do not deny that different modes o f sharing the burden can have considerable repercussions on the total amount o f the burden. For example, shifting part o f the burden o f depression unem ployment from the workers to the Government may, depending upon the method o f finance, re duce the amount o f unemployment. Or making in dustry bear the costs o f certain kinds o f seasonal irregularity o f employment may provide just the stimulus necessary to cause the introduction o f effective stabilization programs. In brief it is the effective carrying out o f pro ductive employment rather than merely the pay ment o f wages that reduces the social costs o f idle ness and instability. Income maintenance with out productive employment stabilization is only a half-w ay measure which transfers the economic burden o f waste and minimizes its social impact, but fails to diminish it. When, as above, we speak o f the economic costs o f idleness, we must be careful to distinguish be tween various kinds o f idleness. O f course, no one can work 24 hours a day. The hours spent in sleep are not wasted, any more than is the time spent in leisure on Saturday afternoon and Sunday, or on holidays and vacations. W e need not regard base ball pitchers who go hunting during the weeks fo l lowing the W orld Series as wasting their time, any more than we would consider firemen on their way back from a fire or for that matter firemen playing checkers while waiting around fo r a new alarm, as being unproductive. In other words there are unavoidable cessations o f activity which are an intrinsic part o f certain productive activities. A ll cost comparisons are relative. It would be pointless to compare our ac tual national output with that which could be ob tained if the laws o f thermodynamics were other than they are; or with what we could produce i f no one required sleep and desired leisure; or with what would be producible if the seasons were con stant. To the extent that work interruptions are unavoidable and intrinsic, it w ill make for clearer thinking if we avoid the term “ idleness” and “ eco nomic waste.” Also, to the extent that curtailment o f labor re sults in a reduction o f human disutility or, what is almost the same thing, an enhancement o f the utility o f leisure—to this extent it would be ob viously incorrect to regard the resulting loss o f economic output as an equivalent total social waste. Thus, nobody laments the fact that our economic system fails to provide 84 hours a week o f work for all, or fails to find work for a subway attendant in between stations. There are only so many activities that human beings can be reason ably expected to attend to at any one time. Involuntary idleness such as is encountered over the business cycle or in many highly seasonal in dustries presents little or no saving grace in the form o f reduced disutility. Rather does our meas uring o f the cost o f idleness by the value o f lost output understate the true loss. In addition there is an extra psychic loss inherent in unemployment itself. A man thrown out o f work against his w ill is not on a vacation. In the present state o f our society, work under reasonable conditions up to some 40 hours per week and 50 weeks per year is not a curse but a boon and a benefit. The social gain o f transferring a man from frustrating in voluntary unemployment to purposeful productive employment may even be greater than the value o f the goods produced. W e may summarize our discussion o f the over all economic cost o f instability briefly as follow s: (1) The primary economic cost is measured by the loss o f national output resulting from unnecessary idleness; (2) in addition there is a secondary cost involved in the individual misery and social dis organization which grows out o f idleness itself. TH E COST OF IN S T A B IL IT Y TO TH E W O RK ER The intermittently employed worker suffers from instability prim arily from the fact that his pay largely ceases during periods o f unemploy ment. T o some degree unemployment insurance benefits, home or work relief, and severance pay may help to keep his income from falling below an intolerable minimum; but only to a limited degree, since coverage is incomplete and o f limited dura tion, and rates are deliberately set low in relation to full-tim e earnings. It appears in general to be an empirical fact that workers subject to frequent lay-offs over the year and over the business cycle tend to fare badly in terms o f average rate o f earnings over the whole year and over the complete business cycle. But it is not at all obvious in theory that this should be so. It could be argued, and it is often argued, that competition in the labor market w ill cause higher hourly rates o f pay to prevail in unstable occupations so as to compensate for their unsteadi ness o f employment. The mechanism bringing this about is supposedly the fam iliar one whereby workers refuse to go into industries offering less total compensation over time than prevails in those other industries which offer steady, fu ll time employment; and in refusing their supply, workers cause wages to be bid up in the highly fluctuating or irregular industries. Before turning to the available statistics upon the subject, let us point out the rather artificial set o f theoretical assumptions necessary to achieve this result. Firstly, the cyclical rhythms in activity o f the industry in question must be o f sufficient regularity so that workers deciding whether to enter or leave the industry can make some sort o f an approximate calculation as to the hourly rate which w ill in the future yield average earnings comparable over time to stable occupa tions. Perhaps to some extent seasonal (building trades for example) and shorter rhythms are o f sufficient regularity to be forecast by the workers to a limited degree, but the vagaries o f the busi 11 ness cycle and other fluctuations preclude even this possibility. Furthermore, there is the second im plicit as sumption that job opportunities are so plentiful that workers can afford to pick and choose between different industries and occupations. During the past few decades, this has been far from the truth. Often workers have been faced with the choice o f accepting precarious unsteady employment, or none at all. However, if high employment con ditions should come to prevail, then this second hypothesis would become more nearly true. Thirdly, an equalization o f average earnings through changes in the supply o f workers going into different industries implies that wage rates are set in different parts o f the labor market by the operation o f competitive supply and demand. Actually, and today this is so obvious that we need not elaborate, wages are to an important degree set by the forces o f collective bargaining and are a resultant o f the pressures exerted by union and management organizations. Finally, there are always individual differences in ability, training, and opportunity, all o f which tend to create inequality in the earnings o f differ ent kinds o f workers. Even if all jobs were equally stable, we should still expect important, persistent differences in earnings in different fields. How then can we tell what part o f the actual reported differences in earnings are due to different degrees o f instability o f work, and what part to other factors? The answer is we cannot, certainly not with quantitative accuracy. One thing, however, is clearly apparent from statistics relating to annual hours and earnings in industries characterized by different degrees o f cyclical and seasonal variability.8 There appears to be little tendency for hourly rates to adjust themselves in different American occupations so 8 The W ashington staff of the guaranteed wage study has tried with little success to find occupations which are alike in all im portant respects except for varying degrees of stability of em ployment in order to make such estimates. The data examined w e re : (1 ) From the Censuses of Manufactures, 1929, 1933, 1935, 1937, and 1 9 3 9 : total pay rolls, numbers of employees month by month and average annual number of employees for selected in dustries. (2 ) From Bureau of Labor S tatistics: average hourly earnings fo r industries selected for (1 ) and studies of annual earnings of employees in selected industries, published in the following issues o f the Monthly Labor R eview : February 1940, pp. 4 0 1 - 4 0 9 ; January 1940, pp. 1 7 2 -1 8 2 ; April 1939, pp. 7 8 1 7 8 8 ; October 1939, pp. 9 2 1 -9 3 1 ; December 1939, pp. 1 4 7 0 -1 4 8 2 ; July 1937, pp. 2 9 - 3 7 ; July 1939, pp. 1 6 3 -1 7 5 ; M ay 1936, pp. 1 3 3 3 -1 3 3 5 ; December 1938, pp. 1 3 9 3 -1 4 0 8 ; October 1940, pp. 8 2 3 -8 3 3 . 12 that average annual earnings over time are equal ized in industries o f different degrees o f instability. Some o f the industries with most intermittent employment—especially those in rural agriculture areas o f the South—also have the lowest hourly rates; while some o f the highest hourly and weekly rates are to be found in lines o f activity which are most stable in employment. Logical purists may assert that the differences in average earnings over time are all due to other factors; and that hourly rates have adjusted them selves over what they would otherwise have been, not so as to equalize earnings, but so as to com pensate for variability o f employment. The form o f this assertion makes it impossible fo r anyone either to verify or refute it. But such facts and analysis as have come to our attention incline us to the belief that any tendency fo r the hourly wage to rise so as to compensate fo r unemployment is only o f a weak and partial character. B y and large, therefore, we incline to the conclu sion on factual and theoretical grounds that the primary burden o f instability o f employment falls upon the worker and his fam ily in the form o f reduced average earnings over the long run. In terms o f the equity considerations which underlie the modern insistence upon progressive taxation according to ability to pay and upon the social desirability o f keeping up the standard o f life o f low-income families, placing the burden o f unem ployment wholly upon the worker represents the least desirable national policy, and one which ag gravates to a maximum the onerousness o f the given pattern o f instability. It does not follow necessarily from this that the best economic policy to remedy this situation would be one o f raising standard hourly rates to what ever level is necessary to compensate for insta bility o f employment. W e do not say this merely because such a solution would involve extreme changes in the present wage set-up o f American industry. Grave inequities may often justify drastic action. More important is the considera tion that simply raising wage rates without reduc ing instability and idleness only shifts the burden off the worker’s shoulders but may fa il to reduce the economic dollar value o f waste. This would tend to be true unless indeed there were some valid reason to expect that raising hourly wage rates would by itself tend to reduce the average instability o f production. Moreover, we need to consider the relative merits o f raising wage rates and o f changing the wage contract in the direction o f guaranteeing longer periods o f employment as means o f prom oting stable production.9 These questions we must defer until chapter Y . But there is one aspect o f the problem which deserves mention at this point. Under some cir cumstances raising the hourly wage rate may actu ally tend to increase the total o f idleness—espe cially where the problem o f “ casual employment” is involved. B y “ casual employment” is meant prim arily a condition where (a) workers are in termittently hired for short engagements, and (6 ) where the first worker who happens to present himself is likely to get the job. A surplus re serve o f workers often results from higher rates o f pay, especially if there is general unemploy ment.1 0 This line o f reasoning is abstract and oversim plified. It should not be pushed too far. But it does indicate one aspect o f reality, and it does indicate that any policy o f raising wage rates should be accompanied by supplementary policies designed to reduce the total pool o f unemploy ment. Even if normal annual earnings were maintained in unstable jobs by means o f higher rates o f pay when working, there would still remain the evil o f irregular earnings. An ideally calculating in dividual could perhaps be relied upon to budget his expenditure in equal amounts throughout the year; or in occupations subject to much longer rhythms, over the whole cycle. But for ordinary human beings spending patterns tend to follow closely upon their current income, so that uneven income payments result in ups and downs in living standards. Added to the evil o f variability o f work and pay, there is the attendant feeling o f insecurity. W ithout some tangible advance guarantee, for every worker who is actually fired a number o f others may quake in their boots. Living appre hensively from day to day, the worker cannot • A further drawback to raising hourly rates is that workers differ in their lay-off experience according to seniority and other matters. 10 A classic case is that o f the London dock and wharf workers, analyzed alm ost 40 years ago by W . H . Beveridge in his significant work, Unem ploym ent: A Problem of Industry (1 9 09 and 1 9 3 0 ), p. 107. efficiently make financial plans for the future, cannot undertake advantageous long-run commit ments with respect to home ownership, life insur ance, and other desirable forms o f personal con sumption and saving. T o be sure, form al and inform al seniority practices already serve to ameliorate, in part, the universal fear o f unemployment. But there still remains a large residue o f insecurity, which, tak ing the long view, we must regard as partly un necessary. This shows that there may even be merit in those limited guarantee plans which do not go much beyond explicitly guaranteeing job security to those who already in fact largely en joy it. The mere announcement effects may be reassuring to some degree. The question may be raised whether stabilizing employment and income over time and reducing job insecurity w ill serve to increase or reduce the average propensity to consume out o f a given income. The answer to this question, which has a bearing on the problems discussed in chapter IV , depends upon a nice weighing o f conflicting quantitative tendencies. On the one hand when people are reduced from a high income level to a lower one, they probably still try to maintain as best they can the higher standard o f consumption to which they have become accustomed. There fore, because o f the “ ups” in the “ ups and downs” o f income, there is a tendency for average saving to be lower and the average propensity to consume to be higher than with stable incomes. On the other hand the persistent fear o f unemployment tends to cause people to try to save more out o f a given income. He would be a rash man who would venture a confident definitive answer to the question o f which tendency is the stronger, no' satisfactory statistics being available. Since there appears to be no strong presumption upon either side, we may ten tatively cancel out o f our appraisal any claim that guaranteeing wages w ill have stimulating or hampering effects upon saving habits. (T o some extent steady wages improve the credit position o f workers and enable them to invest in consumers capital goods such as housing and autos. Such an increase in investment implies a higher rate o f effective demand.) Guaranteeing employment to a lim ited number o f employees raises some o f the same policy prob 13 lems involved in excessive seniority provisions.1 1 Adm ittedly, we need not argue that everyone must be dragged down to the level o f insecurity o f the least fortunate. Having granted this, however, we must still go on to point out the possible evil o f carrying seniority and wage guarantees to a fa vored group so far as to create in effect an economic caste system, in which one fortunate privileged group is protected completely from the vicissitudes o f economic life, at the expense o f throwing the fu ll burden on those left without protection. Some underemployment for all may be preferable to total unemployment for a minority. I f there must always be an irreducible minimum o f the working force unemployed, it would be far better fo r their numbers to be made up o f a revolving group o f different people, with no appreciable group being unemployed for very long. During the 1930’s, those who had been unemployed the longest had the greatest difficulty in finding jobs after business demand began to recover. They were beginning to form part o f the so-called hard core o f unem ployables, until the high level o f wartime demand melted the core and proved their employability. This points a possible way out o f the dilemma that giving security to the m ajority may inequita bly and cumulatively concentrate insecurity upon an unlucky m inority: the maintenance o f a high level o f over-all effective demand by bold public policies w ill serve to help those recently hired employees who lack seniority and who cannot qualify fo r guaranteed wages. In concluding this discussion o f the evils o f job insecurity, we must point out that this is not a matter o f concern only to the workers involved. The employer is vitally affected by the workers’ mental state in this regard. On the one hand, from a realistic hard-boiled viewpoint there is little doubt that fear o f losing one’s job sometimes acts as a whip to spur the worker on to more effi cient production. Fear o f discharge fo r cause cer tainly acts in this direction. But fear o f lay-off is quite another matter, except to the extent that the order o f retention o f men and rehiring is made to depend upon demonstrated efficiency. Even u I t is not a t all paradoxical th at even recently hired workers tend to favor seniority rules. They hope to possess seniority themselves some day. 14 in the fairly rare industrial situations where labor is in fact laid off in order o f inefficiency and re hired in order o f efficiency, there is always the counter tendency that tense and insecure workers may over the longer run be less efficient rather than more efficient producers. Neither o f these last points is so telling as the fact that increasingly in modern industry form al or inform al seniority rules determine the order o f a worker’s lay-off, rather than his own effort and productivity. No worker feels that the chances o f retaining his job are prim arily dependent upon his own efforts. But all the workers together often feel—rightly or wrongly—that by their behavior they can stretch out the work to be done and pro long the average period o f employment o f the whole group. In order to make jobs last they w ill consciously or unconsciously slow down their efforts—with detrimental effects upon productiv ity, costs, and profits. And such is the group solidarity o f any set o f men working in close con tact that even those workers with high seniority and no fear o f unemployment w ill actively and passively pursue these same tactics; fo r just as young workers look forward to the time when they w ill have gained seniority and act accord ingly, so too do the old-timers remember back to the days when they were at the bottom o f the lad der. Besides, rarely does one have so much senior ity but that he is outranked by someone else, and even his job may be thrown in jeopardy whenever there is a really deep falling off in business activ ity. Given the above set-up, increasing the work er’s feeling o f security by means o f guaranteed wage plans may contribute to an increase in labor productivity. This completes our brief survey o f the ways in which the economic burden o f unemployment tend to fa ll with heavy weight upon the workers; through (1) reduced earnings, (2) loss o f oppor tunity for useful and satisfying activity, with attendant costs to the community at large, (3) through variability o f the receipt and expenditure o f income, and (4) through a widespread feeling o f insecurity and tension among both those em ployed and those unemployed. W e turn now to an examination o f the burden experienced by the employing industry. TH E COST OF IN S T A B IL IT Y TO IN D U STR Y The burden o f unemployment fallin g on the workers is only part o f the total story. Heavy un employment causes national income to fall sharply. During such times industry is operating at low capacity levels. The rewards to property decline. Historically, there seems to have been a rather remarkable constancy in the proportionate divi sion o f the total national income between personal effort (wages, salaries, and personally earned entrepreneurial profits) and property income (interest, net rents, and profits com bined). It appears that the burden o f heavy unemployment tends to be parceled out in something like the same proportion between property and labor income. I f we take property income as a whole, probably similar conclusions can be drawn in the case o f seasonal and other noncyclical idleness. I f these irregularities cause an annual loss o f national in come and product o f some billions o f dollars, an appreciable fraction takes the form o f lower earn ing returns to property income receivers. Success in a thorough-going stabilization program would mean more stable and probably higher profits over the cycle as well as higher wages. Therefore, both labor and management have a mutual stake in ironing out irregularities in production and employment. W e have seen that the average take-home pay tends to be decreased by instability o f employment in different industries. W hat may we infer about the profit rate in industries which are subject to the greatest fluctuations in activity ? Can we ex pect that the effective maintenance o f guaranteed wages would be at the expense o f the profits now earned by variable industries? T o the extent that any wage or job guarantee induces, accompanies, or follow s upon successful stabilization o f production, the second question answers itself. I f steady production at a high level tends to increase labor efficiency and to reduce expensive turn-over, the program may much more than pay fo r itself. But i f we must assume that guaranteed wage income is maintained in indus tries that are still subject to irregularities o f de mand and production, the answer to the question o f whether the burden w ill fall on profits is not so simple. It depends in part upon the answer 744705-47-----3 to the first question as to how profit rates have in the past accommodated themselves to differing de grees o f instability. I f we select those industries with the greatest cyclical or seasonal fluctuations and compare their profit rates with steadier industries, there does not appear to be any strong relationship between profits and instability. F or the period 1935-39, which includes a recession as well as years o f ex pansion, the average, profit rate o f 10.1 percent fo r all durable goods hardly differs from the 9.8 per cent fo r all nondurable goods. Automobiles, one o f the most strongly cyclical and seasonal indus tries, has among the highest rates o f profit, whereas iron and steel, also a cyclical industry, has rather low profit rates.1 * I f industries which had previously been stable were suddenly to become cyclically or seasonally irregular, then profits would immediately be a f fected adversely. Something like this probably happened as a result o f the unexpectedly deep de pression o f 1929-33, profits in many lines being very hard hit. But by and large, unstable indus tries have a history o f being unstable. When capital chooses between going into the canning industry and some other line, it does so with the expectation that supplies and operations w ill there be bunched into a relatively few months o f the year. Similarly, an industry like baking is ex pected to be fairly stable, both seasonally and cyclically. T o the extent that these expectations are cor rect—and they certainly are not necessarily so— there is a tendency fo r profit rates to be equalized (over the cycle and year) by the competitive move ment o f capital between and among industries. That being the case, the residue o f the economic burden o f idleness that is not thrown upon the workers in the given industry—this residue is not borne by the industry’s capital. Instead it is shifted forward and backward to workers and investors elsewhere. In overcrowded lines o f activity where rivalry is very keen and profits are chronically low, the cost o f even customarily expected fluctuations in production may not be allowed fo r in price set ting. Many o f our industries with the greatest “ Unpublished study o f Office o f Price Adm inistration, entitled “ Return on Net W orth, Before and A fter Taxes, 1 9 3 6 -4 4 , 2 ,500 Leading Industrial Corporations.” 15 amplitude o f seasonal fluctuation happen to meet this description at least in part. However, for the most part, the low profits experienced in these fields ought properly to be kept analytically sepa rate from the fact o f seasonality o f operations. Were the same total level o f demand to be spread evenly over the year or cycle, while at the same time the conditions o f over-entry into the industry and past excessive investment were yet to prevail, we m ight still expect profits to be low. A real exception to this statement is provided by the case in which people are encouraged to overinvest their capital in an industry because o f the appearance o f high profits during the short-lived busy season. This would be the counterpart o f a similar possi bility in the labor field, where workers overcrowd an industry because o f short-sighted concentra tion upon the transitory high hourly rates to be earned fo r a few months o f the year or a few years o f the cycle. Other fields are to be found such as durable con sumer goods and chain department stores where operations show a distinct seasonal and cyclical pattern but where profits are nonetheless fairly high. In many markets one or a few firms occupy a strategic position o f dominance by virtue o f effi ciency o f operation, advertising, research, trade marks, patents, and so forth. New firms are not able to come in and compete away the monopoly or quasi-monopoly returns o f these concerns. From the analysis o f the last section, a few ten tative hypotheses may be drawn as to the possible effect o f introducing guaranteed wages upon the profit position o f unstable industries. Depending upon the variability o f the industry and upon the limitations placed upon the guarantee, we could expect profits to fa ll in the short run after the plans are first put into effect—unless indeed the plans induced an increase in efficiency. In the longer rim, and to the extent that capital can be made mobile, capacity and output would gradually shrink in the more variable industries. More and more the burden would then be shifted forward to the consumer through higher prices, or backward upon wage rates in general—not through actual pay cuts, perhaps, but through a slowing up o f the rate o f increase o f real wage rates. Part o f the burden might well fall upon monopoly profits. In fact the cut might be covered by increased efficiency. 16 Nevertheless the case should not be overstated. Even if profit rates were equalized over all in dustries in the long run, we could not conclude that profits would necessarily escape all the bur den o f the extra costs entailed by guaranteed wages. True, the rate o f profit in the particular unstable industries would be relatively unaffected. This does not preclude the very real possibility that the “ equilibrium rate o f profit” would be lower under a regime o f over-all guaranteed wages than under the present form o f the wage contract where labor can be made a variable monetary cost. There is no reason why the historical proportion between labor and property income would neces sarily be preserved in the face o f so drastic an institutional change. In technical terms, the “ im puted net productivity” o f capital depends very much upon wage rates, collective bargaining, labor “ productivity,” taxation and other cost conditions. Under guaranteed wages, we cannot tell with cer tainty whether equilibrium profit rates would rise or fall. Any possible effects upon profits would have further repercussions on consumption, sav ing, and investment propensities, as we shall see in chapter IV o f this report. Probably, as we have seen, workers now directly bear the major cost o f instability. Also, it is clear that the portion which workers do not bear may in part be shifted off the shoulders o f the em ployers directly concerned to consumers and to industry in general. But still there remains some element of cost to the industry directly involved. Is it not to the selfish advantage o f the industry itself to regularize its activities by means o f sta bilization programs? The answer is—to some de gree—yes. Doubtless industry is already pursu ing many such policies; the observed degree o f ir regularity o f operations in our economy is by no means the maximum conceivable. Doubtless, too, much more remains to be done in this connection. Business management is far from omniscient. Custom, ignorance, and inertia play a role in preventing companies from actively undertaking policies which are to their own maxi mum long-run profit advantage. Often companies are not fully aware even o f their own irregular pattern o f operations; many devote far less atten tion to stabilization programs than to other as pects of sales promotion, product development and research, quality control, and cost accounting. The pressure of management groups such as the Ameri can Management Association, the Chamber o f Commerce, and the Committee for Economic De velopment, and the pressure o f organized workers through collective bargaining, and o f government and the public-at-large, is all to the good in spur ring corporations on to studying the problem o f further stabilization programs. Engineering and efficiency studies have time and again shown that the surface is as yet hardly scratched. When every corporation o f any size has a major executive and sizeable staff working creatively upon this prob lem—then and only then can we expect even the self-interest o f the corporation, to say nothing o f the economy at large, to be well served. We can expect that the obvious stabilization policies, which hold out immediate economies with little cost and no risks, have already been exhausted in large part. What is needed is for management to display with respect to thorough-going stabili zation efforts the same venturesomeness and im agination that it displays in prom oting a new prod uct or in risking the construction o f a new plant. And yet, despite the above exhortation, there are good reasons to believe that under present com mercial and financial institutional set-ups, the hard-boiled single enterprise has only a limited motive to pursue extensive stabilization policies. During the last 25 years, good w ill, effort, initia tive, and intelligence have been applied again and again to the problem. Yet plan after plan for voluntary stabilization has petered out or become a casualty o f business depression. W ith a loose labor market and wages a variable cost, irregular production is not always expensive production. This condition may not last. W hile it would certainly be premature to conclude that the busi ness cycle is a thing o f the past, nevertheless, both public and private enterprise have learned much; and today there is widespread recognition and de termination that we must do everything in our power to prevent such catastrophic declines in national income and production as characterized the slump which followed 1929. In any event, the longer that high employment conditions are maintained, the greater w ill be the immediate dollar-and-cents incentive to regularize any remaining instability o f productive activities. Depression conditions in the labor market have left corporations with a heritage o f personnel policies, which however appropriate they may have been in the years o f labor surplus before the war, may prove extremely costly in a high-employment economy. Let us take the automobile industry as a case in point. Suppose manufacturers bow down to the seasonal character o f auto demand once the immediate postwar backlog o f demand has been partially filled, and in consequence are able to provide only some 8 months’ work to the average employee. What w ill the result be? W ill workers be compelled to lay off for 4 months, living on unemployment-insurance benefits and previous savings, and borrowing, or on relief? Probably not, if jobs are really plentiful. Then, when auto production is again resumed, many w ill be called but few w ill answer. Large numbers o f workers w ill have been attracted by job offers with other concerns. Any firm with unstable employment would have to raise wage rates and undertake costly recruitment policies to keep a suitable labpr force. During depressions the cost o f labor turn-over to an employer consists prim arily o f the costs to be incurred in securing and training a new worker. In modern mechanized industry, where a premium is placed upon general intelligence, manual dexterity, and basic training in the use o f nonspecialized machinery, it is not at all im possible that the expense o f training new workers is a good deal less than would appear from many exaggerated accounts. A fter not many weeks, a worker transfered to a new department or hired for the first time, may be nearly as efficient as he is ever likely to be; especially since the novelty o f a job often elicits an enthusiasm and zeal which wears off with continued experience. Possibly this explains why the alleged reduction o f cost o f labor turn-over has been so ineffective in the past as a spur to stabilization. But in a full-em ploy ment economy, every employer w ill be pitted against every other employer in an effort to secure more labor in a tight market. Labor laid off may 17 be lost forever. From any one firm’s viewpoint, labor separations then tend to give rise to labor shortages rather than simply to labor turn-over. This leads us to stress once more that produc tion stabilization and the guaranteed wage involve much more than business cycle unemployment. The follow ing chapters undertake to show the rela tionship between guaranteed wages and other pro grams; also, the direct relationship to the over-all level o f aggregate demand and unemployment, and to the efficient use o f resources in a high-employment economy. III. THE GUARANTEED WAGE IN A BROAD SECURITY PROGRAM Substantial security o f income or employment, or both, is everywhere more and more recognized as a primary goal o f modern democratic societies. This goal cannot be achieved by any one method. In all modern societies many institutional arrange ments, programs, and procedures have been adopted, designed to contribute to stability o f in come or employment. Am ong the most impor tant o f these are (a) a variety o f social security measures and (b ) various programs to stabilize employment in the individual firm and for the economy as a whole. D IF F E R E N T PATTE RN S OF JOB SECU RITY Employment contracts under modern conditions typically are terminable without notice or upon so short notice as to give virtually no security o f ten ure. Nevertheless a substantial proportion o f modern workers feel fairly secure in their jobs. In many cases it may be assumed that employees o f long standing regard their positions as more or less permanent. Such employees are not likely to fear dismissal fo r cause. In general, they are not likely to be greatly concerned about lay-offs, ex cept in the case o f wholly unforeseen catastrophes o f a character which are not regarded as very probable. It appears true that human beings are prone to be unduly optimistic and to underesti mate risks o f a character which they are not in the habit o f facing more or less continuously. By and large, employees o f long standing probably experience a feeling o f security greater than is in fact justified by the course o f events. Employees who feel a sense o f security in their job by reason o f long tenure are likely to under estimate the importance o f unemployment insur ance, and related measures, as means o f ensuring some degree o f income security. F or these em ployees it is security and continuity o f the job la already held that is really important, not the eas ing o f transition to a new job. But a large proportion o f employees are not in this favored position. Typically their tenure o f employment with any one firm is relatively brief. F or many workers, a job is, even at best, a compar atively temporary thing. Lay-off or voluntary separation is a continually recurring experience. Workers who are accustomed to only short-period job tenure doubtless fall into different categories with respect to reemployment experience. Some, by reason o f the character o f their occupation, adaptability, resourcefulness, etc., expect easily and quickly to find new employment. In the trans ition to a new job, unemployment insurance plays an important role. In bad times, however, reemployment, even for the alert worker, becomes difficult if not impossi ble. Unemployment insurance becomes then an important means o f support, just as it is fo r those workers who encounter difficulty at any time in finding new work. It provides, however, only a fraction (usually 50 to 60 percent) o f normal earn ings, with maximum limits ranging from $13 to $28 per week; while the maximum duration o f benefits runs for only 14 to 26 weeks. It can serve as an important stop-gap fo r seasonal and transi tional unemployment under conditions in which opportunities for reemployment are reasonably good—that is, in a relatively buoyant labor market where aggregate demand is on a fairly satisfactory level. In addition, the dismissal wage can help to bridge the gap in the event o f technological unem ployment, consolidation o f firms, or a secular de cline in a firm’s business. A considerable proportion o f the American labor force, even though at times laid off, does not typically change jobs. Such workers regard themselves as “ attached to a firm” even though they experience frequent and fairly long periods o f unemployment. Thus, in the automobile in dustry in the period o f 1935-39, workers were employed on the average fo r only about 75 percent o f full time.1 * The table below gives a quantitative picture o f the number o f workers o f different years’ service with a large public utility system. In 1929, when jobs were plentiful, the turn-over o f workers was very high. Consequently, the percentage o f workers with less than 1, 2, or 5 years o f service was relatively high. During the depression, the picture was changed greatly. Few workers had been recently hired, and the number “ separating” from the company was relatively small. Workers with 5 years’ service began to be the rule rather than the exception, so that even workers with high seniority rating would have faced lay-offs if fur ther widespread contraction o f operations had become necessary. Percent of workers with varying lengths of service in a large utility system With 2 years or less With 5 years or less Number of separations as a percent of average number of workers 38 2 12 10 26 61 33 17 22 36 31 10 7 7K 12 Percent of workers Year With 1 year or less 1929..................... 1932....... .............. 1936..................... 1938......... ............ 1941..................... 25 1 9 2 19 In strongly organized industries, workers are less vulnerable than form erly to dismissal, or dis charge “ fo r cause.” So long as times are good, union workers with seniority are also protected from lay-offs in a wide variety o f industries where contraction o f output is a gradual affair involving only the more recently acquired workers. There are other industries, however, where technology requires whole plants and units to be shut down when output is contracted. In such industries, even strong union organizations cannot protect workers with high seniority rating from being u The average weekly hours o f automobile workers for the 1 9 3 5 -3 9 period, using Bureau of Labor Statistics data weighted by the Bureau’s employment indexes, were 36.1 . These were the average weekly hours for employees working. T otal labor force w as estimated from numbers o f employees reported to the Fed eral Bureau o f Old-Age and Survivors Insurance, with a correc tion for inclusion o f nonproduction workers. The automobile labor force appears to have averaged about 18 percent more than the average force at work, so that average weekly hours for the whole force averaged about 3 0.6. intermittently laid off—even in years o f fairly high business activity. The demand fo r a guaranteed wage is likely to be strong in industries o f this type. Such workers, to be sure, not infrequently take advantage o f un employment insurance and job opportunities offered elsewhere in the interim period o f tempo rary unemployment. Y et a large proportion more or less regularly experience temporary unemploy ment each year, and await a return to their old job without obtaining, in the interim, work elsewhere. T o repeat, it is particularly for workers o f this character that the guaranteed wage plan makes a strong appeal. These workers are, in a sense, “ attached to a firm,” yet the number o f days worked per year is low. The conditions o f the industry prevent or make difficult a high ratio o f days worked to working days in the year. I f the firm could diversify its production so as to dovetail employment throughout the year, a high ratio o f days worked to working days could be achieved. In view o f the character o f the modem labor con tract, it is probable that most employers have not yet seriously tackled the job o f giving steady em ployment to workers regularly attached to their firms. I f the labor contract were on the guar anteed basis, these firms would be compelled to explore to the utmost ways and means o f finding steady work fo r those “ attached to their firm.” GU ARAN TEED W A G E S AN D STA B IL IZE D EM PLOYM EN T Where unemployment is prim arily a matter o f seasonal fluctuations, and other types o f work in terruptions o f a short-term character, the guaran teed wage plan seems to be peculiarly suitable as a device to help solve the problem o f job security. I f under such conditions the guaranteed wage became the rule fo r all competitors in an industry, more and more those competitors would survive who proved most resourceful in the development o f stable, year-round employment. A t first, doubt less, relatively little progress would be made and the guaranteed wage might prove to be a heavy burden on the industry. This m ight involve a rather drastic readjustment o f prices or wage rates or both. Eventually, however, as more and more firms increasingly solved the problem, costs would be reduced. Accordingly prices could be lowered 19 again or wage rates raised, or both. The net social gain would be steady year-round employment for the workers and an enlarged real income for the community as a whole. A L IM IT E D G U ARAN TEED W A G E In what has been said above, the assumption is made that the guaranteed wage would be applied to all, or nearly all, the workers employed by the firm. But even firms with short-run irregularities in employment have a certain residue o f workers who are continuously employed. I f the concept o f the guaranteed wage were applied only to these workers, the plan would become at once more fea sible, but also would accomplish relatively little. Similarly, with respect to industries afflicted with cyclical unemployment, a guaranteed wage which involved only that residue o f workers which a firm can reasonably expect to hold through even a deep and prolonged depression, would become highly practical, but correspondingly also o f less value. Nevertheless, even this limited concept deserves consideration. It would indeed not solve the prob lem o f either income or job security. These diffi cult goals must be attacked from a great many angles. The guaranteed wage is only one among many approaches to the problem. Indeed, if too much were attempted by this method, other more promising methods might be seriously impaired. The guaranteed wage, we believe, does have its place as a part o f a larger program. Even the minimum concept o f the guaranteed wage referred to above could achieve something o f substantial value. I f each firm gave assured employment to that fraction o f its peak employment which ex perience indicates to be feasible and practical, a form ally recognized status o f security might be given to an appreciable fraction o f the entire labor force. This would, at any rate, be something to start with. The problem would then be sharply posed: if such a fraction o f workers were protected under the guaranteed wage status, what could then be done (1) to provide in other ways (dismissal wages, unemployment insurance and over-all sta bilization policies) as much security as is possible fo r the uncovered employees, and (2) to expand gradually the number coming under the guaran teed wage plan. 20 T H R E E RO AD S TO JOB AN D INCOM E SE C U R ITY Comparison may usefully be made between (1) the guaranteed wage plan, (2) unemployment in surance, and (3 ) a full employment program. In making this comparison it w ill be helpful to as sume that each measure is a part o f a general pattern in which each reinforces and supple ments the others as means to provide security for workers. Thus considered, what can each o f these three approaches contribute to the problem o f security? The full employment program aims to provide security by ensuring a good market fo r labor services. The goal is adequate aggregate demand, so that workers can readily find and obtain jobs. This plan does not guarantee a job to any individ ual. It provides a favorable market in which he may seek a job. The mere fact that the aggre gate demand for goods and services is sufficiently high to ensure a good market for labor services, w ill not prevent fluctuations in employment in seasonal industries. It w ill not prevent shifts taking place between industries and between regions. A man living in town X may not be able to get a job in his particular skilled occupa tion in town X ; or anywhere if his skill has be come obsolete, or his demands exorbitant. Aggre gate over-all demand does mean, however, that job opportunities are in general plentiful through out the Nation. A man cannot sit on his front porch and expect just the job he wants to come to him. But if he is alert and active, the market is favorable. Labor is in good demand throughout the country. There are many job opportunities from which one may choose. And retraining pro grams, employment offices, travel allowances, and adequate housing facilities (essential parts o f a complete fu ll employment program ) should make the task o f transfer to new jobs and other regions relatively easy. "Such in brief is the goal or function o f a fu ll employment plan for job security. It does not guarantee a job, and at best there w ill be unem ployment. In particular there w ill be seasonal, technological, and transitional unemployment. The plan thus falls short o f adequately meeting the problem o f security. Something more is needed. The unemployment insurance program aims pri marily at a minimum o f income security, and does not directly attack the problem o f job security. The program assumes that workers w ill become unemployed for all sorts o f reasons—seasonal fluc tuations, fluctuations in the business cycle, changes in technology which throw people out o f work, structural changes which adversely affect certain regions, shifts in population, foreign competition, etc. The program is not designed directly to re move any o f these causes o f unemployment. It is designed merely to ensure a temporary minimum income to workers who are out o f a job. As a by product, the program does indeed, it is now be lieved by most economists, make some contribution (if properly financed) toward the maintenance o f aggregate demand. But at best its contribution to the maintenance o f purchasing power is rela tively limited, and its main function is to provide a minimum income to those out o f work. T o re peat, the program is designed to give minimum in come security, not job security. The income re ceived is not paid in return for work. The pay ment is made basically in accordance with the principle o f partial income maintenance,1 not in 4 accordance with the principle o f productivity. I f now we extended unemployment insurance (industry or firm financed) far beyond the limits which are now actually applied in practice to the point o f full coverage o f wage income fo r the period o f a year or perhaps a minimum o f 40 weeks o f the year, instead o f only 50 or 60 percent o f wages for 14 to 26 weeks, we should have ar rived, at least in terms o f the income receipts o f the protected individuals, at the same point as the guaranteed annual wage. Under the guaran teed wage, o f course, the worker not only receives a continuous income payment during the period fo r which the wage is guaranteed; he may also be subject to the call o f the employer for the perform ance o f work. From one point o f view the guaranteed wage can be looked upon as 100 percent unemployment insurance, fo r at least a limited time. But it is often more than that, especially where it involves 14 A rough calculation suggests that unemployment compensa tion covers less than a fifth, and possibly not much over a tenth, o f the wage loss from all unemployment. The system is much more effective in compensating total unemployment wage loss in good times than in bad. “guaranteed work” as well as income. It carries the principle o f merit (or experience) rating to its ultimate degree; the whole burden o f main taining the worker is assumed by the employer. Any use he can make o f the workers is all to the good since his labor costs go on anyway. Under the unemployment insurance plan, the worker is obligated, in return fo r income payments, merely to register at the public employment office and actively to seek work if it is available. I f work is found, the insurance payments cease and wage payments begin. Under the guaranteed wage, wage payments continue with or without work, but the worker is under obligation to work when ever the employer so directs. And especially if the guaranteed wage is integrated with unem ployment insurance, as we suggest, he w ill also be under obligation to seek work elsewhere. The guaranteed wage plan, accordingly, as op posed to the principle o f unemployment insurance, tends in the direction o f bringing work to the worker in the employer’s establishment. Unem ployment insurance, through the unemployment exchanges, seeks, on the other hand, to send the worker to the job wherever it may be. This d if ference can, however, easily be exaggerated since an employer under the guaranteed wage also may seek to place his idle employees in jobs elsewhere and with other establishments.1 1 5 9 Nevertheless, the guaranteed wage tends rather more than unemployment insurance toward fixity o f jobs in regions and establishments to which work must be brought. The maintenance o f labor m obility is no easy m atter; yet much could be done to keep productive resources fluid in refer ence to changes in technology and in demand. Employment offices, operated in conjunction with unemployment insurance, if effectively managed, aim to promote the movement o f workers from old establishments and regions to new ones as dynamic changes in products and in wants arise. Under the guaranteed wage plan the main effort is to provide work for the worker on his old job or at least in the firm to which he is attached; un der unemployment insurance the emphasis is placed upon connecting up workers and jobs in any firm offering suitable work, preferably in the 19 See our policy recommendation in chapter V I calling for an integration of unemployment insurance benefits and guaranteed wages. 21 region where the worker is located, but if not there then elsewhere, at least within reasonable limits. T o sum up, a fu ll employment program is de signed to maintain, within reasonable limits, ade quate aggregate demand. Emphatically, it w ill not guarantee every worker a continuous job. It does no more than to ensure plentiful employ ment opportunities—a brisk demand for labor. And in order to achieve reasonably fu ll and sus tained employment, a fairly high degree o f labor m obility is required. A fu ll employment program must therefore include, as a part o f the plan, re training provisions, travel allowances, and ade quate provision fo r working-class houses where jobs are available. Nevertheless, at best, there w ill still be seasonal, technological, and transitional unemployment. A fu ll employment program, therefore, leaves many workers unprotected unless supplemented by a well-developed scheme o f unemployment insur ance. Social security is therefore necessary, in order that at least a minimum income w ill be re ceived by those workers who suffer the effects o f the unavoidable unemployment which is a by product o f the manner in which production is carried on under modern conditions. This insur ance o f a minimum o f income, since a job cannot continuously be guaranteed to everyone, is the right o f all workers, and is so recognized in all modem democracies. W ithin the framework o f a fu ll employment program, supplemented by unemployment insur ance, a highly important function can be served by the guaranteed wage. There arp already a substantial number o f persons, including profes sional and salaried workers and some wageearners, who have achieved a fa ir degree o f job security, if not in a form al guaranteed wage, at least in its practical equivalent. This group could doubtless be enlarged even without a fu ll em ployment program. Standing alone, however, without the protecting influence o f a steady, brisk demand fo r labor, any widespread expansion o f unlimited guaranteed wages might encounter seri ous difficulties. W ithin the pattern o f a realizable high employment program o f sustained aggregate demand, the guaranteed wage could be extended to include a fairly high proportion o f workers in the labor force. F or some firms or industries the number o f weeks guaranteed m ight be 40, fo r others 45, and for others 50. The guaranteed wage, safeguarded in view o f the risks involved and properly integrated with unemployment in surance,1 could play a significant role as a means 6 o f providing income security. Most important o f all, however, is the task o f securing contin uing high levels o f employment. Income security cannot take the place o f steady productive employment. IV. GUARANTEED WAGES, OVER-ALL EFFECTIVE DEMAND, AND THE BUSINESS CYCLE IN TRO D U CTIO N AN D B R IE F SU M M ARY The subject matter o f this chapter, the quanti tative effects o f guaranteed wages upon total de mand and upon business cycle fluctuations, is intrinsically complex and can only be analyzed in rather technical terms. Perhaps it is fo r this rea son that the voluminous literature on guaranteed wages scarcely scratches its surface. Nevertheless, no economic analysis o f guaranteed wages could be complete without coming to grips with the ques tion as to how effective such plans can be in reduc ing over-all unemployment and in moderating the business cycle. F or readers who are not interested in the techni calities o f analysis, a few o f our more important findings may be briefly sketched. However, to ap 22 praise their validity, there is no escape from the more detailed analysis which follow s this section. In this chapter we do not consider governmental programs o f any kind. I f the guaranteed wage were to be a govemmentally financed program, that would be one thing; and the economic and political implications o f such a plan would then need to be examined. But this is not our concern in this chapter. W e consider here the economics o f a privately financed guaranteed wage. How would a privately financed guaranteed wage, taken by itself, affect (a ) the fluctuations o f the business cycle, and (6 ) the level o f effective demand, pro duction and employment over the cycle. It is im portant, we repeat, fo r the reader to remember lf The problem of integration is discussed in ch. V I. that the problem before us is concerned exclu sively with the economic effects o f a privately fi nanced guaranteed wage. Those who concentrate upon the expansionary effects o f guaranteed wage payments, to the neg lect o f how these payments are financed, come out with an exaggerated notion o f the efficacy o f wage guarantees to combat depression and unemploy ment. It is not inconceivable that the guaranteed wage could be the means o f inducing (by providing worker security and in other ways) a higher man hour productivity. But there is no evidence from experience that increases in productivity, as such, tend to iron out the cycle or to raise the level o f employment over the cycle. Increases in produc tivity are important from the standpoint o f rais ing the standard o f living, but not from the standpoint o f eliminating the cycle. Guaranteed wage payments (whether they induce increased productivity or not) may, however, help to iron out the cycle if they are financed from reserves accumulated in good years and paid out in bad years; or they may, to some extent, raise the level o f employment in the depression phase, through the redistribution o f income when financed on a pay-as-you-go basis. The fact that a guaranteed wage plan might more or less “ pay for itself” is certainly signifi cant. But only insofar as the plan redistributes income over time, or else redistributes income be tween economic groups, w ill it have any effect either on the stabilization o f employment or in raising the level o f employment. Our tentative conclusion is that the guaranteed wage, by itself, is not likely to prove very effective either as a means o f stabilizing the cycle or o f liftin g the level o f employment over the cycle, although the guar anteed wage may be partially effective in regular izing seasonal employment. W ith respect to the redistribution o f income, nobody can say just what percentage o f the finan cial costs o f guaranteed-wage plans w ill fall on corporate stockholders rather than on consumers or wage earners. Except to the extent that the costs fall upon net corporate savings (without affecting dividends or interest outlay, or net cor porate outlays on plant, equipment, or inven tories), guaranteed wages may be regarded as a form o f personal income redistribution. Such a 744765-47----- 4 redistribution may be highly desirable on social grounds. But the available statistical data show ing (a) how people in different income classes allocate their extra dollars between saving and con sumption expenditure, and (&) the relatively small proportion o f the total national income flowing to the high-income groups, suggest that such re distributions are o f somewhat limited quantita tive significance in raising the level o f over-all purchasing power. Apart from such personal redistribution and its relatively limited effect, we conclude that in peri ods when net investment outlays for the economy as a whole are deficient, causing a decline in em ployment and in business profits, the guaranteed wage may, indeed, fo r a time maintain consump tion expenditure. F or this to be true in these circumstances, the payments would have to be financed from accumulated business reserves pre viously set aside to cover guaranteed-wage pay ments in the subsequent depression, or else from current undistributed profits and depreciation charges not invested in plant, equipment, or in ventories. Under the assumptions made (apart from redistribution o f personal income via lower wages or higher prices), these would be the only methods o f finance, government support being ruled out. But the amount o f guaranteed wages that could be financed from depression business profits and surplus would be quite limited, even though no account were taken o f possible unfa vorable effects on the economy which such losses might entail. And i f the amounts set aside in reserves in boom years are large enough to cover the risk involved, new problems are encountered which are rather serious. W hile moderate reserve accumulations could help to iron out the cycle, large reserves (as we shall see in the discussion which follow s) might seriously intensify the savings-investment problem. Thus, either way, there are, we believe, rather severe limits to the extent to which the guaranteed wage can iron out the cycle or raise the general level o f employment over the cycle. W e do not wish to overstate or overemphasize these somewhat pessimistic conclusions. Indeed, in the discussion that follow s they are qualified at many points. Nonetheless, we must once again stress the limitations o f the guaranteed wage as a remedy for the business cycle, and emphasize 23 the need for a well-rounded full-employment pro gram including both public and private policies designed to keep our economy stable at a high and growing level o f real income. JOBS AN D M ARK ETS Let us begin our detailed technical analysis by considering the simple view that guaranteed wages and employment maintain incomes and jobs, that steady incomes mean steady sales o f business, which in turn give rise to high and steady em ployment. In short, jobs make markets, markets make jobs, and so forth, indefinitely. Accord ing to this view, a serious depression like that o f the early thirties is nothing but a vicious cir cle; for one reason or another, the public and business has become frightened enough to con tract spending; this initial contraction o f spend ing has resulted in reduced levels o f business sales and consumers’ incomes; and this situation tends to perpetuate itself at low levels o f national in come and high levels o f unemployment. Some such theory as this is widely held both by many conservatives and many liberals. The latter say, “ A ll we have to do to cure such a situa tion is to prime the pump fo r a little while. Once the system gets back on its feet the Government pump prim ing can stop because the new circular flow o f income-sales w ill be able to maintain itself at high levels o f employment/’ W e believe it fair to say that the m ajority o f modern economic experts reject this view just as they reject its oversimple conservative version which goes something like the follow ing: “ Any depression can be temporary if businessmen and the public w ill keep their heads and realize that prosperity is just around the corner. The Gov ernment should steer clear o f the whole situation, except fo r a little necessary emergency action. Then, if people w ill only act upon the assumption that conditions are fundamentally all right, and if they w ill only ‘buy now’ instead o f later, the corner o f prosperity can be permanently turned/’ The m ajority o f modem economists, let us re peat, do not believe that such a theory squares with the monetary, statistical, and historical evi dence accumulated by students o f the business cycle. They find themselves forced to reject the assumption that, like Gilbert and Sullivan’s allpow erful Mikado ( all we have to do is declare jobs 24 guaranteed, and ordering that this be done is just as good as if it were done. W e do not deny that there is a germ o f truth in the above view point: Pessimism and reductions in spending do involve some elements o f the vicious circle character, and the descent into a depression is o f a cumulative, self-aggravating nature. But fundamental facts o f economic life are ignored by this overly simplified theory. The fact is that people do not spend all their income on consumption goods and services. Retail sales by business to the public constitute only a part o f the total national income produced at high levels o f employment. Around one-eighth o f na tional income,1 when employment is high, has 7 throughout our history o f the last half century gone into the form o f investment goods or net capital formation. And it is capital form ation that basically represents the difference between prosperity and depression. A ll the available sta tistical evidence shows conclusively that capital goods and durable consumer goods experience the greatest amplitude o f variation over the busi ness cycle. Moreover, this variation spreads throughout the economy, in particular affecting the volume o f nondurable consumption expendi tures and replacement capital expenditures, and so it has a magnified effect on aggregate outlays, income and employment. F or the purpose o f this report, it is not neces sary to enter upon the controversy as to just what the long-run prospects for investment are, or just what in the past have been the primary causes o f investment fluctuations. It is enough to record the fairly substantial agreement o f economists that among the important factors affecting invest ment behavior are scientific inventions, tech nological innovations, population movements and grow th; changes in tastes; relations between d if ferent elements in the price, cost and wage struc ture; interest rates and availability o f capital; investor psychology and businessmen’s expecta tions; attitude, legislative acts and policies o f government; governmental public works and pub lic development projects; and so forth. Now it is not at all clear how much, if at all, the guaranteeing o f wages as such would affect 17 I f one preferred to relate gross investment to gross national product, the ratio o f investment to national product would be far higher. the sum total o f these factors in a manner favor able to the maintenance o f the high levels o f total outlays necessary fo r fu ll employment—if, with out guaranteed wages, otherwise adequate demand wer^ inadequate. It is not even as yet clear in what direction or to what degree the over-all balance o f saving and investment would be affected by the widespread adoption o f the guaranteed wage. W ILL GU ARAN TEED W AG ES S T A B IL IZE CONSUM PTION? But cannot the follow ing limited case be made fo r guaranteed wages: W ould not the widespread adoption o f such plans at least succeed in stabiliz ing consumption, with the effect that fluctuations in national income would then be confined to the relatively small sector o f the economy—the cap ital-goods industries? Also, would not the suc cessful maintenance o f consumption moderate, at least in the short run, some o f the wide swings o f investment—those dependent upon the level o f current business sales ? Finally, would guaranteed wages raise the total level o f consumption over the cycle, and so also lift the average level o f investment? These are separate but related questions. The answer to the first is fundamental and prepares the way to an answer to the others. I f it can be shown that guaranteed wages would stabilize the level o f consumption, much (though far from enough) w ill indeed have been proved. F or under the present system o f wage and job tenure, fluctua tions in the relatively minor (say, one-eighth) part o f our economy concerned with net investment give rise to “multiplied and amplified” fluctuations in consumption as well. Thus if full employment net national income in the postwar were about $160 billion, and net investment were to drop away to nothing, the loss to the economy would not simply be around one-eighth o f national in come— or some $20 odd billion.1 As workers and 8 property-owners in the capital-goods industries lost income and contracted their consumption ex 18 Even when gross investment is positive in the sense that some capital goods are being produced, net investment may well be negative if the total value of capital goods output is less than the amount of capital goods being used up. In view of the fact that gross and net investment differ only by the fairly constant magnitude of capital consumption or depreciation, absolute changes in gross and net investment can be used interchangeably for purposes of business cycle analysis. penditure, there would be set in motion a chain o f cumulative contractions in income which might be two or three times as large as the initial decline in investment expenditures. In short, a relatively small failure in investment outlay tends to result, under our present system, in considerably larger changes in total income, because o f the “ secondary” contraction o f jobs and consumption expenditure. I f the consumption outlays could smoothly and instantly rise whenever investment declines, we should not experience any general decline in ag gregate demand in a prolonged depression. But this is simply not possible. Individuals w ill not suddenly change long-established customs and habits with respect to the disposition o f their in comes between consumption and saving merely because, from the social standpoint, the turn in the cycle calls for an increase in spending out o f a reduced income. Thus an offsetting increase in consumption cannot and does not occur. Worse than that, consumption actually falls off when in vestment declines, since unemployment and declin ing business prospects induce a decline in private spending patterns. This induced decline in consumption could in deed be avoided if, by some “ magic,” substitute in come payments could be found to take the place o f the declining capital-goods pay rolls. These sub stitute income payments, since they do not corre spond to any productive output, would be in the nature o f “transfer” payments. Ideally, they might be financed out o f the stream o f saving which always characterizes full-employment income levels. By this “ m agical operation” the income flow would be maintained even though net invest ment should fall to zero and cause considerable primary unemployment. In short, the savings stream, instead o f financing new investment, would be transferred to idle workers and the owners o f idle resources. Total consumption would thereby be maintained. Perhaps a bald statement o f the problem in these terms may help to reveal the lack of realism in volved in over-simplified assumptions with respect to the maintenance o f consumption and employ ment. Unfortunately, economic analysis o f guar anteed wages does not seem to substantiate the claim that this device would, as a first approxima tion, necessarily tend to maintain consumption for 25 any appreciable period o f time at the appropriate full employment levels, i. e., roughly at seveneighths 1 o f the full-employment level o f net na 9 tional income. T o illustrate the point, let us consider the case where investment opportunities are only such as to replace existing capital. It is not enough then for all employers in the consumption goods industries to undertake guaranteed wage plans. Rather, all employers—in both consumption and capital-goods industries—must carry through with such guaran tees. Total consumption and saving typically de pends upon total national income (at least as a first approximation) in such a way that in order to ensure a volume o f consumption equal to seveneighths o f full employment national income, the actual amount o f national income accruing to peo ple must be eight-eighths of the full employment level. I f investment opportunities are fo r the mo ment moribund, where is the “ missing eighth” to come from ? 40 I f investment opportunities are temporarily in large part saturated—plant and equipment hav ing been built up during the preceding boom period to about as far as prevailing technology justifies, and if houses, hotels, office and commer cial buildings have been constructed to the extent (and in fact often farther than is) justified by growth and the requirements o f prevailing stand ards—then investment outlays w ill decline even though consumption expenditures could, in some manner, be stabilized at the boom level. And, if this is so— and the experience o f business cycles fo r 75 years supports this analysis—then the pro ducers o f capital goods w ill not be able to continue to operate at levels adequate to maintain their pay rolls or to pay guaranteed wages to their employ ees. A nd when the market for the capital-goods industries drops out, then a large fraction o f the market for consumers goods w ill also disap pear. This is the situation which cannot be waived by the over-simplified theory which holds that jobs provide pay rolls, pay rolls provide markets, and This includes local community and other governmental serv ices (education, for example) financed from personal taxes. In modern communities this is one important w ay in which con sumers spend their incomes. 90 A ll through this analysis we are assuming, it is to be remem bered, that the guaranteed wage is financed by private industry and not by the Government. I f consumption were automatically maintained by Government outlay, the case would be entirely different. 26 markets in turn provide jobs. The overly sim pli fied version did indeed fit reasonably well into a primitive economy where capital goods played no significant role. And it is fo r this reason, basically, that primitive economies, while poor, nonetheless rarely suffer from unemployment. I f the “missing eighth” continues to be lacking, how can consumer outlay be kept up to the maxi mum level reached in the boom? This is the cru cial question: by what method o f financing is the wage guarantee to be maintained? A nd yet throughout most o f the literature on guaranteed wages, this whole problem is either neglected completely or passed over lightly. T o remedy this defect, the rest o f our chapter w ill analyze in some detail the effects o f different methods o f finance—pay-as-you-go and reserve. I f a guaranteed wage plan keeps a man from being laid off and makes possible the maintenance o f his income, then it is not unreasonable to expect that it w ill result in an increase in his consumption purchases over what they would otherwise have been in the absence o f such a plan. So far, so good—but we have not carried the analysis very far if we stop with the expansionary effect o f guaranteed wage expenditure. I f the Government were paying the bill, not out o f tax revenues, but by the printing o f crisp new dollars or by loan-financing, that would be one thing. The discussion o f the im plication o f such a policy, together with relevant comparisons with other antidepression programs (which might be more justifiable and prom ising), would take us very far afield and far beyond the scope o f our present study. When only private finance is in volved, it is necessary to give attention to some important factors which offset the favorable in come maintenance effects, as we shall see. BUSINESS CYCLE S TA B IL IZA TIO N AN D PRO D U CTION E F F IC IE N C IE S RESU LT IN G FROM G U ARAN TEED W A G E S Before entering upon a discussion o f reserve and pay-as-you-go financing, we must digress to consider a point which is given considerable weight in the literature o f guaranteed wage plans. Such programs, it is claimed, often more than pay fo r themselves because o f the production efficien cies resulting from lower labor turn-over, greater labor efficiency, and fuller utilization o f capital. Especially in connection with the ironing out o f seasonal and irregular fluctuations is this likely to be the case. I f the general business cycle is not brought under control, it is more doubtful that the increase in efficiency w ill be nearly as great as the costs o f maintaining income payments in the face o f conditions which fluctuate independently o f the firm’s control. However, in any case there may be some sav ings in cost to offset against the expense o f guar anteed wages. The question w ill inevitably be raised, therefore, as to whether widespread guar anteeing o f wages cannot by itself iron out the business cycle, or at least maintain income steady throughout the cycle, and do this without any net cost to the business community. The remaining paragraphs in this section aim to throw light on these questions and clear the ground for the anal ysis o f reserve and pay-as-you-go financing. I f the savings in cost through increase in effi ciency were very important, the individual em ployer would be helped to finance his guaranteed wage, but it would not necessarily solve the prob lem o f unemployment for the economy as a whole. Increased labor productivity, meaning among other things that less employment is provided by a given volume o f output, would raise the same problems o f unemployment and purchasing power as technological inventions. About this there is a large literature, the consideration o f which would lead us rather far beyond our immediate task. The mere reference to this literature is, however, sufficient to show that in the absence o f a positive full employment program, it cannot safely be as sumed that technological and other developments designed to increase worker productivity, would automatically increase and stabilize aggregate em ployment and aggregate demand in the economy as a whole. I f the guaranteed wage could induce an increase in productivity sufficient to cover its cost, this would be a great social gain leading, as with all advances in efficiency, to a potentially higher real incom e; and in an expanding economy in which technologically displaced labor quickly became reabsorbed, this potential would actually be real ized. But the mere fact that a guaranteed wage plan or an invention “ pays fo r itself,” w ill not alone tend to reduce the fluctuations o f the cycle. There is no evidence from past experience (nor could it be expected on grounds o f general anal ysis) that gains in productivity tend to have this effect. Nevertheless, it is possible fo r the cost-savings o f guaranteed wages to be used in such a way as to promote stability over the cycle—that is, by in vesting the financial cost-savings in good times into a reserve fund out o f which payments are made in depression. This leads us into the whole problem o f pay-as-you-go and reserve financing. A fter we have analyzed these devices, we shall return to efficiency cost-savings.2 1 TW O M ETH ODS OF FIN AN CE There are two methods by which widespread guaranteed wage plans could be privately financed. First, there might be some sort o f reserve plan, according to which funds would be collected during the expansionist boom phase o f the business cycle; and then during the contractionist depression phase o f the cycle, these funds would be dispersed in the form o f guaranteed wages. This same reserve principle is already in effect in connection with our Federal and State unemployment compensation program. The best case could be made fo r the anticyclical reserve method o f financing if our economic system could be regarded as suffering only from cyclical swings above and below some “normal high-employment” level. The second general method o f financing guar anteed wages, is that o f so-called pay-as-you-go. In contrast to the anticyclical reserve plan, in any given year the costs o f guaranteed wages would be approximately balanced by current charges against sales or business gross income. Such a plan would involve relatively small payments in the boom years and therefore relatively small costs. In depression years, however, the pay ments would be large and the redistribution o f current income would be considerable. The ques tion to which we wish the answer is how much the payment o f currently financed guaranteed wages might so redistribute income as to raise the level o f consumption expenditures (or in more technical language the consumption function) and thus in crease total aggregate outlay beyond what it other n See below, footnote 2 6, p. 445. 27 wise would have been in the depression period. How would the cost be distributed between busi ness net savings, wages, and consumer prices ? P A Y -A S-Y O U -G O FIN AN CE Except for the earlier discussion o f any possible induced increases in productivity, pay-as-you-go financing necessarily involves a process o f redis tribution o f income between the various claimants to the national income—a redistribution which would fall mainly in the depression phase when large guaranteed wage payments would be made. How such a redistribution affects aggregate de mand is one o f the most difficult questions in the whole field o f economic analysis. A t the present time, there does not exist sufficient theoretical and factual knowledge to permit a confident or pre cise answer. It involves the same perplexing problem as that o f the incidence o f the taxes raised to finance our whole social security program ; or for that matter, o f the incidence o f corporate and business taxes in general. As we have already noted, we assume that any increase in productivity induced by the guaran teed wage has already had its effect in total real income. The problem to which we address our selves, therefore, is the redistribution effect o f the guaranteed wage. Three extreme positions can be stated, with the truth probably lying somewhere in between: (1) The cost o f guaranteed wages is borne completely by profit-receivers through re duced earnings o f their properties; (2) the cost is shifted forward completely to consumers; (3) the cost is shifted back onto workers, who lose in wage rates what they gain in steadier income. A survey o f the literature on social security tax ation would show widespread doubt that most o f the increase in pay-roll costs could, except in the shortest run, be at the expense o f corporate profits. Economists who have written on this subject seem rather to incline to the belief that the primary in cidence would be on wages or on consumer goods prices. W e would be rash to attempt a pro nouncement on this difficult quantitative question, since not even the two o f us would be in perfect agreement as to our rough guesses. In depression when profits are already low or negative, the heavy extra costs o f guaranteed wages are perhaps even less likely to fall in large part on capital. This conclusion is reinforced 28 when we remember that guaranteed wages are to be superimposed on a system o f developed collec tive bargaining, where already labor is attempting to keep wages high and is already taking some advantage o f management’s ability to pay. But let us first, fo r the sake o f the argument, suppose that the cost o f guaranteed wages does fall largely on profits. Then wage income w ill be larger, and dividends smaller. Since the bulk o f dividends does not go to poor widows and orphans, there w ill probably be some favorable effect upon the propensity to consume. But the extent o f this should not be exaggerated. The difference between the average propensity to con sume 2 (and to save) o f the rich and poor is known 2 to be very great. But the available evidence from governmental fam ily budget studies seems to sug gest that the marginal propensities to consume (and save)—and it is only these which are rele vant— differ less sharply between rich and poor than do the average propensities. Dollars redis tributed from wealthier common stockholders to workers add something to total consumption pur chases, but less than is usually supposed. Never theless, to the extent that such a transfer is feasible consumption is thereby increased. Not all corporate profits are paid out in dividends. Some go to that other silent partner o f business enterprise, the government. Con ceivably a shift from profits to wages might lower tax revenue and, other things being equal, con tribute to larger depression loan financing and so to a slightly higher level o f over-all expenditure; in other words to a slightly higher level o f effec tive demand. It should, however, not be hard to find other more obvious ways o f expansion if de pression and unemployment call for expansionary fiscal policies. A considerable part o f profits goes into net cor porate saving, representing undistributed profits or additions to earned surplus. Part o f the in crease in guaranteed pay rolls might be expected to come out o f such net corporate saving. (This is true even in deep depression, when net corporate saving may be negative if dividends are being paid out in excess o f current earnings; because o f guaranteed wages the figure might become 22 “ Average propensity” refers to the proportion of income con sumed (and saved). “ M arginal propensity” refers to the propor tion of any sm all additional increment of income which is con sumed (and saved). algebraically more negative.) A n increase in the payment o f wages at the expense o f corporate net saving would, taken by itself, tend to raise the level o f total effective demand. But this con clusion would be true only on the assumption that there would not at the same time be adverse effects upon the corporation’s inducement to invest. In time o f deep depression, most firms do not even replace much o f their depreciating equip ment. Yet they still, on their income statements, include charges to depreciation reserves as an item o f cost to be subtracted from gross earnings in ar riving at net profit. But these depreciation charges are expenses that, at the moment, involve no outlay. It would therefore be wrong to con clude that a firm which was suffering business losses was necessarily flooding the community with purchasing power. A more meaningful measure o f a firm’s expansionary or contractionist in fluence, taking all factors into account, would be the difference between its gross investment ex penditures (including replacement outlays) and its gross saving (including depreciation); or, what comes to the same thing, the difference between its net investment and its net saving. W ith net and gross investment already cut to the bone, in a period o f depression a further re duction in earnings (and net saving) might not have any very significant effect upon net invest ment. But in the first years o f recovery from a depression, or during the first years o f descent into depression, there might be some counter balancing adverse effects upon investment—al though not necessarily dollar for dollar. These adverse effects might arise ex post simply because in an imperfect capital market many firms must depend upon their own funds for their investment outlays. Probably more important would be the effects on expectations aroused—the adverse effects upon investment decisions at all phases o f the cycle. During the boom, businessmen will be somewhat wary o f undertaking certain doubtful marginal investments if they expect to be unable to cut their pay roll when times become bad. The harmful total effects upon the average level o f net invest ment over the whole business cycle could conceiv ably be greater than the quantitative reduction in net business saving. Or less. No one can say. W e may sum up the case usually considered most conducive to an expansionary effect o f guaranteed wages—where the costs fall completely on p rof its—by saying that stimulus to the level o f pur chasing power and effective demand is o f somewhat doubtful quantitative magnitude over the business cycle as a whole. In depression times, the quan titative effects may be somewhat more important, but still should not be exaggerated. Let us turn now, more briefly, to the more realistic case where a considerable part o f the cost o f guaranteed wages is assumed to fall on wages or upon the consumer’s real income. I f only seasonal or short rhythms o f particular industries are being considered against the background o f a full-employment society, such a method o f meeting the costs may lead to a more nearly optimal system o f pricing and use o f economic resources. But from the standpoint o f business cycle fluctuations, such a method o f bearing the costs o f guaranteed wages is fairly neutral in its over-all effects upon production and (real) effective demand. I f the cost falls upon wages or raises consumer goods prices, then income is being shared by those who are employed with those workers who would be without pay if it were not for the guaranteed wage. I f such benefited workers are not already in high-paid industries and do not already have high current incomes (apart from the guaranteed wage) this whole process may involve some re distribution o f income from average income re ceivers to those who are more needy. Just as with schemes for sharing work through reduced hours, this redistribution might result in some slight in crease in the propensity to consume for the com munity as a whole. However, the point made earlier concerning the smaller difference between marginal propensities o f different income classes would apply here with even greater force—since the difference between incomes is in this case relatively small. The guaranteed wage would amount, in these circumstances, to a sharing o f the income between those productively employed and those unem ployed. In other words, those who are earning incomes at productive employment would be de fraying the cost o f the guaranteed wages. This is all the more true in view o f the fact that under depressed conditions, profits could cover no appre 29 ciable part o f the cost. Thus, the guaranteed wage, under these circumstances, would in fact be a “ share-the-income” program. In place o f a “ share-the-work” plan, we would have a “ sharethe-income” plan. There would occur a shift o f income among the wage-earner and consumer groups with no sizable effect upon the aggregate propensity to consume. W e may conclude, therefore, that a pay-as-yougo financing o f guaranteed wages which does not fall on profits would result in little change in the level o f income determination at the different phases o f the business cycle. W ith real investment fluctuating over the cycle, real total consumption (all groups in the economy considered) would not remain stable, and real income would still fluctu ate more than the absolute variation in net invest ment. Only to the extent that the cost can be saddled on to corporate profits, and more specifi cally upon corporate saving and not upon invest ment, w ill the guaranteed wage plans have any appreciable expansionary effect upon the level o f over-all production and employment. This does not quite exhaust the analysis o f guar anteed wage financing on the total level o f effective demand. As yet no mention has been made o f the possibility that the enhanced security engendered by guarantee plans might cause workers to save less over the cycle, and to purchase with their sav ings more durable goods such as autos and houses. This is a possibility, and to the extent that it is effective, it would work toward some strengthening o f the average level o f effective demand. Can a similar case be made out for the proposi tion that the guarantee device w ill cause employers to cast about fo r new productive activities and thereby raise the average level o f employment? As fa r as a single concern is concerned, there may be something in this. I f a locomotive or other capital-goods producer were to take on a consumers-goods line, such as breakfast food, he would no doubt smooth out the variation o f his employment and production. But fo r the economy as a whole, little or nothing has been gained. The average income elasticity o f demand and the over-all fluc tuations in income are the same as before. Simi larly, i f an auto company, to make work fo r its own employees, cancels subcontracts with sup pliers and begins to make its own parts, there is no over-all gain. I f the beer industry, by heavy 30 advertising, gains at the expense o f the bakery or confectionery trade, the over-all situation is the same. Only too often, the devices which seem to stabilize one concern’s production are at the cost o f somebody else’s destabilization. The only true exceptions, from a business cycle point o f view, are policies which involve additional net investment fo r the community as a whole, or an expansion o f consumption standards. “ P ro ducing fo r stock” is one stabilization device fre quently mentioned by writers on guaranteed wages. Such a policy does involve net investment. How important can this factor be over the busi ness cycle? The answer is somewhat discourag ing, except fo r brief cycles. Production fo r stock might in 1930 have kept employment from slip ping off as rapidly as it did. But such positive net investment in inventories is only too likely to result subsequently in commensurate disinvest ment, making 1932 and 1933 even more depressed. A brief recession in business believed to be tempo rary might be bridged by this device. But a sharp inventory recession o f the 1921 or 1938 type is likely to be accompanied by so much price un certainty as to render this device relatively ineffective. W e must also not lose sight o f the fact that many o f the policies o f businessmen that lead to larger total investment and employment do not give rise to much employment within their own plants. When railroads order new equipment, the extra jobs are not provided fo r railway labor. When a merchant or manufacturer “buys for stock” the task o f guaranteeing jobs is eased for his suppliers but not fo r himself. It is true that he is some body else’s supplier, and that if all together pursue expansionist policies, all together may benefit. But it is not less true that anyone who holds back w ill also benefit from the expansion o f the others; and it is somewhat Utopian to rely upon collusive decision-making by business as a body or upon altruistic social planning by isolated businesses. The figures o f seven-eighths consumption and one-eighth saving are given by way o f illustration. They should not be interpreted literally, especially since in many downswings o f the business cycle, autonomous net investment (caused by advances in technique and the like) w ill not disappear com pletely. Less than one-eighth o f the full em ployment income level would then have to be made up by guaranteed wage programs in order to keep total consumption stable. There is also the further possibility that some appreciable part o f net investment expenditure is not “ autonomous” but is “ induced” by, or is de pendent on, the level o f consumption or income. According to this type o f argument, if guaranteed wage plans succeeded in keeping consumption at a higher level, then an appreciable amount o f net investment would be induced (over what there would otherwise have been).28 The weakness in this argument stems from the fact that a given stationary level o f national in come, no matter how high it is, cannot on balance be supposed to include any continuing net—as distinct from gross or replacement—investment. It is only the change in the level o f income that induces net investment. When the stock o f capi tal has become adjusted to any new level o f in come, no matter how high, net investment would again disappear 2 —in the absence o f technological *4 changes and interest rate variations. These rather long-run considerations w ill not seem decisive to the adherent o f guaranteed wages as a business cycle cure. He w ill ask, “ Isn’t the business cycle a short-run phenomenon? W ould not the maintenance o f consumption lead to in duced investment (over what it would otherwise have been) for at least the few years that a depres sion usually lasts?” The answer to both o f these questions may well be in the affirmative. But this does not mean that the temporary, once-and-for-all, induced invest ment effects o f consumption maintenance can be regarded as raising the average level o f demand and employment over the whole cycle. W hat is 28 In technical terms, some economists would assume a "m argi nal propensity to invest” (out o f income) as well as a marginal propensity to consume. T his would tend to increase the "m u lti plier” effects of a given increase or decline (as the case may be) in autonomous investment or autonomous consumption. But so long as the two marginal propensities add up to less than one, then guaranteed wage plans will incur some over-all losses (payments for idle labor) whenever autonomous investment declines. How ever, those writers who assume large marginal propensities to invest and large multipliers, seem to imply that the historical changes in autonomous investment (the "m ultiplicands” ) were correspondingly smaller— in order to lead to the same observed fluctuations in total national incomes. Therefore, to such writers the "losses” resulting from guaranteed wage plans would tend to be correspondingly smaller. 24 This situation means that there are no changes in technology involving new processes o f production or new products for con sumption. Changes in methods of production or in products would often require new autonomous investment, which would cause a rise in national income. gained in investment during the depression is likely to be at the expense o f investment during subsequent periods. By virtue o f induced invest ment effects, guaranteed wage plans may have some stabilizing effects on the cycle—by expanding de pression demand at the expense o f the follow ing boom-time demand. In this it resembles reserve financing, which we are about to discuss, with the difference that reserve financing usually fills in the trough o f depression at the expense o f the previous rather than the subsequent peak. But neither method is likely to expand demand appre ciably over the cycle, unless Government loan financing or private corporation net savings are used to finance the guaranteed wage plans. More over, as we have seen, any favorable induced investment effects must be set against the unfavor able effects upon net investment and venture cap ital resulting from the recognition by potential risk-takers that guaranteed wage plans are likely to increase their costs and risks. One final point favorable to the guaranteed wage. Over time, the consumption standards o f people increase as people get used to higher levels o f living. Any success achieved in maintaining high real incomes tends to breed further success. Unlike induced investment effects, “induced con sumption” effects may be permanent rather than transient or once-and-for-all. Once employment is guaranteed, labor becomes in the nature o f an overhead cost rather than a variable cost. W e should at least explore the pos sibility that this might lead to a pricing policy over the cycle which would make for greater sta bility o f production. The beginner in economic theory is taught that really fixed costs should, at least in the short-run, be disregarded in output decisions and price form ation; any business should not be refused which can at least earn something more than variable costs. I f a large part o f labor is no longer a variable cost, this might permit downward flexibility o f prices to rather low levels. In more advanced courses in economic theory (particularly those dealing with so-called “ ad ministered” prices and monopolistic competition) we are taught that “ fu ll” costs (including both overhead and variable costs) actually have an im portant influence on price form ation, and this more sophisticated analysis appears to be in line with business practice and business opinion. In all like 31 lihood, this pattern o f “ full-cost pricing” tends to characterize those semi-monopoly fields where industry is likely to guarantee wages. Therefore, it is something o f an academic exercise to inves tigate the effect o f flexible downward prices which abstractly might be related to the conversion o f labor costs from a variable to a fixed cost. Twenty years ago most economists would prob ably have considered such downward price flex ibility a highly effective way of meeting depres sion. Today we are not so sure. The modern economy (given our present-day institutions) does not seem to go through price deflation painlessly or easily. Still there is this to be said for the kind o f downward price flexibility referred to above in connection with the guaranted wage. Such price reductions would not, it is assumed, be accompa nied by wage reductions or any substantial decline in worker income, so that one o f the worst aspects o f a deflationary spiral would not be present. Still, increasingly it is the modern viewpoint that the price elasticity o f demand is much lower in de pression periods than used to be thought the case. Consequently, we should think it unwise to at tach too great quantitative importance to the phenomenon under discussion even if it were like ly—which it is not—that prices w ill show the proper cyclical flexibility rather than perverse flexibility. R E SE R V E FIN AN CIN G W e must now analyze the possible effects o f re serve financing. This method has most signifi cance, and is most likely to be resorted to, in a so ciety subject to pronounced cyclical swings. I f, on the contrary, employment is reasonably stable, whether at a high level with low unemployment or at a relatively unsatisfactory level with chronic unemployment, one year is very much like another, so that there can be little shifting o f the financial burden over time. A gradual and experimental introduction and extension o f the guaranted wage plan might be financed in part on a self-liquidating basis. A gradual expansion o f the program as experience justified might be accompanied by cost reductions which could “ carry” the reserve plan without re sort to higher prices or wages lower than could otherwise be paid. On this basis, it is likely, how ever, that progress toward the guaranteed wage 32 would be slow and would probably be largely re stricted to firms whose employment experience is in any event favorable. I f anything substantial were going to be achieved, it is probable that there could be no escape from a program involving, for the great m ajority o f firms, the accumulation o f reserves financed out o f profits, wages, or higher prices. W ithout substantial reserves, the program would be discredited by a wave o f defaults at the first onslaught o f any considerable volume o f mass unemployment. An examination o f the experience o f various firms in the 1929-34 depression reveals how costly different guaranteed wage plans might have been. The high employment commitments o f the late 1920’s would have entailed heavy cost outlays all through the early and middle 1930’s. I f guaran teed wages had then been in existence, in a real economic and accounting sense, the men necessary to produce the high 1929 outputs would frequently have cost the companies in the thirties more than the total wages actually paid out to them at that time. Prudent bookkeeping would have sug gested that this expense be recognized in 1929 as part o f the cost o f that year’s output, and that funds be set aside to meet this contingent, but nonetheless necessary, deferred expense. Accordingly, if we mean by a guaranteed wage program something more than a cautious and tentative introduction by a few firms with highly favorable employment experience, we must en visage a substantial program o f financing—in other words, the accumulation o f large guaranteed wage reserve funds. Now it is just this problem which we need to face in assessing the feasibility o f the guaranteed wage plan, unsupported by governmental financial assistance. How would the accumulation o f large guaranteed wage re serves affect the functioning o f the economy? An analysis o f the impact o f the guaranteed wage system (financed on the reserve basis) upon the flow o f income and expenditure involves a con sideration o f the manner in which such a program withdraws funds from the income stream on the one hand, and pours funds into the income stream on the other. I f we envisage a highly fluctuat ing society, and if we concentrate attention for the moment exclusively on cyclical unemployment, it follow s that a guaranteed wage reserve plan w ill, for the economy as a whole, withdraw funds from the income stream in boom years, and pour back approximately equivalent funds in the de pression years. The same problem o f “ incidence” o f the costs arises here as under pay-as-you-go; but the tim ing is different. I f we assume that the annual charges to the reserves are financed largely from higher consum er prices or lower wages, it follow s that the level o f consumer expenditure would be less than other wise in the boom years, and higher than otherwise in the depression years. The “ reserve method” lops off the boom and fills in the valley o f depres sion. It is a method o f redistributing expendi tures over the time span o f the cycle. The reduc tion o f expenditures in the boom period might re duce employment, or it might merely have the effect o f holding down price inflation. In technical terms, the accumulation o f reserves during prosperous periods tends to exert a de flationist influence at that time by reducing the effective propensity to consume and increasing “ average thriftiness” or the effective propensity to save. The spending o f the reserves in the de pression period exercises an expansionist effect by raising the effective propensity to consume at that time. Taken as a package, this double-barreled pro gram is socially desirable. But i f the period o f so-called prosperity is like that o f 1936 or 1937, in other words not one near to fu ll employment, the deflationary influence o f reserve accumulation is, taken by itself, a factor reducing employment, consumption, and in all likelihood capital form a tion as well. However, if the prosperity period is one characterized by manpower shortages, by too much effective demand, and a tendency toward inflation, then the enforced reduction o f consump tion is all to the good. W e need not treat in detail the mechanics o f the process whereby reserves are accumulated and dispersed, because this problem has been widely discussed in the literature on social security. I f guaranteed wage reserves consisted simply o f building up o f cash or bank deposits in good times and their dispersal in bad times, the process would be relatively simple. The quantity o f money would not necessarily have to undergo any changes but its rate o f activity or velocity would be lessened in good times and increased in bad. I f reserves are treated as trust funds investable only in Government bonds or similar gilt-edge se curities, the mechanics might be slightly different but the final effect substantially the same. As a matter o f fact so long as the Federal Reserve banks or the commercial banks are w illing to buy or sell considerable volumes o f Government se curities, these assets possess many o f the liquidity properties o f money. I f, in depression, the bonds accumulated in the reserve funds were bought by the banks, then the stimulus to effective demand resulting from guaranteed wage disbursals might appear as an increase in total bank deposits rather than simply as a change in the velocity o f money. In good times, if the growing reserve funds ab sorbed Government bonds held by the banks, the result would be to curtail the volume o f deposits during the boom below what it would otherwise be. Any thoroughgoing discussion o f this problem would have to take account o f the effects upon the interest rate structure o f both the accumulation and deaccumulation o f such reserve funds. Broadly speaking, these secondary reactions upon the interest rate structure and the capital markets are to some degree perverse. The accumulation o f reserves in good times would to some measure tend to raise capital availability and lower interest rates, so that some o f the enforced reductions in consumption would seep over into investment rather than curtailing over-all demand.2 Sim il 5 arly in depression, attempts to liquidate reserve funds might have some upward effects upon in terest rates and consequently, some depressant effects upon investment. However, these second ary effects can be minimized if reserve funds are confined to Government bonds. Moreover, the 25 So long as the funds are not hoarded, there is always the possibility that some o f the purchasing power diverted away from consumption will filter into investment. (It has been argued along similar lines that raising taxes in excess of Government expenditure to retire debt may prove not to be deflationary if the owners of retired Government bonds plough the money back into productive investment expenditure.) In a perfect capital market, where the favorable effects upon investment can only take place through a reduction in the interest rate, this secondary reaction to a primary reduction in the consumption schedule would in all likelihood be quantitatively smaller, dollar for dollar, than the original primary effect. T his for two reasons: the doubtful responsiveness of investment to reduced rates of interest, and the adverse effect on investment outlay of reduced consumption sales. To the degree that the actual capital market is an imperfect one, always subject to informal rationing, investment may react some* w hat stronger to larger loanable funds than if the investment rate alone were involved. T his modifies but hardly alters our con clusion. 33 monetary authorities can easily, and should, take account o f any perverse interest and other effects o f reserve accumulation and deaccumulation and take appropriate action. A new possibility would arise if governmental fiscal policy could be assumed in the years ahead to be effective in preventing inflation when the economic system is operating at full employment. Whenever private investment demand was exces sive, fiscal and monetary policy, on this assump tion, would curtail the excess. In such an environ ment, the widespread accumulation o f guaranteed wage reserves would reduce to some degree the need fo r alternative deflationary monetary and fiscal policies. I f carried to excess, however, such reserve accumulations might have to be offset by actual expansionist governmental policies. There is no assurance that the level o f income and employment over the entire cycle would be any larger under reserve financing. Among other effects, the program would bring about a redistribution o f income, partly from profits, partly from consumers in general, and partly from productively employed workers to workers who, if there were no guaranteed wage plan, would cease to draw wages. I f carried out on a relatively modest scale—one which would not seriously dis tort price-wage relationships—the effect would be to level off more or less the boom demand and to raise the effective demand in the depression period. But if really huge reserves were accumu lated in boom years, adequate to pay annual wages to a large part o f the labor force in a depression as serious as that beginning in 1929, the boom could easily be converted into a depression. Reserves o f the magnitude required would cause a serious distortion in the economy. Some would prefer to state this distortion in terms o f the wage-price structure, others in terms o f a concealed tax on consumers and wage-earners, and still others in terms o f a savings-investment problem.2 * 8 Apart from these theoretical considerations is the practical matter that actuarial calculations o f the unemployment risk are not feasible. The reserves that are built up might prove too large. In this event the accumulation o f reserves would amount to a compulsory savings scheme not simply in the boom years, but also over the entire cycle. Aggregate demand might thus tend to become deficient over the long run and create a condition o f underemployment. (O f course if the long-run factors were inflationary, such compulsory saving would tend to be a stabilizing factor.) Finally, in a grow ing society, the setting aside o f reserves, even though actuarially exact, nevertheless would still result in the growth o f reserves over the cycle, and to this extent, would tend to aggravate the problem o f maintaining adequate levels o f aggre gate demand. On the other side, the inability to make reason ably accurate actuarial calculations o f the risk in volved might result in wholly inadequate reserves. Unemployment insurance reserve funds, though they undertake a very limited liability compared with a comprehensive guaranteed wage program, have had to be supported in various countries by the Government treasuries. W hile large reserves have been built up in the State unemployment funds in the United States, a serious and prolonged depression could easily bankrupt many o f them. A guaranteed wage plan, even one limited to 40 weeks per year, could be considerably more costly than unemployment insurance. The reserve funds, to withstand a serious depression, might have to be far larger than those needed for unem ployment insurance. Put to the test, guaranteed wage plans might need Government support even more than unemployment insurance. Thus, what was intended to be a scheme based on private industry financing might face default under the stress o f serious depression, or else be compelled to call upon Government support. This risk * I f technological cost savings from the guaranteed wage plan (assuming for the moment that such plans do increase man-hour productivity) were placed in a reserve fund during the boom phase of the cycle and paid out in the depression phase, the effect would be a redistribution o f income over time. The total income payments made to individuals in the boom would be made less than would otherwise have been the case ; and during the de pression the income payments would be increased. Financed as here assumed, out o f increased productivity, there would result no redistribution o f income between individuals, and no absolute re duction in corporate profits or corporate net savings. I f how ever, the gains in productivity were nil or inadequate to cover the guaranteed wage payments, the added costs would have to come either out of corporate net savings or else from a redistri bution of the income of individuals. Only the reserve method o f financing is capable of redistribut ing the income over time, i. e., from boom to depression. Such redistribution over time has nothing to do per se with the ques tion where the funds come from , whether fr o m : (a ) gains in productivity or (b) redistribution of income between individuals and social groups. I f the reserve method of financing is not used, gains from productivity will necessarily be passed on in the form of (a ) larger income payments, (b) increased outlays on plant and equipment, or (c) larger corporate retained earnings. But none of these, apart from reserve financing, would tend to iron out the cycle. 34 serves to stress all the more the importance o f lim iting the employer’s liability, as suggested in chapter V I below. It would not be possible to calculate with any high degree o f accuracy the actuarial risks involved in introducing guaranteed wage plans in different industries even in a society committed to a program o f sustained full employment, and prepared to use fiscal, monetary and other measures to achieve this goal. A number o f countries, including England, Canada, and Australia, have issued white papers on employment policy proclaiming high and stable levels o f employment as a deliberate aim and responsibility o f Government. The United States Employment A ct o f 1946 does not go quite that far. A ll the British countries referred to above have recognized the difficulties involved, and none is confident o f anything more than moderate suc cess. Thus, even with the best o f good w ill and favorable circumstances, no one knows what degree o f success may attend these plans. W e know that the war has left us the heritage o f innumerable distortions, and these w ill plague us fo r many years to come. Let us suppose, however, that we have surpassed this difficulty—that experience has, in fact, demon strated a high degree o f success with a stabilized full-employment program for a period sufficiently long to warant general confidence. Some fluctua tions there would nevertheless be, but more moder ate and presumably fairly well under control. Cyclical fluctuations o f employment would be m ild, yet relatively large in the heavy capitalgoods industries. Seasonal and transitional un employment would also be experienced. In the “ ideal” society described above (and we repeat that we have no illusions that such a society can quickly be realized) guaranteed wage plans, privately financed, might tend to moderate any re maining cyclical swings, and to facilitate an at tack on seasonal unemployment. Transitional and frictional unemployment might be intensified if the guaranted wage had the effect o f reducing labor mobility. W ithin the framework o f the society envisaged above, cyclical unemployment might become a more or less manageable risk, and accordingly, the reserve funds needed to meet the financial lia bility involved would not need to be large. In this event there would develop no long-run prob lem o f excessive compulsory saving in the form o f reserve funds. The reserves accumulated in the good years would be dispersed in the rela tively low years. This leveling process would tend to smooth any remaining swings in the cycle. A buoyant labor market, such as that here as sumed, would moreover be favorable to an attack on the problem o f seasonal unemployment. Em ployers would be stimulated by the introduction o f the guaranteed wage to find year-round work fo r their employees. I t is one thing to place such a responsibility upon employers in a society such as the one we are living in, with its grave uncertainties with respect to the stability o f the market and the adequacy o f aggregate demand. It is quite a different matter to introduce the guaranteed wage in seasonal industries in a society in which, within reasonable limits, the adequacy o f aggregate demand is assured. The develop ment o f supplementary lines is feasible in an ex panding and buoyant society, whereas it might only supplant the product o f some other producer in a limited market in which there was not work enough to go around. Seasonal unemployment might be largely eliminated in a full-employment economy. But even so, this important goal could not be reached unless ingenuity and resourceful ness were brought to bear on the problem. The guaranteed wage might offer a powerful financial inducement to employers to find steady year-round work fo r their employees. The technical job o f dovetailing short-run fluctuations in employment is no mean managerial task, but a sustained high level o f aggregate demand would provide an en vironment within which the job could be attacked with high promise o f success. W e have spelled out these various prospects o f the guaranteed-wage program within the frame work o f a society which had achieved a high and stable level o f aggregate demand, not because we think such a society is going to be reached tomor row or the next day, but rather to lay bare the difficulties confronting the guaranteed-wage pro gram. The conclusion is not that the guaranteed wage must await the dawn o f a perfect society. The conclusion is rather that we must take account 35 o f its limitations, and especially avoid procedures that may defeat the very purposes which it is intended to achieve. Having in this chapter analyzed the effects o f guaranteed wages upon the problem o f busi ness-cycle unemployment, we turn in the next V. GUARANTEED WAGES AN The problems to be discussed in this chapter are o f a rather abstract and general nature. Nevertheless, the theoretical analysis involved, even though briefly stated in this report, is, we believe, important in an over-all consideration o f the guaranteed wage. T H E M E R IT-R A TIN G CON TROVERSY In chapter I I it was shown that the economic cost o f idleness to the community could be pri marily measured by the loss o f output which it entails.2 Suppose that a particular industry or 7 line o f activity has a record o f great instability o f production and employment. How should its prices and costs then be set? This identical question was involved in the ex tensive discussion o f merit (or experience) rat ing in connection with unemployment compensa tion. The original Wisconsin Unemployment Insurance A ct, drawn up under the strong in fluence o f John R. Commons and economists o f his school, required employers with unusually poor records o f stability to pay higher rates o f taxa tion. Those demonstrating unusually good con tinuity o f employment were granted lower rates— in much the same way that steel-frame buildings, relatively immune to fire risk, are granted low fireinsurance rates. Two principal considerations are usually in voked to justify this procedure. First, and per haps o f least importance, is a feeling that fairness (justice or equity) implies that an employer should have to pay only for the evil which he creates, but that he should be made to pay for that. This particular argument is a highly con troversial one. Numerous writers, perhaps even the m ajority, have dissented, arguing that it is the essence o f social insurance (in contrast to private ” Our earlier discussion also dealt w ith the additional evil arising out of the effect o f idleness itself on the worker and com* munity, and also upon the problem o f equitably sharing the costs involved in idleness between different individuals and groups. 36 chapter to a consideration o f its influence upon the pricing system as it relates to the efficient use o f economic resources. Even if the problem o f depression were brought under control there would still remain the challenge o f seasonal and other irregularities in production. EFFICIENT USE OF RESOURCES insurance) that there be a sharing o f burdens be tween groups; the essence o f merit rating, on the other hand, is the perverse principle o f “ to him that hath, shall be given.” The second argument in favor o f making the industry bear the costs o f its own instability has usually been the more important one. Under such a system, management is supposed to be encour aged to stabilize its operations. It is financially penalized if it does not stabilize; it is rewarded to the extent that it does. There is certainly some thing to this; as we saw in our earlier discussion o f the present wage contract, labor appears to each firm as an avoidable variable cost. Current prac tices therefore tend to warp and distort the socially desirable adjustments o f production. W e do not wish to enter upon the details o f this controversy. It is enough to note that the oppo nents o f merit rating point to empirical data to show that employers are typically not able to do much about their own instability. Experience, it is said, seems to indicate that employers’ reactions to higher penalty rates are not sufficiently sensitive (particularly in response to the rather moderate differences in rates usually involved in merit rat ing) to be very significant. Further, it is argued that employers with steady records are not usually better planners but are more often the lucky bene ficiaries o f special market situations (often o f a monopoly typ e). This economic controversy finally burned down in the years after 1935 when the Social Security A ct was passed. The theoretical honors, if one may judge from the literature, seem to have gone against merit rating. Nevertheless, in recent years merit rating has been introduced in modified form by almost all States, principally as a one way street making fo r reductions in taxation and tending to keep unemployment reserves from growing so rapidly during the high-employment wartime years—that is, during precisely the time when such reserve accumulation might contribute toward the fight against inflation and excess aggre gate demand. This, and not the establishment o f effective differentials spurring planning for sta bilization, has become the primary function o f merit rating. BUSINESS CYCLE VERSU S O TH ER ID LEN ESS When we turn to guaranteed wages, which is equivalent almost to 100-percent unemployment compensation borne completely by employers, what was previously a rather stale academic exer cise becomes a really vital economic problem. Should the form o f the wage contract be changed so as to raise, substantially, the costs o f production o f unstable firms? A quite different answer must be given to this question depending upon the kind o f instability involved. W e saw, in the last chapter and earlier, that a typical business concern cannot really do very much by itself to moderate the cycle, and often what it can do represents no corresponding net gain for society as a whole but rather only changes the place where instability appears. W e also saw that widespread guaranteed wages by all employers together are scarcely likely, taken alone, to prove a quantitatively important stabilizing factor in an economy subject to wide swings o f investment, which in turn have a cumulative effect on income and employment. The John R. Commons view that the price o f each commodity should be made to bear the cost o f the idleness which it involves—we do not say causes—cannot validly, we believe, be applied to unemployment o f the business cycle type. In this chapter, we wish to consider whether or not a similar conclusion would hold with respect to un stable industries which operate in a continuing high employment economy; or to unstable indus tries which operate in an economy whose general over-all activity is steady but at a relatively low level involving considerable chronic unemploy ment. A t the risk of being wearisome, let us repeat once more why the problem o f business cycle unemploy ment differs from that o f seasonal or other ir regular rhythms. The unstable sector o f our econ omy is predominantly the capital goods industries. To argue that costs and prices should be raised in such industries flies in the face o f certain com monly accepted programs for fighting depression— programs which stress the need to reduce construc tion costs, interest rates, and capital costs in gen eral so as to stimulate investment outlays. It would be bad social accounting to discourage em ployment in the capital goods industries by sad dling additional costs on these industries. Such a procedure is only too likely to reduce the total vol ume o f capital goods outlays over the cycle, thereby leaving the system at a lower average level over the entire cycle. The same matter may be put in still another way. When unemployment is the alternative to an industry’s employment o f a worker, the true social (opportunity) cost o f hiring the worker is nil. Let us suppose that we undertake to saddle—in the form o f an overhead—the financial cost o f an idle worker upon the industry in question. I f indeed, it were true that the capital goods industry were characterized by a high degree of perfect competi tion, in the short-run these increased overhead items would not cause any increase in the selling price o f the product (a capital good) since under perfect competition the selling price would be forced, in periods o f falling demand, down to a level below full unit costs. This means that sad dling the extra labor cost (in the form o f an over head item) need not necessarily result in curtail ment o f output. Instead, it would cause the shortrun burden to fall heavily upon the industry’s stockholders. This might well be better than hav ing it fall upon the workers, but is not necessarily better than spreading the burden of depression more widely on the community as a whole (assum ing for the moment that the Government does not act to prevent depression). The above result would hold only for a highly competitive industry and then only for the short run. In the longer run, firms would be forced out o f the industry as a result o f chronic losses (in cident to saddling the industry with extra labor cost in the form o f a guaranteed wage overhead) until enough firms had left to cause output to de crease and prices to rise to the level o f fu ll aver age cost (including overhead and taken over the whole business cycle). But such increases o f the price o f capital goods, as we have already seen, tends to reduce total employment. I f the industry is one characterized by consid erable imperfection and by administered prices 37 set by the firm, we might still have short-run ruinous competition. But more likely the firms w ill be aware o f their competitive interdependence in pricing and o f their monopoly power to raise prices without losing any considerable part o f their customers. Full cost considerations are likely to provide a powerful rallying ground for the firms in the industry, so that, even in the short run, prices might rise. In the long run, the same result as already indicated would follow , although we must also be prepared for the possibility that so-called monopoly profits may be made to bear part o f the cost. In the absence o f effective Government fu ll em ployment measures,2 it would not be good policy 8 to permit the whole scale o f capacity and labor force in the capital goods sector to shrink to a point where the average price over the cycle would cover fu ll costs including the guaranteed wage overhead. This might only make the saving and investment problem worse. And yet, it is patently unfair to let workers (attached to the capital goods trades and who have no other place to go) receive lower earnings over the cycle than other workers. One obvious solu tion would be to let the Government make up the difference in relief payments or social insur ance, it being manifestly undesirable to permit capital goods prices to rise over the cycle. How ever, if the policy o f Government aid is to be fo l lowed, it becomes rather obvious that the Govern ment might better get something for its money in the way o f useful public projects. Thus we are brought to a planned program o f public expendi tures to fight depression; or if the people demo cratically feel that high priority should be given to private consumption activity, the logical policy implication is Government transfer expenditure (or tax reduction) which w ill increase private consumption and thereby provide job opportuni ties outside o f the construction and capital goods industries. And just as taxes should be raised 88 W e deliberately refrain from discussing the problem of letting wages (and prices) fa ll in an attem pt to reduce the surplus sup ply of unemployed labor. Such a discussion would be highly technical and would involve a volume by itself. Except for a sm all fringe, the m ajority of writers dealing with this problem in the la st decade are o f the opinion that cutting wages and prices in a hyper-deflation spiral is either undesirable or unfeasible (or both) in a modern capitalistic country. I t w ill be clear from our emphasis that we believe that any alleged advantage o f such a deflationary process could better be achieved by expansionary policies. 38 during inflation and Government expenditure carefully trimmed, so in lean times tax reduction is in order, while useful governmental investment and public services w ill need to be expanded. CON DITION S OF CHRONIC U N EM PLOY MENT But what if there is no effective fu ll employment program and the economy operates at a steady level with considerable unemployment? H ow does the view that the prices o f unstable industries should be made to bear the costs o f their instability then stand up ? It w ill be clear from the discussion above, that the answer must be “ very poorly.” In effect we are asked to place a sales tax on the capital goods industries in order to keep up the income o f the unemployed. This is very nearly the worst type o f financing that anyone could devise. O PTIM A L PR IC IN G U N DER H IG H EM PLOYM EN T CON DITION S When we deal with a world o f adequate aggre gate demand and generally high employment, however, the John R. Commons reasoning comes into its own. Then any industry with great seasonal variability or with some other irregular pattern w ill be withdrawing the labor it hires from other productive employment.2 9 In a high employment economy any industry that draws resources to itself and then uses them only part o f a year ought properly to be charged with the cost o f those resources. This industry’s output must bear the responsibility o f causing re ductions in output elsewhere, equal in value to the productive resources taken away from all other industries. Even if an industry is unable to stabilize its operations, it should still be charged up with its full costs, inclusive o f idleness created by it. Otherwise too many resources w ill be going into that industry, its price w ill be too low, and the con sumer o f the goods in question w ill be subsidized at the expense o f the group least able to afford the 89 These remarks are not meant to apply to those seasonal in dustries which hire workers who come for limited periods Into the labor market, from schools and from homes. Such industries perform a highly useful function in providing short-period em ployment at ju st the season o f the year when there occurs a temporary increase in the labor force. burden—the workers subject to intermittent in come. Higher costs are more than a whip to the em ployer; they are a device for contracting an unstable line o f activity. I f it is unavoidably necessary for a municipal fireman or an auto worker to stand by in idleness much o f the time, we must redefine our concept o f productive labor to include such necessary “ ac tivity.” A t the same time we must tender to the auto worker what we already concede to the fire man, adequate pay fo r all working time that he is unable to perform elsewhere. Undoubtedly the gradual carrying out o f a pro gram such as is implied in the above remarks would result—gradually but surely—in some drastic quantitative changes in the present pattern o f prices and output. So much the better, i f it is true that the present-day structure o f prices is grossly distorted away from an efficient and equi table pattern o f resource use. Some people may grant that an industry should be charged with the fu ll cost o f the resources it takes away from other industries, but still argue that higher hourly wage rates rather than a longer guaranteed tenure is the proper solution to the problem. The statistical data referred to in chap ter I I suggest that there is no universal tendency fo r hourly rates to rise so as to compensate fo r variability o f employment. However, it is un doubtedly true that in some very seasonal lines, hourly rates are very high; and variability o f work is one reason offered by unions fo r even higher hourly rates. Often such unions are opposed to guaranteed wages, i f this is to mean lower hourly rates or slower increases in future hourly rates. In such situations, higher hourly rates may be better than no adjustment to heavy seasonal unem ployment. But so long as steady work does not involve extra disutility, the longer tenure provided by guaranteed wages would be a more economically efficient arrangement because then, and only then, would there be an incentive toward regularization o f employment, an incentive to put the available worker to some sort o f useful work. The only real economic drawback to guaranteed wages in such a situation arises out o f the very real possibility that the workers laid off by the industry in question might possibly (had there been no guaranteed wage) been able to find fill-in productive employment elsewhere. Such possibili ties are no doubt lim ited, but to the extent that they exist, it is not proper to regard the industry in question as being responsible fo r totally with drawing labor from other useful w ork; it is there fore incorrect to saddle the industry with their fu ll guaranteed wages. In fact, in such situations, the guaranteed wage can have harmful effects on the mobility o f labor. Our proposal o f chapter V I with respect to “ integration” with the unemploy ment insurance system may remedy this in part.*® D O V E TA ILIN G AN D LA R G E M ONOPOLIES T o the extent that the idle time o f the workers hired by one industry can be dovetailed with the operations o f another industry or with schooling and leisure, we need not attribute to the industry the fu ll cost o f its laid-off workers. Ideally, dove tailing o f activity is one solution to the irregular ity o f seasonal and other short rhythms o f different industries. Hence it is essential that unions relax their objections to changes in function and ac tivity o f workers. Dovetailing is easier the larger the units, since when many intermittent lines o f activity are merged, the extremes o f their sums is much less than the sum o f their separate extremes. Or to put this in another way, by merging activities we can cancel off the peaks o f some against the troughs o f others. This would seem to provide some argument in the direction o f large-scale mon opolies, necessarily possessed o f tremendous amounts o f economic power. And so it does, i f the dovetailing is to take place within the firm ; i. e., if the work is to be brought to the worker and not the worker to the work. However, dovetailing tends to stabilize pri marily to the extent that the components dove tailed have different cyclical rhythms and phases; in the ideal case o f the coal and ice dealer, the com ponents w ill have opposite phases so as to be mutually canceling. Therefore, the firm which wishes to stabilize its time-profile o f production cannot do so simply by growing in scale or by add ing closely duplicating lines subject to the same timing o f demand. It must branch out into new territories, take on new and different lines. Con-8 0 80 W orkers in some lines of activity employment. To the extent that they o f leisure, they need not receive a full collective bargaining in such cases w ill guaranteed wage. may prefer intermittent value occasional periods annual wage. No doubt have other goals than a 39 sequently, the fact that it grows in size need not entail a commensurate growth in its monopoly power. Monopoly power depends prim arily upon a firm’s relative importance in comparison with the total demand for each o f its products. I f it grows by branching out into new lines the firm is also spreading itself over wider areas o f de mand. T o revert back to the coal and ice dealer, there is no special reason to believe that his monop oly power in either o f the two markets is greater that that o f two separate dealers each handling one line. H is over-all efficiency may be greater; but that is a good thing rather than a bad.3 1 PO O LIN G L A B O R R E SER VES TH RO U G H A W E LL ORGAN IZED M A RK ET I f we pursue the argument concerning the de sirability o f pooling the labor reserves o f many lines o f activity in order to minimize unnecessary idleness, we are confronted with the problem o f the general labor market itself. W e can con ceive o f a vastly improved labor exchange to which workers go fo r placement. In the fantastically extreme case where each day every person and every firm turn anew to the central labor exchange for reassignment, idle labor reserves would be at an absolute minimum, the whole being steadier than its parts. Common sense tells us that there are insurmount able objections to such a fantastic case. In any factory this morning’s production w ill depend upon yesterday’s activity in such a way that en tirely new men coming into the plant would be at a great disadvantage, however good their general experience and training. Moreover, even if the inefficiency o f such an arrangement could be over come, we must still reckon with the intrinsic prop erties o f human nature; the worker would find it profoundly distasteful and disorganizing to move on to a new job at frequent intervals; man requires continuity in his working as well as in his home life. W e need not prolong the discussion o f what is admittedly an extreme case. There is no reason why a proposal to pool all workers together into one organized market should not be perfectly com patible with job continuity for the great m ajority * * Those who regard “ bigness” per se as an evil, without regard to its market consequences, may still object to the firm’s growth. 40 o f workers, most turn-over being confined to the “marginal” workers as is now the case. But this discussion does high light the problem o f recon ciling job continuity for the individual with maxi mum national productivity. Thus, if guaranteed wage plans were universally adopted prior to any successful attack upon the problems o f stabilizing production, the effect might be to attach workers to a given firm when there was no useful work fo r them to do there, even though, at the same time, there was produc tive employment to be found elsewhere.3 Boon 2 doggling is boondoggling, and bad economic policy whether it is pursued by the Government or by private enterprise. Boondoggling is particu larly costly during periods o f high employment when productive employment opportunities are plentiful. This illustrates that it is as necessary in this chapter, as in the last one dealing with aggregate demand, to distinguish carefully between the d if ferent conditions o f economic life upon which guaranteed wage plans impinge—between high employment and depression economies. There is a danger that guaranteed wage plans, not coupled with effective stabilization planning, could immo bilize men in such a way as to increase labor hoard ing and the inefficient use o f resources. S T A B IL IZA T IO N AND D IF F E R E N T IA L PRIC IN G Thus far we have been discussing the costs o f unavoidable idleness. Even if employment were not stabilized, guaranteeing steady wage income might so improve labor productivity and turn over rates as to be, in effect, costless. This would be all to the good. O f course, it would be better still if in addition to making labor more produc tive while at work, it were also possible to find useful continuous work to do. Stabilization may, however, be costly at the out set. It takes effort and planning. Under the present wage contract, which permits the layingoff o f workers on short notice, there may be little, ** Because o f the considerable difficulties and costs involved in workers going from one job to another in order to utilize their special skills, there is also much to be said for mobility on the part o f the employer in shifting between production lines rather than throwing the fu ll burden of mobility upon the workers. T his in no sense means that we minimize the importance of labor mobility between and within firms. if any, counterbalancing savings o f cost from steady operation.8 3 But under the more optimal system o f pricing discussed in earlier sections, where each employer is made to pay fu ll wages to all workers whom he withdraws from productive employment else where, the costs o f stabilization w ill often be lower than the costs o f not using the workers whose pay goes on anyway. Accordingly, stabilization plans, under this procedure, might be promoted. Being available, such workers might as well be used for whatever useful goods they could produce. Under high employment, the ingenuity exercised by an employer in financing odd jobs, or less than fully productive tasks, for his men to do in off seasons is highly commendable. The construction industry provides an illustra tive hypothetical case to show how guaranteed wages might give rise to a different system o f pricing, leading to a more efficient and equitable use o f our national resources.8 Let us suppose 4 that a house requires, say, 2,000 man-hours o f labor to build in the summertime and 4,000 in the win ter. I f labor can be hired and fired by the season at the same hourly rates, say $2 per hour all year around, no one would ever build in the winter, because by waiting until the follow ing season, the buyer could save 50 percent in labor cost. Construction workers would be out o f jobs for, say, 6 months o f the year. Houses would be sold for less than their true social cost, most o f the differ ence coming out o f the workers’ loss o f income. Theoretically, it might be alleged that a way out o f this situation would be for workers to ac cept half wages in the winter. But in view o f union insistence on uniform ity, and if it were as sumed that similar workers on steady jobs else where receive around $1.50 per hour, we may re ject this suggestion as unrealistic. 88 One saving, mentioned earlier, results from the fact that less plant and capital equipment are necessary when loads are spread evenly over time. Against this saving in interest and deprecia tion (resulting from time rather than use) there is the fact that in many industries the extra capacity serves as a stan dby; in case of heavy sales, the busy season is lengthened. A plant op erating near to capacity all the time cannot expand its production in this w ay, but m ust depend on overtime or more intensive production. 34 A s indicated in ch. V I, we do not think that any union should be compelled to adopt guaranteed wages except through voluntary collective bargaining. However, the present hypothetical case of the construction industry throws some light on the economic de sirability of collective bargaining taking one form rather than another. Under a system o f guaranteed wages the em ployer would realize that his summer contracts really involved more than the wage rates paid then. Moreover, during the winter his labor force would cost him nothing extra if he kept it at work. The result: He would produce houses all year around, selling them in all seasons at about the same price (based on the hourly wage times 3,000 man-hours, the average between winter and sum mer labor requirement). I f the old hourly wage rate were retained, the labor cost o f a house at the site would now be $6,000 summer and winter. This increase in price results from the product bearing its full economic costs. However, if part o f the $2 hourly rate repre sented a premium to compensate for the winter idleness, and if similar labor elsewhere received only $1.50 for steady jobs, then the hourly rate might now fall and workers would still be better off than under the old regime with a winter lay-off. Their annual take-home pay would be up by 50 percent. The labor cost (at the site) o f houses, summer or winter, would now be about $4,500 (or 3,000 times $1.50), which is just where it should be.*5 Apart from the social desirability o f subsi 3 4 8 dized low-cost housing, the price indicated would draw the right amount o f resources into the build ing industry, and all o f such resources would be efficiently used. This is only a numerical example; it is meant to be illustrative rather than to describe an actual situation. Moreover, it assumes a condition o f adequate aggregate demand and high over-all employment so that the excess labor in the unstable, seasonal industry could readily find jobs else where, once the industry were stabilized. Often a system o f variable pricing may prove to be economically optimal. Such a system would mean high prices during peak demand seasons when (m arginal) costs are high and output must be rationed; and lower prices during off-peak periods. In addition to price reduction, terms o f sale and delivery may be made more agreeable in off periods and less agreeable in peak periods. Alternatively, a seller in peak seasons may offer a reward in terms o f priority in delivery dates to 85 In a moment we shall consider the possibUity of seasonal variations in demand, giving rise to different prices a t different seasons. 41 those customers who have bought steadily in off periods. T o be effective, and profitable, the proc ess o f stabilization may require something more than rewards to good customers; it may also be necessary to apply penalties to bad ones. And the firms may be compelled to refuse the casual busi ness o f new customers at the peak o f a sales boom. Were our markets not so im perfectly competitive and characterized during normal times by “ buyers’ markets,” businessmen would realize more keenly that peak-load, fair-weather customers are often very expensive customers who do not pay their way. Most o f the above patterns o f price and market behavior have already been tried on a lim ited scale. Often they have later been abandoned be cause they were not economical. It might well prove to be otherwise in a world o f guaranteed wages, where the penalties for unstable produc tion would be considerably greater than at present. VI. POLICY RECOMMENDATIONS The analysis o f previous chapters shows that there is much to be said fo r guaranteed wages, particularly for seasonally or otherwise unstable industries operating in an economy with relatively high and stable levels o f aggregate demand. But it also suggests that guaranteed wages cannot be regarded as a cure-all for depressions. There are important limitations to the device as an anti depression measure—especially in comparison with other economic policies such as fiscal policy and other full-employment programs. In view o f these limitations, so long as our society suffers from rather wide fluctuations in business activity, guaranteed wage plans should be introduced only with proper safeguards. This conclusion is so important that we think it worth restating in a slightly different form . W e do not believe that the limitations referred to above warrant the conclusion that guaranteed wage plans are economically unsound and should be abandoned as goals. However, these reserva tions are sufficiently important to put us on guard against viewing the guaranteed wage as a solu tion fo r the problem o f unemployment; and especially to warn us that it cannot and should not become a substitute fo r social security mea sures or a full-employment program. In this con cluding chapter, we should like to put forward for discussion four policy recommendations. Becommendation No. 1. Guaranteed mages should he a m atter fo r voluntary collective bar gaining without legislative compulsion by the Government.— Our analysis has not led to the con clusion that the Government should seek to inter vene actively by legislation to require that indus try introduce guaranteed wages. In the present state o f our economy, this is a matter that should 42 be left to voluntary free collective bargaining be tween management and labor. Becommendation No. 2. In many lines guaran teed^ wage plans may properly include limita tions as to the employer's liability.—From our previous analysis, and from a factual survey o f earlier depressions, it appears that unqualified guaranteed-wage plans, indiscriminately applied to industries subject to wide cyclical swings o f demand, might involve such extensive financial costs as to prevent industry from undertaking such program s; or if industry had nevertheless undertaken such programs, they might have to be prematurely abandoned in order to avoid insolvency. Some sort o f limitation is therefore necessary. Existing guarantee plans provide for various kinds o f lim itations: coverage o f less than all workers; guarantee o f work for a limited number o f weeks per year (fo r example 40 w eeks); guar antee o f less than full-tim e weekly pay fo r weeks covered; provision for nonrenewal o f the guaran tee ; and so forth. W hat is above all needed, how ever, is a form o f limitation which w ill in normal times give a maximum o f security and protection to the worker, and to the employer a maximum incentive to stabilization efforts; but which, in emergency times, w ill come into effect to keep the firm’s financial liability under the guaranteed wage within reasonable limits. Because o f the ever present danger o f cyclical slumps we tend to look with favor on some sort o f limitation on the percentage o f the total guar anteed pay roll that the firm shall be liable for in any year or contract period. So long as production declines are not too great, workers w ill receive 100 percent o f their fu ll normal income. But if and when deep depression years come along, and lay offs from productive work involve more than, say, 10 percent (or some other suitable ratio) o f the guaranteed payroll, then the workers may have to receive something less than the fu ll guarantee. The precise percentage liability is a matter for experimental determination over time, and would no doubt vary from industry to industry depend ing upon their differing degrees o f cyclical vari ability.8 F or many lines o f activity, 10 percent 6 may be a reasonable percentage limitation. W ith such a limitation, it should not be necessary fo r private industry to accumulate unduly large reserve funds. Moderate reserve accumulation may be considered necessary fo r prudent financial management; and, if not excessive over the whole period o f the cycle, the accumulation and subse quent expenditure o f the reserves may contribute to dampening the business cycle, helping to con tract excessive boom-time purchasing power and to expand it during depression. But i f ever the guaranteed wage should be widely adopted, and if substantial private reserve funds were accumu lated in addition to the continually accumulating governmental reserves for old-age security and unemployment insurance (currently around 14 billion dollars) the problem o f maintaining high employment might be made more difficult. Momentarily, the continued growth o f these social security trust funds is helpful as an antiinflationary device. Nevertheless, over the long pull we believe these programs should be placed more nearly on a pay-as-you-go basis. Similarly, we should not like to see a long-run accumulation o f large guaranted wage reserve funds. Whether for social insurance or guaranteed wages, moderate reserves that could be drawn on in the depression phase o f the cycle would, however, not only be permissible but desirable. They would help to serve as an automatic anticyclical compensatory mechanism. W ithin the collective-bargaining contract itself, there would probably have to be included certain special stipulations further lim iting and qualify 86 The Latim er Report on guaranteed wages presents data show ing w hat the costs to a number o f industries of an unqualified guarantee would have been in earlier years. See ch. V II I, pp. 7 0 100. The report discusses the rather intricate problem o f how the guaranteed wage funds are to be allocated among the laid-off workers whenever the drop in production turns out to be great enough to bring the percentage limitation into effect. ing the guarantee o f work. For example, there might be a stipulation that wages would not be payable to workers out on strike. Collective-bar gaining clauses might be introduced into the guar antee also lim iting wages o f workers in the plant, not themselves on strike but rendered unable to work by virtue o f strikes outside the plant—or for that matter inside the plant but in another department or craft. Thus a transportation strike might so impair the flow o f supplies as to make a continued guarantee unbearably costly. A ll these are matters for free collective bargaining. The primary purposes o f guaranteed wages are to insure workers more security and promote sta bility o f production. The device is not intended to create permanent quasi-feudal rights o f a worker to a given job. Thus, if a sweeping technological change makes it no longer necessary to keep the previous number o f workers attached to an in dustry or firm, or if there is a secular change in people’s tastes and demands, then the guaranteed wage should not serve to freeze workers to their old jobs and techniques and to thwart our econ omy’s increases in efficiency and responsiveness to demand, upon which the ever-rising standard o f American living fundamentally depends. Most changes are not introduced discontinuously and suddenly. The normal attrition o f the labor force o f a firm, through turn-over and sepa ration, makes it possible to reduce gradually the number o f workers without undue lay-offs. Where this is not the case, the guaranteed wage, together with severance pay, may perhaps be used to soften the impact o f the socially-desirable proc ess o f change upon the workers directly involved. But we feel it necessary to point out the danger that, with guaranteed wages introducing length o f tenure as a new variable in the collective-bar gaining negotiation, the long-term flexibility o f our economic system may be impaired unless or ganized labor is w illing to waive or lim it the scope o f the guarantee in situations involving perma nent changes in technology and demand. Recommendation No. 3. Corporate income tax lam and practice should be changed so as to en courage guaranteed mage programs.—Under pres ent tax law and practice, corporations may for tax purposes deduct as an expense all payments made to workers under guaranteed wage pro grams, during the year when such payments are 43 made. Suppose, however, that a corporation wishes in good times to set aside a reserve fund in anticipation o f the time when it w ill have to pay out guaranteed wages to the workers it is cur rently hiring. Such appropriations to guaranteed wage reserve funds—unlike retirement pension and profit-sharing programs—cannot now be treated as an expense to be deducted from taxable earnings in good times; only later, when actually paid out, can they be so treated. The present tax situation is, therefore, in a sense discouraging to guaranteed wage plans. It is, however, easy to exaggerate the quantitative im portance o f this whole matter. A few years’ de lay in receiving a tax reduction or credit means prim arily the loss o f interest on the sums involved. This is all the more true to the extent that cor porate income tax rates remain stable, and to the extent that a system o f carry-forward (and carry back) is in force so that the accounting period for tax purposes becomes longer than a year and varia tions in income are averaged over a longer period. There seems to be no question that any change in existing practice would require new legislation by Congress. This might take at least two possible forms. First, Congress might specifically provide that any guaranteed-wage payments might be carried back over previous years as deductible ex penses. In effect, the company would be reim bursed fo r its previously overpaid taxes, resulting from its previously overstated earnings—over stated because the accruing liability for guaranteed-wage payments had not been recognized as an expense o f production at that time. A second method o f handling this problem would be to permit appropriations to the reserve fund to be deductible as expenses during the year when they are being set aside. Then, later when the reserve is being used to finance guaranteed wages, such outlays would not again be chargeable as an expense and as a tax deduction—for, other wise, there would be an unwarranted tax loss to the Treasury and an unfair subsidy to the guaran teed-wage plan. So far this second plan seems simple enough; but under some circumstances d if ficulties may arise. W hat if more is appropriated to the reserve than is ever necessary to meet guar anteed wages, either because o f good luck, or be cause o f good stabilization planning by the employer, or because the employer deliberately 44 builds up an excessive reserve? Certainly under these circumstances it would be wrong to permit the employer to recover the excess reserves, and to escape corporate income taxation completely on the sums involved. For this reason, it would be necessary to insist that the reserve be a “ trust,” created to provide guaranteed wages or other benefits for the workers. The trust would neces sarily have to be irrevocable as far as the corpora tion was concerned.3 7 Our previous economic analysis and the specific accounting aspects o f guaranteed wages justify, we believe, the recommendation that the present tax deterrents to guaranteed wages be removed by one or another o f the above described m ethods: that is by a specific carry-back on guaranteed wage costs, or by tax deductibility (w ith safeguards) o f appropriations to guaranteed wage reserve funds. Recommendation No. 4. Social security legis lation should he altered so that guaranteed wage plans can he integrated with the unemployment compensation system to the mutual advantage o f ecwh.—Guaranteed wages should never be looked upon as a substitute fo r our existing unemploy ment-insurance system. This system admittedly needs substantial improvement and strengthening. And we believe that guaranteed-wage plans if in tegrated with social security could operate in this direction. Such integration would serve to strengthen rather than weaken the system o f un employment insurance. The workers o f employers who maintain wage payments under guarantee plans should be per mitted to draw unemployment benefits, provided that they are unemployed in the sense that would qualify them for unemployment compensation in the absence o f a guarantee.3 I f the employer 8 had to pay only the difference between the full guaranteed wage and the amount paid by the un employment insurance system, then he might be more w illing to undertake such wage guarantees and more w illing to broaden the coverage and ex tend the duration o f the guarantee. 87 Any excess reserve funds might be finally used up if aU further appropriations to the reserve fund were curtailed. I f even this does not meet the problem, the terms of the trust should stipulate that any remaining excess funds be used for the benefit of the workers through providing old-age pensions or for some other welfare provision now permissible from the tax standpoint. 88 Some im portant technical and administrative problems have to be faced. These are discussed in M r. Latim er’s report. Moreover, the employer would then be in the position o f wishing to support the extension o f Social Security instead o f—as under the present system o f merit rating—having a financial in terest in keeping its benefits limited. The only alternative to the method o f “ integration” here proposed, that would tend to remedy the present discrimination against the employer who volun tarily agrees to guarantee fu ll incomes to his idle workers, would be the method o f further extending merit rating—perhaps to the point where all em ployers with effective guaranteed-wage plans would be excluded completely from all unemploy ment insurance taxation. W e should consider the adoption o f such an alternative as something o f a calamity, designed to divide and weaken the forces making for an effective social security sys tem. The method o f integrating social-insurance benefits with guaranteed-wage plans would have just the opposite effect, as we have already indi cated. It would serve to strengthen rather than weaken our social-insurance system. In addition, the integration with present unemployment com pensation would have the further advantage o f reducing the tendency fo r guaranteed wages to freeze workers to a given employment when it may be socially desirable for them to transfer to other occupations. Under the integrated plan, in order to draw unemployment benefits, workers would have to be registered with the central employ ment exchange, and would be obliged to accept suitable alternative employment. This would lessen but not obliterate the tendency for 100 per cent income maintainance to have deleterious effects upon the m obility o f labor. Important as these legislative changes are, we do not mean to im ply that the growth o f guaran teed-wage plans should wait upon the reforms here suggested. On the contrary, in some lines o f industry the guaranteed wage can be economically introduced under present conditions to the ad vantage o f all parties. As the number o f such industries grows, both employers and employees would have an increasing incentive to press for the needed legislation. These legislative reforms, especially the inte gration with social security, would make it pos sible to strengthen such guaranteed-wage plans as may in any event be developed without waiting for new legislation; and they would encourage the spread o f plans to other industries which (even with some kind o f limitation o f employer liabil ity) might otherwise find it difficult to set up ade quate and satisfactory programs. 45 ADDENDUM-DEPRESSION POLICY AND THE GUARANTEED W A G E N o t e : This Addendum is presented as a basis fo r further discussion. It does not constitute a part o f our Recommendations. The analysis o f previous chapters points to the conclusion that the guaranteed wage would be a weak rod to lean upon, by itself alone, as an anti depression measure. Taken in conjunction with other measures—especially fiscal policy and social security—it can serve a useful purpose as a supple mentary device. The guaranteed wage is not a cure fo r depressions. However, as we have seen in chapter V , it could perhaps serve its most use fu l purpose by overcoming seasonal or other irregularities o f employment in industries oper ating in an economy with relatively high and stable aggregate demand. TH E F E A R OF D EPRESSIO N But industry cannot safely assume that we shall in fact achieve in future a continuing high and stable level o f demand for goods and services. It is precisely the fear o f recurring major depres sions that makes management reluctant to under take commitments under the guaranteed wage. So long as the prospect o f serious depressions remains as a m ajor but incalculable risk, guaran teed-wage commitments must be rather severely restricted with respect to the liability involved. Such restrictions could be eased somewhat by building up in good times guaranteed-wage re serve funds. Guaranteed-wage plans, privately financed, could reasonably be regarded as a manageable risk in a society freed from the fear o f m ajor de pression. And even though this danger o f recur ring depressions must be faced, plans can, we be lieve, be devised along the lines discussed in chapter V I, which are made as “ cyclone-proof” as possible. It now remains to consider whether some direct Government financial support o f guaranteed wages might not be justified from two standpoints: (1) as a means o f encouraging, in 46 an unstable world, more widespread adoption by private industry o f guaranteed wage plans; and (2) as a part o f a rounded governmental anti depression program. ECONOMIC A SPECTS O F GOVERNM ENT A ID The purely economic aspects o f such Govern ment support deserves to be considered on its own merits. In terms o f political realism, so long as we are not prepared to accept direct Government contributions to social insurance programs (a policy adopted at the very inception o f social in surance in England a generation a g o), there is no likelihood that governmental grants-in-aid to guaranteed-wage plans w ill be acceptable. Social insurance should be first in line for governmental contributions from general revenues. This is true, among many reasons, partly because social insur ance is o f universal application (within limits legally established) and does not apply only to such individual firms as may voluntarily choose to take advantage o f Government aid; and partly because social insurance provides a lower minimum income than is contemplated in most guaranteed-wage plans. Virtual universal coverage, taken in con junction with the principle o f a national minimum, commends social insurance as a program deserving direct Government financial aid and support. The case is less strong for the guaranteed wage. Until direct Government contributions to social insur ance become a reality, it is probably somewhat utopian to discuss such aid fo r guaranteed wages. We should ourselves wish first to see direct Federal contributions to social insurance, thereby dimin ishing the deflationary effect o f the regressive pay-roll taxes. Nevertheless, the economics (apart from the politics) o f support for the guaranteed wage de serves study. In terms o f employment and anti depression policy, there is much to be said for governmental grants-in-aid to guaranteed-wage programs. Indeed, even with respect to social in surance, Government contributions commend themselves in large part by reason o f their implica tions with respect to fu ll employment policy. The more social insurance is financed from progressive income taxation, the more expansionist w ill the program be. Our exploratory discussion, in this addendum, o f partial governmental aid to the guaranteed wage, contingent upon emergency depression con ditions, is very much in the spirit o f many o f the less controversial Government programs in other spheres. Federal insurance o f bank deposits and the capital salvage activities o f the Reconstruction Finance Corporation are more or less analogous instances. A third and more venerable precedent is that o f central banking (Bank o f England, the Federal Reserve System, etc.) whose function has long been recognized as that o f providing succor o f “ last resort” in times o f crisis and depression. Accordingly, we wish to raise the question whether the program o f financing guaranteedwage plans might not, to a degree, be incorporated in a comprehensive antidepression program, such as the Government may find itself called upon to undertake in periods o f substantial unemploy ment. W e believe that there are few, i f any, who doubt that we shall in fact encounter more or less violent fluctuations in business activity. A slump once started can cumulate, under modern condi tions, with incredible speed. The Federal Gov ernment will not be able to stand by without under taking some positive program o f action. In such periods, governmental intervention in various di rections w ill be required in any event. In such periods o f crisis, therefore, governmental support for existing guaranteed-wage plans might provide one additional compensatory mechanism, thereby reducing somewhat the danger o f unplanned and makeshift depression projects which otherwise are likely to be undertaken. W e believe, as we have repeatedly stated, that worker income security can best be promoted by strengthening our already established system o f social security and by a comprehensive full em ployment program designed to maintain a high and stable level o f demand, production and em ployment. W ithin this pattern, the guaranteed wage can be made a useful supplementary device. It is not something upon which we should embark with a view to making it carry the whole load. It is not and cannot, we believe, be made the basic foundation o f worker income security. But it can be made a significant part o f a larger program. T o this end we explore below two highly tenta tive suggestions. W e present them, not because we believe them currently to be within the range o f practical politics, but to round out our economic analysis o f the guaranteed wage, and to provide a starting point for further discussion. W e suggest, in line with the discussion in Chapter Y , that guar anteed-wage programs shall be wholly privately financed except in abnormal emergency periods o f serious depression. W e consider governmental aid to such plans only as a last resort measure, and as a part o f a comprehensive attack on depression. W e therefore explore two criteria indicating at what point in the depression phase o f the cycle governmental aid might be invoked. The first is based on the level o f profits o f the firm in question; the second is based on the broad general level o f business activity in the nation as a whole. GOVERNM ENTAL G R A N T-IN -A ID BASED ON P R O F IT LEVELS OF FIRM S Precisely in a condition o f serious mass unem ployment, government support for the continua tion o f guaranteed wage plans may be far more sensible than ill-designed antidepression measures, which in the absence o f adequate planning are likely to be improvised. I f a firm were assured that, upon a fall in the annual profits o f the com pany to a small fraction o f, say, the previous 5-year average, the Government would come to its support, the guaranteed wage plan might be saved from extinction; and what is more, many firms might be induced to establish guaranteed wages in good times, if. they were assured that the Govern ment would bear part o f the costs contingent upon a deep depression when profits are low and losses are likely to be sustained. W hile it is true that special conditions may cause low profits for an in dividual firm even though general business activity is high, nevertheless experience shows that the profits o f most firms fluctuate in close correlation with the cycle. A more complicated formula than the one we suggest below might be designed so as to combine both the criteria o f (a) the firm’s profits and (5 ) the general state o f business. 47 Using the more simplified formula, however, let us suppose that the firm’s profits have fallen to 30 percent (or some other appropriate figure) o f the previous 5-year average. The Government might now undertake to contribute a grant-in-aid on a sliding scale basis, and increasing at a specified rate as profits continue to fall. The scale might, by way o f illustration, be somewhat as follow s: Percent profits o f previous 6-year Government ehare of average : guaranteed wage cost 3 0 ___________________________________________ 10 2 5 __________________________________________ 20 2 0 __________________________________________ 30 1 5 ___________________________________________ 40 1 0 __________________________________________ 50 According to this formula, the Government would make an outright grant-in-aid to the current cost o f the guaranteed wage plan (that is, to the guar anteed wage payments fo r workers declared by the company to be totally unemployed) beginning at 10 percent o f the cost when profits fa ll to 30 percent o f the previous 5-year average. The grant-in-aid might rise on a step-up basis as profits fa ll, finally reaching a maximum o f 50 percent o f the cost when profits fall to 10 percent or less. (A more precise formula is given below.) In the next section we wish to discuss an alter native grant-in-aid procedure, based not on the profits o f the particular firm, but on more general employment and business conditions in the Nation as a whole.8 ® GOVERNM ENTAL FIN A N C IA L A ID B A SE D ON LE V E L O F GROSS N A TIO N A L PRODUCT The feasibility o f a guaranteed wage, financed by private business, depends in large measure upon the kind o f economy we shall have in the future—how violent w ill the fluctuations in the gross national product (G N P ), employment and business activity be? T o the extent that we suc ceed in narrowing the over-all range o f fluctuation in the gross national product, the guaranteed wage becomes more feasible as a practical proposi tion. It is not feasible fo r guaranteed wage plans to include any substantial proportion o f the total w Again it may be noted th at a more complicated formula m ight be constructed which combines both criteria. On this basis, the governmental grant-in-aid would apply only when both general business conditions and the firm’s profits had fallen to a certain level. 48 labor force in a society in which the gross national product fluctuates from 100 billion dollars to 50 billion dollars and then up to 200 billion dollars, as it actually did from 1929 to 1944. I f, however, we could narrow the range o f fluctuations in the gross national product to within 15 to 20 percent, it would become feasible to include in guaranteed wage plans a substantial percentage o f the total labor force. Grants-in-aid to guaranteed wage plans, once the gross national product had fallen to a cer tain level—fo r example, to 85 percent o f a fu llemployment level—might well be included as a part, however modest, o f a general antidepression program. There would indeed be no guarantee that the gross national product might not fa ll far below the level at which grants-in-aid would begin. The program would amount to no com mitment with respect to fu ll employment. It presumably would mean that the Federal Govern ment would seek to promote fu ll employment by such politically acceptable measures as were avail able. I f these were reasonably successful, the gross national product might not fa ll to the level at which the grants-in-aid program would begin. Up to this point, private business would be ex clusively responsible fo r the financing o f such guaranteed wage plans as had voluntarily been put in operation. The grant-in-aid could be stepped up if the GNP should continue to fa ll to lower levels. It is here suggested, by way o f illustration, that at 85 per cent o f a full employment GNP the grant-in-aid might equal 25 percent o f the current cost o f any guaranteed wage plan, i. e., 25 percent o f the amount actually paid out in wages to workers cov ered by the plans over and above the wages re ceived by workers fo r productive employment to gether with unemployment compensation, during the year or period covered. I f the GNP should fa ll to 80 percent o f the fu ll employment level, or below, the grant-in-aid might be raised to 50 per cent o f the cost. (A more precise form ula is given below.) The criterion has been stated in terms o f the gross national product (G N P ). In point o f fact, the statistical calculation o f the GNP is neither sufficiently current nor sufficiently precise to provide the basis fo r a legal determination o f the point at which a grant-in-aid would be made. Some other statistical index which is precise and unequivocal, but which reflects fairly accurately the fluctuations in the gross national product, would have to be found. This is a matter which w ill require much careful study. Possibly the ratio o f the number o f workers receiving bene fit payments under the unemployment insurance system to the total number o f workers covered by the system might serve as an appropriate index. But the ratio should be applied so as to permit the entry o f grants-in-aid (together with the step-up) as nearly as may be at some such points in the level o f the gross national product as were indicated above. It may be appropriate to note that the system o f “ compensatory payments” to farmers, advo cated by some leading agricultural economists, offers an interesting analogy to the suggestion to support guaranteed wage plans at say 80 or 85 percent o f a “ full employment” GNP or below. The agricultural “ compensatory payments” plan, very briefly, may be explained as follow s: 4 The 0 plan does not envisage any price controls or mar ket price supports. Assume that, as a result o f a depression, the market price o f a given farm commodity falls below 85 percent, say, o f the pre depression base. Assume the predepression price was $1 and it now falls to 60 cents. The Govern ment then undertakes to make compensatory pay ments to all farmers covering the difference between the price actually received (60 cents) and 85 cents, the guaranteed 85 percent o f the prede pression base. Such payments would provide minimum income security to farmers without in terfering with market forces. Prices would not be raised. The commodity would not be “ priced out o f the market.” But the farm er would be protected. A nd the compensatory payments would be made by the Government at just the phase o f the cycle when it is important for the Govern ment to find sensible, useful and justifiable out lets fo r public expenditures—expenditures needed to swell the diminishing income stream and stop the cumulative collapse. A s soon as recovery had proceeded to a point at which the price o f the com modity had again reached 85 percent o f the pre 40 See Theodore W . Schultz, Agriculture in an Unstable Econ omy (M cGraw -H ill, 1 9 4 5 ), pp. 2 2 0 -2 3 5 . This is a research study for the Committee for Economic Development. depression base, the Government support would cease. So also in the Government’s support o f the guar anteed wage; as soon as the GNP had recovered to 85 percent, say o f the “ fu ll employment level,” or the more specific and definitive index indicative o f the degree o f full employment (such as the ratio o f current beneficiaries to the number o f covered employees), the Government’s support o f estab lished guaranteed wage plans would cease. Our two suggestions involve one possible tech nical difficulty from which the agricultural com pensatory payments plan is free. According to that plan, as farm prices drop from 86 percent o f the predepression base down to 84 percent, nothing very drastic or discontinuous happens. Such a small change is accompanied by a commensurately small change in Government payments. A ll trouble o f “ discontinuous brackets” is avoided. A M ORE PRECISE FORM U LA FO R EACH PRO PO SAL Our two suggestions seem to involve this tech nical difficulty. As a firm’s profits suddenly fall to below the critical 30 percent level, or as GNP falls below the critical 85 percent level, suddenly and discontinuously the payments o f the Federal Government go from nothing to the first bracket rates. This might give rise to troublesome litiga tion concerning the last decimal point o f profits and GNP, and might even occasion deliberate efforts on the part o f firms to cause their profits to fa ll! These bracket difficulties can be avoided by making our grants-in-aid formula or table slightly more complex. Thus, just as the Government financial share grows from 10 to 20 percent as profits fall from a 30 percent level to a 25 percent level, so we might arrange for the Government’s share to rise grad ually to the first 10 percent level, corresponding to the gradual fall in profits from 35 to 30 percent. Our exact formula might be something like the follow ing: ( а ) When percent o f corporate profits to nor mal are above 35 percent—nQ Government aid. (б ) When percent o f corporate profits to nor mal are less than 10 percent—Government aid equal to 50 percent o f cost. 49 (c) Anywhere in between these two extremes we suggest the form ula: Percent o f Government aid equals 70 minus (2 times the percent profits are o f norm al). This could be written as an equation as follow s: G = 7 0 —2P, in which G is the percent o f govern ment aid and P is the percent which current profits are o f normal profits. This w ill be found to agree with the earlier table, and to avoid all bracket difficulty. Sim ilarly, our second formula based on GNP too might be as follow s: (a ) Percent o f GNP to fu ll employment GNP above 90 percent—no Government aid. (b ) Percent o f GNP to full employment GNP below 80 percent—Government aid equal to 50 percent o f cost. (c) Anywhere in between, the form ula: Percent o f Government aid equals 450 minus (5 times the percent o f GNP to full employment G N P ). This could be written as an equation as follow s: 6r'=450 —5P ', in which G' is the percent o f Government aid, and P ' is the percent which the current GNP is o f a fu ll employment GNP. V A R IO U S ASPECTS OF TH E GOVERN M ENT A ID PRO G RAM The two alternative grant-in-aid procedures which we have outlined would have a number o f advantages. They would leave to collective bar gaining, or to the employer alone in the absence o f collective bargaining, the choice o f whether or not to set up a guaranteed wage. Under the pro cedures here suggested, if a guaranteed wage plan is set up, the responsibility would devolve wholly upon private enterprise so long as (a ) profits equalled 35 percent o f the preceding 5 years, or (b ) alternatively the gross national product did not fa ll below 90 (85 in the cruder form ula) per cent o f a fu ll employment product. The respon sibility o f the Government, up to this point, would not relate to the guaranteed wage, but rather to policies and programs that promote high business activity and high employment. Thus up to the point indicated the Government would not enter the picture at all with respect to the guaranteed wage programs so far as direct grants-in-aid are concerned. Should, however, the situation with-respect to individual businesses, or alternatively with respect 50 to the economy as a whole, become serious, Govern ment support in a variety o f directions would sooner or later become necessary. Grants-in-aid for established guaranteed wage plans might well prove, in such circumstances, to be one measure which along with others might stop a deflationary spiral. There is increasingly general support for the thesis that early and vigorous support in many directions is less dangerous than delay, and is much to be preferred to risking all on any one policy. Reasonably early action may prevent a bad situation from getting worse. A long with other antidepression measures, grants-in-aid to established guaranteed wage plans could help to put a cushion under a recession and prevent a further downward cumulative move ment. The governmental grant-in-aid would in volve the payment o f funds into the income stream at a time when a Nation-wide depression is becom ing really serious and needs to be stopped quickly and effectively. Planned public works and public development projects can play an important role, but many such projects cannot be timed quickly nor can they be adequate to check, without the aid o f other measures, a cumulative downturn. W hat other programs may be necessary or desirable (in cluding tax adjustment) is not a matter which properly comes under discussion here. The plans discussed above would be automatical ly self-disciplinary. I f the particular guaranteed wage plan introduced by an individual firm were overly cautious, it would be relatively ineffective. It would cost the firm very little, and by the same token, it would cost the Government very little; and it would not satisfy labor. I f the plan were overly ambitious, it would be costly to the firm up to the point at which the governmental grantin-aid would begin. This fact could be counted upon to deter the company from setting up too am bitious a plan. I f the firm had failed to make good on its financial responsibilities under the plan, having dropped it entirely either in accord ance with the terms o f the plan or in violation o f these terms, the Government would have no re sponsibility to undertake the grant-in-aid. Gov ernment support would be accorded only such plans as were actually in fu ll operation at the point where grants-in-aid were scheduled to begin. As soon as the GNP again rose to a level above the point indicated (or alternately profits rose) the Government support would cease. On this basis every firm would know that any plan it might inaugurate would entail financial obligations to cover the entire cost o f the guaran teed wage fo r its covered employees only under reasonably favorable conditions. In the event o f a nation-wide depression, it could rely upon Gov ernment support. The business firm must indeed prepare a plan which could reasonably be expected to weather the storm up to the point at which gov ernmental aid enters. And even then it must be prepared to carry its share o f the load. W ith such guaranteed support, a firm could at least go farther both with respect to the proportion o f its workers covered under the plan, and the percent o f the fu ll time wage guaranteed under the plan, than would be the case without such support. How far each firm is w illing and able to go, under these condi tions, is for it to determine fo r itself, either on its own volition or on the basis o f collective bargain ing with the union representing its employees. There is no governmental compulsion. Any firm which objects to a guaranteed wage is quite free to go on without any plan. The governmental aid, available when conditions become difficult, is in deed an inducement, but it represents no compul sion whatever to set up a guaranteed wage plan. It w ill be noted that both o f the suggested plans are free o f any element o f unfair government wage subsidization o f one producer as compared to his competitor. The Government contributes not a cent toward the wage o f any worker who is to any degree productively employed.41 Only after the employer has declared a man completely unem ployed would the Government’s aid be forthcom ing. W e do not wish to spell out the details, but presumably such workers would be required to be registered with the unemployment insurance labor exchange, and would be required to accept suitable alternative employment. Also, as discussed in chapter V I, we suggest that unemployed workers should be eligible for regular unemployment com pensation benefits, with the employer and govern " I t m ight be objected that an element of discrimination and Inequity is involved in the respect that government aid to the maintenance o f wage income is applied only to workers who happen to be employed by firms with a guaranteed wage. This, however, we should not regard as decisive. W h a t is important is that firms be encouraged to introduce the guaranteed wage, and that all firms doing so shall be accorded equal treatment. ment (under the guaranteed wage plan) making up the difference as specified in the guarantee. To some extent these suggestions avoid the criti cism that worker m obility w ill be adversely affected. The employer is still always left with an incentive—perhaps greater than that now existing under experience rating—to see that his laid-off workers accept suitable other employment. None theless, we must expect that any system o f almost 100-percent out-of-w ork compensation w ill weaken the employee’s exertions to find another job and w ill make him more particular in accepting prof fered alternative offers. This, however, is a basic difficulty o f all guaranteed wages, whether com pletely privately financed or financed in part by the Government. Moreover, the loss o f workers’ incentives toward m obility is a much less serious economic problem at a time when there is wide spread unemployment in all fields. Since the Government comes in with financial aid when guaranteed workers are declared com pletely idle, there w ill always be some tendency for firms with guaranteed-wage plans to relax in their efforts to find low-productivity “fill-in” jobs for the workers to do in their own plants. As soon as the Government support came into operation, the costs incident to the guaranteed-wage plan in any firm would be carried in part by the Government and increasingly so as the depression deepened. There would thus be less incentive left for the firm to hold down costs by putting the largest pos sible number o f covered employees to productive work or by securing work fo r them elsewhere. Nevertheless, there would always be some consid erable incentive left, since the firm would cover at first all, then most, and finally at least 50 per cent o f, the cost. The suggestion here made has the great merit that it could reasonably be expected that active full employment policies may be sufficiently suc cessful so that the GNP would fall below 85 to 80 percent o f the calculated “ full employment” level only under relatively unusual circumstances, when the underlying factors were such as to threaten a really serious m ajor depression. Under this sug gestion one may hope that the Government guar antee would be invoked on any substantial scale only in periods o f serious depression when vig orous measures to combat the downswing would, in any event, be necessary. 51 I f the GN P has fallen below 80 percent o f the “ full employment level,” the labor market would already be flooded with a large excess o f appli cants. Accordingly, from the social viewpoint, it would relieve the situation somewhat if the em ployees covered in guaranteed wage plans did not come on the market at all. Moreover, it might not help the general situation, under conditions o f falling and low aggregate demand, for firms hav ing guaranteed wage plans to enter the market with competing sidelines to give productive em ployment to some o f their covered workers. Such action, in a period when total demand was sharply restricted, would contribute little, if at all, toward an expansion o f aggregate demand. In such peri ods, taking on new lines by one firm probably sup plants jobs elsewhere rather than adding to the total. In contrast, in periods o f brisk demand and full employment, the elimination o f seasonal and frictional unemployment would represent a net gain in social efficiency and total real income. I f all workers were given steadier employment throughout the year, there would be a net gain in total output and total real income. Neither the payment o f compensatory supple ments to farmers nor the payment o f wages to idle workers covered by guaranteed wage plans rep resent the ideal situation. The ideal situation is one o f sustained aggregate demand and full or high employment. But if, in fact, fu ll employ ment policies prove to be only partially effective; or if wholly abnormal and unexpected factors emerge which cannot, in the short run, be coped with adequately, then compensatory payments to farmers and wage payments to covered workers under guaranteed wage plans may be valid, justi fiable and helpful under the circumstances. The situation is a temporary one in which unusual and rather drastic measures may become necessary. Financial aid to guaranteed wage plans is one channel o f depression Government spending. In terms o f the encroachment o f Government activity on the private sector o f the economy, it is a rela tively conservative measure, since it involves only “ transfer payments” to individuals, who in turn spend this money as they wish on the products o f private industry. Thus, it is neither competi tive with private industry nor “socialistic” in the 52 sense o f expanding the productive or collective consumption role o f Government. The problem is one o f spending wisely, effec tively, and equitably. Giving one favored group o f workers 100-percent unemployment insurance while others receive limited partial benefits fo r a limited time, or no insurance benefits at all but only relief, cannot be defended as equity. From this point o f view it might well be desirable to confine Government aid to plans with some sort o f time limitation with respect to the period in which any one worker could continue to draw guaranteed wage payments. As a first line o f de fense against depression, relatively generous treat ment o f certain workers for a limited period can be defended. But it would be more important still to strengthen our general social security, relief, and public works and development programs dur ing the next serious depression. In the final analysis, productive and useful employment should be the goal. The primary purpose o f the suggestions de scribed is to contribute in some measure to the general insurance against the social plague o f deep depression. These suggestions, like all insurance measures, would serve their purpose best if it never became necessary to invoke them—if the system always remained at high levels o f employment and general profits did not decline to low levels as a result o f adverse business conditions. T o conclude: An adequate full employment program would make governmental grants-in-aid to guaranteed wages quite irrelevant as an anti depression device. But in the event, as is not im probable, that we shall at times face rather se rious unemployment, such grants-in-aid could be a useful “ last resort” instrument, to be applied along with other measures as one part o f a many-sided antidepression program. But while it is necessary to be prepared with antidepression policies, we must never lose sight o f the positive program de signed to provide continuing and high levels o f useful and productive employment. Guaranteed wage plans are not adequate substitutes fo r other public and private economic policies, even though they may provide supplementary protection against instability o f production, employment and income. SUBADDENDUM-COMMENTS ON THE HANSEN-SAMUELSON REPORT By Prof. J. M . C lark , Columbia U niversity; Prof. E dw ard S. M ason , Harvard U niversity; Prof. Sumner H. Su ch te r , Harvard University, together with Final Com ments by Professors H ansen and Samuelson 1. COMMENTS B Y J. M. CLARK In general, the writer is in hearty agreement with the actual recommendations o f the report, and approves o f their cautious and realistic char acter. He has, however, a feeling that in the body o f the discussion this caution is not aways fully maintained; and here and there, by tone and im plication, an impression is conveyed o f promise o f greater results, both in extent o f voluntary adoption and in effect on stabilization o f employ ment, than are wholly consistent with the cautious character o f the actual recommendations. The frequent qualifications which are introduced do not entirely offset this optim istic impression. A few major and minor comments follow . Perhaps the chief aspect which seems insuffi ciently recognized is the probability that i f a guaranteed-wage plan were installed, covering all or most o f the employees in an industry, the hiring o f additional workers might involve such a heavy commitment on the part o f the employer that he would go slow about such hiring in times o f active demand, with the result that employment might not simply be stabilized, but the total reduced. This would place a heavier burden on the highemployment program which, it is assumed, goes with the wage program. It might be replied that, since the wage program is voluntary, employers would not be likely to undertake commitments so heavy as to have this effect. This seems not fully convincing, since the pressure o f organized labor toward such programs may be very strong, and may even develop quasicompulsory effects, im pelling employers toward assuming heavier com mitments than caution would dictate, in the light o f the possibility here considered. A second point is that, in view o f the voluntary nature o f the proposals, more attention might be given to modified plans, even if they may not so far have been proposed by interested parties. F or example, it might be regarded as a defect o f a simple guarantee from the standpoint o f workers’ incentives, that under it the worker would get paid no more fo r working the guaranteed amount o f time than for working less. This could be avoided, and the employer’s commitment be at the same time lightened, by a plan under which the worker would be paid full wages for full-tim e work, and something less though presumably more than half wages fo r idle time, down to some absolute minimum. The guarantee should properly be so figured as to give the worker something more than his unemployment benefits under social security, but less than fu ll pay fo r idle time. This would avoid asking him, in effect, to work part o f the time fo r nothing. Methods o f financing reserves deserve more dis cussion than is given them. In voluntary private plans, the employer would be under strong pres sure to invest them productively, in private securities. Then when they had to be drawn on, in bad times, i f these securities were thrown on the market in order to realize funds, the effect might be to accentuate depression, offsetting the alleviating effects o f the distribution o f the proceeds to the workers. This dilemma seems to deserve consideration; it might be avoided by some credit arrangement. A fourth point concerns the effect o f stabiliza tion on the productivity o f workers. Here I agree with the general conclusion that the net effect is likely to be favorable, and my reservation is largely concerned with the tim ing o f the effect. Experi ence seems to show that when bad times threaten, workers work harder, and increase their output. It is when a lay-off does not immediately impend 53 that workers are urged to make their jobs last as long as possible, by restraining output. This is complicated by the effect o f reduced output in increasing indirect costs per unit produced, so that total labor cost per unit o f product may rise, or output per man-hour fall, on account o f this latter factor, which has nothing to do with the attitude o f the workers, or how well they work. Dr. Julius Hirsch, in a paper delivered before the American Economic Association, January 25, 1947, at tempted to separate these two effects on produc tivity, and found that the kind that registers the efficiency o f the workers behaved in the way I have indicated. A fifth point concerns the incentive the em ployer has to make capital outlays in slack times, because o f the lower prices then available. This is an incentive which must, in the nature o f the case, stop far short o f fu ll stabilization, since it is only if demand is slack that prices are likely to come down. A single industry could get this gain, in an economy where the rest were still follow ing the usual practice and concentrating their capital outlays in active times. But if capital-expansion in dull times became the standard practice, the price-incentive would be destroyed. Further, it is still true that, the longer the employer waits, before a revival is expected, the shorter is the time during which his capital is tied up, idle, and the risk he runs o f installing equipment that does not embody the latest improvements. Therefore, this incentive seems to be subject to a much heavier discount than is indicated in the report. The incentives o f the employer, arising from the combination o f a guaranteed wage with social security, seem to deserve some analysis. It seems to be implied that the employer could, at his option, lim it his liability to the excess o f the guaranteed wage above the regular unemployment-benefits, by declaring his worker unemployed, and leaving him to find another job if one is offered. It ap pears that this would, in practice, involve him in little likelihood o f losing a worker through the worker’s getting another job, because it seems only natural that the employment office would give priority to job-seekers who were not covered by guaranteed-wage plans—this on the assumption that such plans are voluntary, and have not be come universal. Then if social security pays the worker, fo r example, 55 percent o f his regular 54 wages, it would pay the employer to let his worker go onto social security i f the only work the em ployer could find fo r him to do were worth less than 55 percent o f fu ll wages. I f it were worth more, it would pay the employer to keep him. In this connection, the discussion o f the loss, or cost, o f unemployment, gives the impression that it is the worker’s full wage, or its equivalent, that is lost, though elsewhere it is noted that the best employment that is available in depressed times is likely to be something that is not worth the fu ll amount o f the standard wage. Other minor points might be mentioned. The report suggests the maintenance o f adequate total demand as a goal, and takes the position that a fair approach to attainment o f this goal is a pre requisite to any general successful adoption o f guaranteed-wage plans. But no ways and means are suggested fo r a fu ll employment program nor is any estimate made o f the probable degree o f success o f any measures o f the sort which might be adopted. This is perhaps a matter in the pur view o f the President’s committee o f advisers, set up under the Full Employment A c t; but this com mittee does not seem as yet to have filled this gap. In chapter H , section 1, the report states: “It is perhaps not too hard to imagine an economic system in which the wage contract with all em ployees was more nearly like that with executives and white-collar workers than like the present pre vailing mode o f wage payment and tenure.” I agree that present modes o f wage payment show a remarkable lack o f imaginative adaptation to the facts o f modem industry. But if this means that all workers could be put on a basis o f fixed salary, it is hard fo r me to imagine this under a system o f private enterprise, in view o f what seems to be the admitted fact that private enter prise is inescapably subject to some degree o f fluctuation, and probably a fairly substantial degree, even with the best that can be done toward stabilization. It is easier fo r me to imagine a system o f cooperatives in which the workers are their own employers, and hire capital fo r a fixed return; but in that case they would necessarily take the risks that now fa ll on equity capital. Their salaries might be fixed, but they would be combined with sharing o f profits and losses. In conclusion, let me repeat that the actual recommendations o f the report appear to deserve approval. 2. COMMENTS B Y ED W A RD S. MASON Messrs. Hansen and Samuelson rightly em phasize the wide range o f variation in types o f guaranteed wage plans, and in employee coverage and employer commitments involved. Taking this range into account it is possible to conceive a situation in which guaranteed wage plans would be widely used without any appreciable effect on worker security or on the functioning o f the econ omy. This would be so if the only workers cov ered were those the employer could reasonably ex pect to retain at the lowest expected rate o f oper ations, and if his commitment were subject to fre quent revision. Under these circumstances the acceptance o f the plan would offer no special in ducement to the employer to explore ways and means o f stabilizing employment; it would offer no appreciable increase in the employees’ security o f tenure; and it is difficult to see how it would in any way significantly alter work incentives, in come distribution, or, in general, the functioning o f the economy. Even a wide acceptance o f such guaranteed wage plans would have the same negli gible effect as minimium-wage-rate legislation set ting rates substantially below the market mini mum. The problem becomes interesting and important only if plans are adopted committing the em ployer to wage payments he would not otherwise assume and offering a greater measure o f security to employees than they would otherwise enjoy. Under what circumstances would the employer be likely voluntarily to accept such a commitment? A guaranteed wage plan might be adopted if it promised to give the employer the pick o f the labor market, if it were judged to offer new and import ant work incentives, and i f possibilities existed for lessening the disadvantages o f a new overhead com mitment by stabilizing operations through diversi fication or otherwise. These are all real possibilities. But it must be recognized (a) that the advantages o f the plan depend largely on its not being extensively adopted by competitors in the labor market, and (6 ) that the stabilization effects are mainly limited to the firm and have no appreciable ad vantages for business cycle stability in the whole economy. I f all—or even a large number o f— employers accepted additional commitments under guaranteed wage plans in the expectation that their selection o f employees would be improved or that their operations could be stabilized by diversification, the expectations o f a large per centage would be unavoidably disappointed. It must also be recognized that the advantages o f a guaranteed wage plan, which may be real enough as long as the number o f such plans is small, may create additional disadvantages for the firms not adapting to the scheme. Some possibilities fo r im proving efficiency or stabilizing employment, which are net fo r the econ omy as a whole, may exist, but these possibilities would seem to be slight. It is possible that the productivity o f workers already employed may be increased by the additional security offered by a guaranteed wage plan. Hansen and Samuelson consider this possibility and, in my view, rightly discount its certainty or quantitative importance. There is also the possibility that the adoption o f guaranteed wage plans might lead to the merging o f operations which are complementary over time (e. g., coal and ice). Although this would make no contribution to over-all business cycle stability, it might lessen the number o f necessary changes in employment. Again, however, the main effects o f diversification by particular firms might often be to increase the instability o f operations o f other firms. A ll this is simply an elaboration o f the point that the main influences on stability for the whole economy lie outside the control o f single firms. Recognizing this fact, the authors raise the question whether it is possible for employers in unstable industries by acting together to contribute to stability. Although this might be done inde pendently o f guaranteed wage plans, it is possible that the common introduction o f such plans in an industry could provide an incentive for stabili zation otherwise lacking. The question is whether the causes o f instability lying outside the control o f individual firms may be to some extent within the control o f an industry acting as a unit. Com menting on the notorious instability o f employ ment and output in the automobile industry, Han sen and Samuelson emphasize that, to the extent that this instability is cyclical, the search for 55 remedies must look in other directions than toward guaranteed wage plans. “ But,” they say, “ it is not too much to hope for, and to ask o f manage ment and labor, that measures be taken and prac tices be altered to moderate the incidence o f irregular and seasonal employment and o f eco nomic idleness. This may require many farsweeping changes in the organization o f the indus try ; increases in inventories and storage facilities, limitations on undue coddling of consumers’ de sires with respect to modification o f style and model changes, perhaps acceptance o f lower hourly wage rates in exchange fo r higher annual earnings.” It is possible that action o f this sort might make a substantial contribution to stability o f em ployment in automobile production. It is prob able that the suggested changes would increase costs and prices to consumers. It seems to me unlikely that standardization, for example, could be brought about without collusive action or a degree o f cooperation among competitors be yond the limits permitted by present public policy. Perhaps existing policy should be changed. But this question raises issues o f considerably greater importance than those involved in guaranteed wage plans. It seems doubtful whether the sta bilization possibilities lying outside the control o f individual firms, but within the control o f the industry, are o f sufficient importance, in them selves, to justify the degree o f industry planning— with the concomitant necessity o f Government supervision—it would be necessary to achieve in order to realize these possibilities. In any case the question opens vistas too broad for adequate con sideration here. I f the scope o f guaranteed wage plans is pri marily determined by profit considerations and if industry-wide stabilization operations (which might broaden the opportunities for such plans) are excluded, it does not seem probable that a system o f wage guarantees w ill be widely adopted, at least in an economy as unstable as the American economy has historically been. Yet the most important and interesting questions that Hansen and Samuelson discuss assume a substantial cover age o f the economy by guaranteed wage plans. It is possible that, if accompanied by effective counter-cyclical stabilization measures, the cover 56 age o f wage guarantees would in fact be substanti ally increased. This w ill be discussed presently. The important questions referred to above have to do with the changes in aggregate demand and its stabilization over the cycle and with the effect o f the increase in overhead costs involved in guar anteed wage plans on pricing policy. W ith respect to the first question, the authors point out that the level o f aggregate demand is not likely to be raised by guarantee o f wages unless such guarantees transfer incomes from profits to wages and unless these transfers increase marginal propensities to consume. They are properly dubi ous about such effects. Even if wage guarantees clearly worked in this direction, it would still be difficult to see why employers would voluntarily accede to them unless we suppose a general under standing and agreement among businessmen that their own long-run best interests are served by the higher and more stable level o f employment that the acceptance o f a lower rate o f profit might be supposed to entail. It seems probable that the “ Keynesian revolution” has not yet been this successful. The effect o f guaranteed wages as a stabilizer o f aggregate demand over the cycle depends pri marily on the method o f financing. I f reserves accumulated in good times are paid out in bad, some stabilizing effect w ill be produced. But un less firms can be supposed to undertake commit ments o f a character and magnitude that would substantially increase business risks this effect is bound to be small. The authors recognize this fully and their tentative support, in the appendix, o f Government grants-in-aid to guaranteed wage plans, stems from this recognition. In part the case for grants-in-aid depends on the expanded coverage by guaranteed wage plans that might be expected to follow from the adoption o f grants; in part, the case rests in the fam iliar argument for deficit spending in depression. T o the extent that deficit spending is the remedy, this type o f spend ing has to be justified as against other claimants. The main advantage, stressed by Hansen and Samuelson—though not in these words—is that it can be “ built in” to a functioning economic system rather than being hastily improvised as so many kinds o f Government depression spending are. Although a number o f practical difficulties (hav ing to do mainly with inequality o f treatment o f different firms and industries) are obvious, the suggestion merits careful attention. W ith respect to the consequences o f guaranteed wage plans for price behavior, the conclusion o f the authors appears to be that the conversion o f vari able into overhead costs would probably have only a slight effect. W hile this conversion might, un der competitive pricing conditions, be expected to lead to greater variation in prices over the cycle, they think that, in industries likely to adopt guar anteed wage plans, market structures are such as to favor the use o f “ full cost” pricing policies. I f, therefore, it can be shown that such plans increase the “ full cost” per unit o f operations over the cycle, the effect on prices might be considerable. Regardless o f the extent to which firms may or may not undertake fu ll cost pricing, it is clear that any substantial increase in the relative im portance o f overhead costs w ill accentuate the dan ger to profits to be expected from competitive price cutting. The tendency to stabilize prices, by agree ment or otherwise, w ill thereby be encouraged. Insofar as the effects are limited to an increased resistance to price declines in depression, the writer would agree with Hansen and Samuelson that the weight o f modern opinion considers such effects to be contributory to stability in the whole economy. I f “ cooperation” among competi tors is seriously promoted through a widespread adoption o f guaranteed wage plans, it is improb able, however, that joint action would be limited to resistance against depression price declines. It is clearly possible that the substantial conversion o f variable into overhead cost inherent in these plans might, through the increased incentives it offers to industry-wide planning, considerably re duce the scope o f desirable competitive adjustment in the economy. It must be admitted, however, that this question is highly speculative. It is the view o f the authors o f this report that guaranteed wage plans are unlikely to play a sig nificant role except within the framework o f an effective national policy directed toward the main tenance o f a high level o f employment. In the un stable American economy o f the recent past, the advantages to the employer o f lay-offs on short notice are o f such importance as to make it improb able that more than a small number o f firms w ill adopt such plans. Furthermore, as already em phasized, the advantage o f these plans to the firms adopting them are likely to be substantial only if the number remains small. I f an effective national employment policy can be assumed, the situation is different. Even under these circumstances, however, the authors caution against exaggerating the merits o f a guaranteed wage program. They state explicitly that “ if it ever had to come to a choice between guaranteed wages and other important depression and anticyclical public and private policies, * * * then it would be guaranteed wages that would have to be rejected.” It is a great strength o f the report, in the eyes o f this reviewer, that the relative merits o f guaranteed wages as against other reform pro grams are carefully weighed. In a situation in which the stimulus to reform is strictly limited it is important to establish a carefully considered priority o f objectives. I f a high level o f employment were expected to be maintained the coverage o f voluntarily initiated guaranteed wage plans might, for a number o f reasons, be very substantial. A much larger num ber o f firms could apply such plans to a much larger percentage o f their employees without an appreciable increase in business risk. Employers, faced with a continually serious problem o f re cruiting and maintaining a labor force would find that the relative merit o f guaranteed wage plans as compared with the privilege o f lay-off at short notice had measurably increased. Although the authors find, even in the recent past, little evidence that instability o f employment is compensated by higher hourly earnings in unstable industries, it might be expected that with a high and sustained level o f employment this condition would be recti fied by competitive forces. I f this were so, firms, even in unstable industries, would probably find it possible to adopt guaranteed wage plans without incurring a large increase in labor costs. On the other hand, although the coverage might be much larger, the importance o f worker in security under such circumstances would be much less. The authors, nevertheless, believe that guar anteed wages might make an important contribu tion to stabilizing “ seasonal and other short rhythms in production and by equitably allocating the economic burden o f the residual amount o f unavoidable idleness! and unemployment.” B y shifting the burden o f seasonal and short-term unemployment to the shoulders o f employers, the 57 incentive to integrate operations having comple mentary time periods and to undertake industry wide planning to increase stability o f output would certainly be increased. As we have seen, there are real disadvantages involved in this process that must be set off against the advantages. In addition the substantial lessen ing or disappearance o f worker incentives to seek other employment during slack periods would handicap short-term adjustment. As so often is the case in the area o f economic policy, there seems to be a real conflict between security and efficiency involved in this question o f guaranteed wages. It is not obvious to the writer that, even if the main tenance o f a high level o f employment could be assumed, the probable disadvantages o f a largescale acceptance o f guaranteed wages would not outweigh the advantages. 3. COMMENTS B Y SUMNER H. SLIC H TE R The guaranteed wage is a proposal to impose an additional liability upon business owners. A l though the proposals vary substantially, they have one common characteristic—the liability to pay fo r labor which would not otherwise be used is one which would have to be met in the main, not in periods o f expansion and optimism, but in periods o f contraction and pessimism. Further more, the amount o f the liability in most indus tries would be conjectural in the extreme because the severity o f business cycles cannot be foreseen. This second characteristic could be altered some what by modifications in the plans. In fact some existing wage or employment guarantee plans provide ways fo r reducing the employer’s liability in the event o f severe depression. Finally, the liability is o f such nature that it is increased in amount by the very conditions which decrease the business enterprise’s ability to meet it out o f cur rent income. Hansen and Samuelson consider two types o f plans: (1) Pay-as-you-go plans, and (2) reserve plans in which, however, the liability is not limited to the amount in the reserve funds. I believe that a third type o f plan should be examined, namely reserve plans with liability limited to the amount in the reserve fund. Hansen and Samuelson make a suggestion closely akin to this in their recom mendations. They suggest that the liability o f employers be limited. They do not, however, sug 58 gest that it be limited to the amount in the reserve funds. Three basic questions are raised by wage guaran tee plans: (1) Their effect upon the distribution o f incom e; (2) their effect upon the utilization o f resources; and (3) their effect upon the business cycle and the growth o f industry. In general I agree with Hansen and Samuelson with respect to the effect o f wage guarantees upon the distribu tion o f income and the utilization o f resources. W ith much o f their analysis o f the cyclical effects o f wage guarantees I am in agreement. There are, however, some points o f difference. Conse quently, I have confined these notes to a discussion o f the cyclical effects o f wage guarantee plans. Before I start this discussion I wish to call attention to the fact that wag'e guarantee plans would require during boom times a larger spread than would otherwise exist between the price o f labor and the prices o f goods. This would be necessary because o f the risk that employers would be compelled during slack periods to make wage payments which are not covered by the value o f the services rendered. The greater spread between wages and prices during boom times might be accomplished either by a rise in prices or by a slower rise in wages. It is uncertain which o f these ways would be more important in bringing about the greater spread between wages and prices. .I f one assumes that the amplitude o f boom and depression is unchanged, the distribution o f in come over the period o f the whole cycle would not be significantly changed. W age earners and con sumers would fare worse than now during boom periods and better than now during depressions; business owners would fare better than now dur ing booms and worse during depressions. Pay-as-you-go wage guarantees Pay-as-you-go wage guarantees would accen tuate business recessions. This result would occur fo r several reasons. One effect o f pay-as-you-go plans would be to stimulate managements to strengthen the liquid position o f their enterprises at the first signs o f contraction in business. Busi ness concerns, o f course, now do the same thing— especially enterprises which have short-term debts or other important liabilities to meet. The very efforts o f business concerns to protect themselves against the drop in business aggravate the decline. The pursuit o f liquidity by enterprises during periods o f contraction ought to be discouraged rather than encouraged. Pay-as-you-go wage guarantee plans, however, would encourage it. A ll o f this amounts to saying that most pay-as-you-go plans could not be kept strictly on a pay-as-you-go basis. Even if assets were not earmarked and called “ reserve funds,” they would be accumulated against the liability imposed by the wage guaran tee plans. Furthermore, under the so-called “ payas-you-go plan,” the accumulation would occur just when it would do the most harm. A second effect o f wage guarantees (whether payas-you-go or not) would be to make cost-price rela tionships during depressions more unfavorable than they would be in the absence o f guaranteed wages. Since workers would have to be paid whether they produced or not, managements would be w illing to produce goods which have to be sold at prices which would not cover cost o f production computed by conventional methods. Hence wage guarantees would depress prices and make prices even lower relative to costs than they would other wise be. The effects o f wage guarantees upon pro duction would not probably be very great in oligo polistic industries. It would, however, be an im portant influence in industries where there are many small competitors, particularly the “ soft goods” industries. The unfavorable cost-price relationships thus produced by wage guarantees and the resulting drain on liquid assets would put managements un der greater pressure than ever to cut expenditures fo r maintenance, replacement, research, and prod uct development. It would also discourage the starting o f new enterprises. Even in severe de pressions there are many thousands o f business births each year and substantial expenditures on capital goods for replacement, reduction o f costs, and expansion by established enterprises. Special mention should be made o f the effect o f pay-as-you-go plans upon the operation o f the credit system during depressions. The liabilities imposed by these plans would make banks more eager to reduce outstanding credits to enterprises subject to the liability and less inclined to extend new credits. Thus, pay-as-you-go plans would be a powerful influence in enforcing a contraction in demand deposits. The effect o f pay-as-you-go plans upon the de mand for labor and upon wages would be unfavor able. Naturally enterprises which had wage guarantee plans that were about to expire would seek to anticipate their orders for goods after the expiration o f the plans by making large quantities o f goods before the expiration o f the plans. Thus plans which were about to expire would build up inventories o f partly manufactured and manufac tured goods that would prove troublesome. Cer tainly the fact that this process was going on and that inventories were overhanging the market would discourage both the starting o f new enter prises and the execution o f many projects by exist ing concerns. The fact that employers who had built up large inventories under the stimulus o f wage guarantees would be in an unusually strong bargaining position in negotiating renewals o f their contracts with unions would introduce un certainty into the whole wage structure o f the in dustries affected. These several difficulties might not be serious if the depression were mild, but the existence o f a large number o f pay-as-you-go plans would in crease the difficulty o f keeping depressions mild. I f the depression were severe, pay-as-you-go wage guarantee plans would appear to be well designed to make it more severe and to weaken the bargain ing position o f unions in renewing contracts which expire during the depression. W age guarantee plans with liability not limited to the amount in the reserve funds W age guarantee plans financed out o f reserve funds but with the liability not limited to the amounts in the reserve might mitigate depressions provided business managers felt quite confident that the depression would be mild and that the re serve funds would be more than adequate. The disbursement o f the reserve funds by raising the propensity to consume would lim it the effect o f a given drop in investment upon incomes and upon employment. Furthermore, if businessmen were quite sure that reserve funds would be more than adequate, the disbursement o f the funds would make managements less inclined to cut postponable expenditures, such as outlays fo r maintenance and replacements and some expenditures for expansion. A ll o f the above assumes that managements are governed by the belief that reserve funds would 59 be ample. This would certainly not be true in severe or even moderately severe depressions. As it became apparent that the depression was going to be deeper than businessmen had anticipated, wage guarantee plans financed out o f reserve funds would produce more and more the same unfortu nate effects as pay-as-you-go plans. Furthermore, even in m ild depressions wage guarantee plans with the liability o f employers not limited to the amount in the reserve funds would produce some unfavorable effects. Some firms would be new and would have had no opportunity to accumu late reserve funds; some would have had a bad employment record. Hence, at least a few firms would be led by their liabilities under wage guar antees to strive fo r greater liquidity than they would otherwise have sought. Many more firms, as a result o f their liabilities, would be viewed as unattractive credit risks by banks and would be forced by the reduction o f their lines o f credit to curtail expenditures on maintenance and re placements and possibly to liquidate inventories at distress prices. Reserves with liability limited to the amount in the reserves Guaranteed wage plans financed by reserves and with the employer’s liability limited to the amount in the reserves would make a moderate, but never theless important, contribution to the stabilization o f employment. The limitation o f the employer’s liability to the amount in the reserve fund would prevent the wage guarantee from stimulating em ployers to attempt to increase their liquid assets during periods o f contraction. Thus wage guar antee plans o f this third type would not have un favorable effects upon the expenditures o f business concerns fo r maintenance, replacements, and ex pansion. On the contrary, the reserves would have two favorable effects upon the level o f em ployment. First, by increasing the propensity to consume it would diminish the unfavorable effect o f a given drop in investment upon income and hence upon employment. Second, by lim iting the losses o f business enterprises and im proving the terms on which they could liquidate their inven tories, the disbursement o f the reserve funds would make managements less inclined to cut expendi tures on maintenance and replacements and to suspend the execution o f plans for expansion. 60 W ould not the favorable effects o f the disburse ment o f reserves during periods o f contraction be offset by deflationary effects from the accumula tion o f reserves during periods o f expansion! I do not think so. In the first place, favorable e f fects o f the disbursements o f reserves during periods o f depression would cause prospective in vestors to take a slightly more favorable view o f investment opportunities. In the second place, the accumulation o f reserves would make invest ment-seeking funds available on slightly more favorable terms. W ould not the accumulation o f reserves during periods o f expansion lim it expansion by raising the propensity to save? There are some m ajor uncertainties in the effect o f the accumulation o f reserves during periods o f expansion. It is quite conceivable, as I shall point out presently, that the accumulation o f reserves would have inflationary repercussions on the credit system. More likely is the possibility that the accumulation o f reserves would m odify the character o f booms by lim iting the expansion o f credit and the rise in prices. Booms are periods when incomes rise because the planned savings o f individuals and business enter prises fail to meet the demand for investment-seek ing funds. The impatient demand fo r invest ment-seeking funds causes the rise o f income to be accelerated by the expansion o f credit and usu ally by a rise in prices and often by speculation in inventories. The rise in prices limits the favor able effect upon employment from a given rise in expenditures. Furthermore, after this type o f expansion has gone on fo r some time, it creates market conditions which discourage business enter prises from immediately executing long-term in vestment plans. Hence the nature o f booms under the modern credit system limits the capacity o f the economy to utilize investment opportunities created by technological change. W ere expan sions financed to some extent out o f reserve funds rather than by the expansion o f credit, the expan sion would probably go farther and the expansion would consist to a greater extent in a rise in em ployment and to a lesser extent in a rise in prices. Hence, although the effect o f the accumulation o f reserve funds would be deflationary in terms o f prices, it would not necessarily be deflationary in terms o f employment. There is a real possibility, however, that the ac cumulation o f reserve funds w ill stimulate the expansion o f credit. The reserve funds would un doubtedly be invested'largely in the short-term obligations o f the Government. The buying o f these obligations by the reserve funds would inter fere with the ability o f the banks to maintain the volume o f their investments in short-term govern ments. A t times when the business outlook is re garded as favorable, banks are averse to per m itting a drop in their earning assets. Hence, failure o f banks to replace their short-term govern ments would undoubtedly lead them to be more enterprising and aggressive in putting money to work in commercial loans and term loans. As an aftermath to the unfortunate inflationary effects o f the expansion o f bank credit during the boom, would be the even more unfortunate deflationary effects produced by the repayment o f these loans in periods o f contraction. This analysis o f the probable effects o f different types o f wage guarantees should not be concluded without pointing out that the judgment o f the probable effects o f the plan must be predicated upon estimates o f the number o f workers whom such plans may cover and to whom they might make an important difference. About one-fourth o f the workforce would not be included because they are self-employed rather than employees. A substantial proportion o f the workforce, even if covered, would get no appreciable benefit from guaranteed wage plans because they are virtually assured o f indefinite tenure under present conditions. These include nearly all employees in the public service, a substantial number o f employees in the industries and oc cupations with little seasonal or cyclical fluctua tions, and a substantial number o f additional employees who are protected from seasonal or cyclical lay-offs because o f seniority rules. A third large group o f employees work under conditions which would make a 12 months’ guarantee (or even a 6 months’ and, in many cases, a 3 months’ guarantee) impractical. These include people in intermittent employments such as most people in the construction industry, the amusement industry, the vacation industry, and the large number o f people who enter the labor market for a few weeks only during each year. As a very rough guess I would estimate that about one-fifth or one-fourth o f the entire w orkforce are employed under such conditions that a guarantee would appreciably affect their annual income. This does not mean a guarantee might not be a useful device in the case o f this fraction o f the workforce. Nevertheless, it does indicate, however, that the possible favor able effects upon the total volume o f consumer in come from all sources would be quite limited. A t the best, the incomes o f people receiving only about one-sixth o f all consumer incomes would be directly augmented to an appreciable extent by an annual wage guarantee. 4. F IN A L COMMENTS B y A lvin H. H ansen and P aul A. S amuelson The comments o f Professors Clark, Mason, and Slichter are particularly welcome and valuable because the complex imponderable factors in volved in an economic evaluation o f guaranteed wages cannot definitively be appraised by one team o f investigators. Accordingly, we wish to confine our reaction to their comments to a few o f the more important points where there still remain some substantive difference in analysis and emphasis. 1. Depressant effects upon investment due to higher wage costs.—A ll three writers emphasize— as we do too (pp. 29-31)—that guaranteed wage commitments may act as an appreciable deterrent to business investment. Taken by itself, this is a damning consideration. But it should not be taken by itself, since there may also be a favorable effect upon income when wages are maintained. There is a universal dilemma to be faced here, which goes fa r beyond the guaranteed wage. Compulsory workmen’s compensation legislation, minimum wages and conditions o f work, collective bargaining and pressure fo r higher wages, social security legislation—all o f these can be thought o f simply as a hampering drag upon “ venture capital.” But they each also have favorable aspects from the standpoint o f social and individ ual welfare, equity, and income maintenance. Realistically, they are probably here to stay; our private enterprise system cannot shirk their chal lenge. Nor, taking the long view, is it likely that it can hold the line against their further extension over future decades. 61 Fortunately, the trend o f economists’ opinions over the last two decades does not appear to be definitely toward the view that the favorable in come and purchasing power side o f the wage ques tion is overbalanced by the unfavorable investment effects. But to discuss this further would take us far afield. 2. A nticyclical effects o f reserve financing.—Be cause o f our stated belief that all reserve funds should be invested in Government bonds or equiv alents, and because it is now definitely unfashion able to invest pension funds in the private securi ties o f the employing company, we did not feel it necessary to discuss the problems raised by such private modes o f financing. High-grade Govern ment bonds tend to go to a premium in bad times. F or this reason, and because the Federal Keserve authorities have ample powers to stabilize their marketability, such securities offer the ideal me dium for reserve financing. Professor Slichter expresses concern that guar anteed wage reserve funds w ill purchase Gov ernment securities at the expense o f the banking system; in the process excess reserves would be created; thereby private loan expansion would be encouraged. This represents an instance o f our statement (p. 33) that the deflationary impact o f boom-time reserve accumulation may be partially offset by an easing o f interest rates and capital availability. A t worst, this could only be a partial offset: Partial because interest rates weaken only in consequence o f a final decline in total income. Even this partial perversity can be easily offset by having the Federal Reserve authorities do what they always should do with respect to actual or potential excess reserves during inflationary boom times. W e cannot wholly acquiesce in Professor Slichter’s appraisal o f the absence o f deflationary influences resulting from boom-time reserve ac cumulations. The reasoning o f his preceding paragraph seems to hold that an increase o f cor porate saving—for that is precisely what reserve financing means—is, by itself, an “ expansionary” factor. In our view it could be so only in the fo l lowing inverted sense: A deliberate deflationary policy which successfully offsets an inflationary boom-time situation may prolong the duration o f prosperity and fu ll employment. Only in this sense, could a “ deflationary” policy be “ expan sionary” ; and this does not deny its deflationary character or differ with the accepted view that re serve financing is anticyclical. Moreover, only in those special boom periods where employment and purchasing power tend to be “more than fu ll,” would such a deflationary policy be desirable. During depressions, replacement expenditures o f corporations often fa ll below depreciation charges; they often fall by more than the total o f corporation losses. Consequently, there is a “ run off” on fixed assets which w ill to some degree mod erate the need o f the corporation to rely completely upon previously funded reserves. Guaranteed wages may also be paid at the expense o f depression dividends. But as we pointed out (on p. 29), in the absence o f funded reserves, there may still be some harmful effects o f guaranteed wage payments upon the ability o f the firm to maintain some o f its depression gross investment, and this is in line with Professor Slichter’s discussion. 3. Guaranteed wage plans with alternative limitations.—Professor Clark (p. 53) and Profes sor Slichter (p. 60) have each suggested a modi fied guaranteed wage plan. There are an infinite range o f such programs and it is useful to explore a variety o f possibilities. In our Report we have suggested the need in many fields o f some lim ita tion upon employers’ guaranteed wage liabilities. In particular our recommendation No. 2 in chapter V I suggested the advisability o f the type o f lim ita tion o f liability developed in Mr. Latimer’s final report (whereby the liability o f any employer in a given period is limited to some agreed-upon per centage o f his total guaranted pay r o ll). U. S . GOVERNMENT PRINTING OFFICE: 1 9 4 7 F or sale by the Superintendent o f Documents, U. S. Government Printing Office, W ashington 25, D. C. 62 - - - Price 25 cents