Keynesian Economics 9780773595538

Citation preview

Keynesian Economics

Mabel F. Timlin. 6 December 189 1- 19 S~pt e m ber 1976. Pon rail by J. Lodoen (1974).

Keynesian Economics

Mabel F. Timlin

With a Biographical Note by A. E. Safarian and a Foreword by L. Tarshis

The Carlelon Library No. 107 Published by McClelland alld Stewart Limited, in association with 'h e Illsfitllle of Calladian Studies, Carlelon University .

THE CARLETON LIBRARY A series of original works. reprints and new co llections of source malerial relatin g to Canada. issued under th e editorial supervision of th e Institute of Canadia n - Studies of Carleton University. Ottawa. DJRECfOR OF THE INSTITUTE

Davidson Dunton GENERAL EDITOR

Michael G. Gnarowski EXECUTI VE EDITOR

James H. Marsh EDITOR IAL BOARO

B. Carman Bicke non ( His/o,y) Dennis Forcese (Sociology) David B. Knight (Geogrophy) J. George NeuspieJ (Law) Thomas K. Rymes (Economics) Derek G. Smith (Alllh,opology) Michael S. Whjttinglo n (Political Science) © McClelland and Stewart Limited, 1977

ALL RIGHTS RESERVED.

0-771 0-9806-5 Keynesian Economics was first published by the University of Toronlo Press in 1942. This edition appears with th e permissio n of the

Un ivers ity of Taronro Press: The Calladion Publishers

McClelland and Stewan Limited 25 Hollinger Road, Toronto Printed and bound in Ca nada

Biographical Note Keynesian Economics was published in 1942, the work of a woman assistant professor of economics_who began graduate work in economics in her forties and whose department. in the precedi ng three decades. had graduated sixteen Masters in eco-

nomics and not man y more undergraduate economics majors. This may be a ma Her of so me surprise [0 those who are accustomed on ly to the recent age of relative amuence in Canadian socia l sciences, all the more so since major studies in economic theory from,scholars at Canadian universities a re still a rare occurrence. II is no surprise to those familiar with the remarkable scholar who produced this work or the rema rk ab le se tting in

which she worked. In th is nole an outline will be prese nted of Mabel Timlin's background and of her work in economics generally. while a more specific comment will follow on Keynesian Economics itse lf. This is very much a persona l no te about the author. whom I met only in 1956. J know that she will be the first to recognize this, for she left her own highly independen t personal stamp on all that she did. Mabel F. Timlin. or "Timmie" to all who knew her we ll. was born in Wiscons in on December 6. 1891. After a two-yea r teacher tra ining course she taught schoo l in Wiscons in . Her mother and fathe r both died in 19 16. She was particularly close to her mother. a woma n of great characte r who took particular pains to see that eac h of her four chi ldren was widely read . In 191 7 she moved to southern Saskatchewan in sea rch of work. After teachi ng school for a peri od. she joined the Department of Agricultural Exte nsion at the Un ivers ity of Saska tchewa n in 192 1 as a sec rerary . From 1929 to 1943 she was in charge o f th e adminis trat io n of correspo nd ence courses at the Univers it y. while also sCIV ing as rea der in econo mics. In 1935 she was appointed as in structor in econo mics. nnd received a n ap poin tm ent as Assistant Professor in 194 1. She had co mpleted he r BA in 1929 al th e University of Saska tchewan. ,]Jld her Ph.D in econom ics in 1940 at the University of Washington. She was promoted (0 the rank of Professor in 1950 and became Professor Emeritus seve n years huer. That brief ske tch can on ly hint -.11 th e enormous obstacles Mabel T im lin had to overcome a long the way. Some of these v

vi

Biographical Note

were shared by most: th e Grea l De press ion of th e thirties lasted more th an a decade on the Prairies. nol least in the pars imony with which university facu lty were rewarded . both in sa lary and pe nsions. as prices soared afte r the wa r. Ot he r obs tacles we re

specific: being a single woman in a ma n's world. without persona l wealt h. acquiring on ly late in life an advanced degree by resea rch on a difficult and conten ti ous lopic. Only grea l abili ty. great determina tion. and a prope nsity to loo k on th e chee rful side could acco unt fo r success in such ci rcumstances. She was helped co nside ra bly by he r sellin g. the De pa rt me nt of Econom ics a nd Pol iti ca l Scie nce at the U niversity of Saska tchewa n whe re she spen t he r e nt ire professiona l caree r. The

distin guished trio

l

or G.

E. Britne ll, V. C. Fowke, and M. F.

Timlin fo rm ed the core of the Depa rtm e nt for th e two deca des prior to her re tire ment. streng the ned after th e Second Wo rl d Wa r by such able yo unger sc holars as No rm an Wa rd and Ken· neth Buck ley. Th e re pu ta ti ons o f the th ree se ni or sc holars in their a reas respec ti ve ly of political economy, econom ic history and economic theory: th eir capaci ty 10 wo rk cl ose ly despite dif· fe re nces in a pproach: a nd th eir conce rn fo r the we lfa re of th eir colleagues al1d th eir st udep ts all co ntri b uted to an un us ually produ ctive and harmoni o us se ttin g in which to work. Professo r T imli n thrived persona lly a nd acade mica lly in this sett ing. During term she worked hard at classes and in Depart. mental affairs generally. alo ng wi th the maj or added b urde ns of responsibility fo r correspondence courses in the ea rl ie r years. Summers were devoted to research. to. meetings and travel. and to reading. exte nd ing her intellectual capital. la rgely. as . was com mon at the time. at grea t persona l financ ial cost. T he annual" meetings of the Learned Societies were an important occas ion each yea r both for academic reasons and fo r contact with her wide circle of frie nds. Work fo r the Execut ive Comm ittees of the Canad ian Political Science Association. the American Econo mic Association. and the International Federation of Univers ity Women furt her enla rged the scope for her grea t energy. Professor Ti mlin's contribut ions were recognized over a long pe riod. She was elected lO the Royal Society of Canada in 1951. the on ly woman member of Section II a t that time: rece ived a Guggen heim Fe ll owship as early as 1945 and a Canada Council Special Sen ior Fe ll owship in 1959-60: became Preside nt of the Ca nad ian Po li tica l Science Associat ion in 1959 -60: and was awarded an LL.D. by her universi ty in 1969 and the Order of Ca nada in 1976. As va lued as a ny of these. however. was the pleasure she acqu ired in her ho use. a sma ll COll age beaut ifully loqtted on the banks of the Saska tchewa n Rive r. en live ned by .

Biographical Note vii her be lo ved paintings. bright colou~ and flowers as well as by the personality 'of the occupant. It was her first house a nd was acquired just as she retired . She had already developed a considerable interest in Keynes' thought as a graduate student when the ilppea rance o f his great work in 1936.just as she com pleted he r gene ra l exam in a tions for the doctorate. prov ided he r with the framework she needed to organize her research. Her major work. Key nesian Economics. is a critica l synthesis of the importa nt set of ideas wh ich grew out of the General Theory of Emplo)lIIlel1l. I llIeresl and Money an d the debates whic h followed it. Her stu dy is much more than a restateme nt, for both her own ana lysis. and the effective use of concepts developed by J. R. Hicks and Oscar Lange. clarified and added to Keynesian thought. G. L. S. Shack le remarked in his review in Ecoll omica (A ugust. 1943. 260) : " In a sense this book is. I believe. a statement of the authors faith. She has found in the Ge nera l Theory so met hing w hich is both intell ec· tually fascinatin g in itself. a nd hopeful for mankind. and she has wished to restate it in her own manner. The resu lt is a boo k

which makes. plain on every page the high com petence of the author and the very great and sustained care she has used in writing it. In its possession o f clear-cuI purposes and meth od. in ils detailed thoroughness. and in consiste ncy. good architecture. and exactness and clearness of state ment, it is outs ta ndin g:' Timlin's wa rm tribute to th e memb ry of Keynes. published in the Canadiall Journal oJ Economics and Polilical Science in Au gust. 1947. shows how mu ch she a dmired him fo r his vision fo r humanity as well as for his co ntribut ions to econo mic theory. Th e th emes in her book were further de veloped in the fi rs t post-wa r de c~de through abo ut a dozen journa l articles coverin g points of economic theory a nd c urre nt pol icy question s. Of particu lar interest is an article presented 10 the 1953 meetings of the America n Econo mic Associa ti on. sha rply c ritical of pos(.\var monetary policy in Canada . and ren ecting also the d irec t and vivid style which marked much of he r writin g. Another maj or se t of q ues tion s had a lso begun to inte rest her. however. In 195 1. after resea rc h in Ottawa and e lsewhere. she published Does Canada Need More People? (O xford University Press). Dea ling mainly with econ o mic variable s. and emp hasizin g the comp le xity of these. she neve rth e less conclude d that a la rger popula ti on for Ca nada should mea n a higher phys ica l prod uct pe r ca pita and he nce highe r real in comes fo r Ca nadian s. Th is conclusion depe nd ed. as she emph as ized. o n a fa vo urabl e interna tiona l cl ima te fo r our products a nd a ppropri a te na ti ona l econo mic policies. It was a conclus io n which. while qu a lified .

viii Biographical Note

differed sha rply from that of a number of other wrilers. who were not slow to react to it. She decided the lop"ic was important enough to warrant further resea rch and embarked on a longerterm study of the volume a nd character of Ca nadian immig ra-

tion. in parti cular the extent to which it could be imputed to non-economic as well as economic factors. As usua l. she went at

her work with great energy and imagination: for example. a careful reading of the Lauri e r. Sinon and other pape rs bener to

understand the policy decisions of the period of major western Canadian se ulcment: fi rs t-h a nd slUdy o f the proced ures in our

immigration offices in Europe: and analysis of the data on post-war immigrati on patlems. While other papers show the de· tailed outcomes more fully, her methodology and the developing conclusions are perhaps best shown in her presidential address to the Canadian Po litical Science Association (published in the November. 1960. issue of the Associ:'lIion's Journal). A vas t amount of detailed resea rch h;ls been judiciously prese nted in a framework broad enough to encompass both economic and non·economic forces. Economic history and analysis combine effectively to point us to certai n conclusions: that attitudes to emigration held by other countries were as critical as attilUdes to. immigrat ion here; the soc ial and political stra in in Canada of a rapid influx of people from ab road, whet her British. Japanese, Indian or Chinese, especia lly where inad equate preparations had been made to receive them : the mi x of ideol ogy and pragmatism. indeed of naivete and opport unism, which ofte n dominated pol· icy: and. not least, the inadequacy of much of conventional ceo· nomic th eory in exp laining the outcomes. It is unfortunate for our understanding of this critical aspect of the Canadian experi. cnce, indeed for current policy questi ons, th at her vo lumin ous research on this topic was not completed. Professor Timlin's resea rch on immigration had raised wider questions about the approach of economics and other social sci· ences to the advance of know ledge and to resea rch on public policy. An opportunity arose to conside r these questions directly. Her final major work was a study commissioned by the Social Scie nce Research Counci l of Canada in 1964. wit h the cen tral task of consid ering the institutional structure for the support of socia l scie nce research in Canada. It was published by the Cou n· cil as The Social Sciences in Call(u/a: TlI'o Swdies. by Mabel F. Timlin and Albert Faucher (Ottawu. 1968). In addition. as a supplement to her repo rt. shc prepared a nd presented a Brief to the Special Senate Commitlce on Science Po licy the fo llowing . yea r. In her usual mcth od ica l way she lOok her task to be the attainment of much grea te r knowledge of the na ture of indi vid·

• Biographical Note ix ua l socia l sc iences: an investiga ti on into th e relationships among (he na tural. socia l a nd engineering sc iences: a nd a n examination of th e institutions for supporting resea rch in the social sciences in some o lhe r count ries. Intensive contact foll owe d wit h ma ny soc ial scie nce depa rl ments a nd indiv id ua l sc ho lars across Ca nada. In a major re trospective sec tion of the study Professor Tim lin deta ils the changes in government function s over time a nd the implica tions fo.r social science, the consequences o f the neglect of support for the socia l sciences in Ca nada, and the problems in (he institutional forms designed fo r support of the socia l sciences. She traces the weakness in the social sciences in Canada bo th lO the relative pa uci ty of financ ial suppo rt and a lso to an in stituti ona l structure wh ich is in capa ble of d eve lo pin g m ult idisciplinary research proj ects which serve the ends of scientific resea rch while a lso prot ecling the interests o f resea rche rs wh o wish lO work ind epende ntly. She placed particul ar emphasis in th ese studies on th e unity of fields o f knowledge, and also on the need fo r multi -discipl inary resea rch of high qu ali ty if the basic socia l iss ues we re to be successfully a u acked. There is effective use o f the concrete image here, pa rt ic ula rly on the poss ibi lities fo r frui tful interactio n betwee n the biological and the soCia l sciences, inte rac tio n whic h she saw as the key to progress bo \h in many a reas o f scie ntifi c resea rch and in tackling many socia l problems. She was critica l of the prevailing approaches to such issues. such as th e in siste nce of many eco nomis ts that transfers through th e tax system will resolve poverty. a nd th e com parative lack of work in socia l psychology in Ca nada. She saw possibilities also in mo re interac tio n between the socia l sciences and ea rth sciences. Not a ll of he r r~c omme nd a t i o n s were easily received. or implemented . Some of her fellow econo mists. for exam ple. th ough t they implied a diminution of emphasis on individual resea rch. Nevertheless. she ha d the sat isfaction of achieving her majo r goa l. a revi tal ized Social Science Resea rch Council. a nd of rece ivin g suppon in principle from the Macd ona ld Re port and the Lamontagne Report for a se pa ra te federa l council for support o f research in th e humanities a nd soc ia l sciences. I doubt if she wou ld be satisficd. despite som_c improvements. a t thc pace a t which multi-disc iplin ary research o f high quality on key iss li cs of soc ia l impo rta nce has proceeded in Cmada. In addi ti on to the three main bodies of work notc d above. there is a significant unpublished literature. much of it of quite . high quality in term s of econo mic a nd social th o ught and a ll of it wriuen in a stimulating manner. Professo r Timlin is not a person

x

Biographical Note

to let im port a nt ideas res t at the leve l of discuss ion wi th co lleagues. dearly as she loves such excha nges. J no te here only two ilems: a long leller da led April 8. 196 1. de a lin g oslen sibly wilh pe nsion s. but whi ch was in fact a sea rin g indictmen t of th e wretched financial tre a tm e nt w hic h all e ntire ge ne ration of univers ity scholars endured: and a j oint submiss ion. dated Nove mber. 1959. with olher departmen ta l colleagues. c riticizi ng a number of proposed amendments 10 th e ~o mbine s In vestiga tion Act and the C rimin a l Code of Ca nada . Bo th in her majo r published

works and in her writing genera ll y the fundamental unity is not in subj ect matter. but in approach and in phi losop hy- Ihe work. first a nd foremost. of a th eor ist who is prepa red to re neet ca refully, and at length -a nd in a deep co ncern abou t th e ways in w hi~h good schola rs hip and a ppropriate polici es could a ll evia te ' th e cataclysms visited on indi vid ua ls a nd who le communities in her ge neratio n. As a majo r ea rly co nt rib uti o n to the most im po rt ant develo pment in rece nt economic tho ught. Key"esian Economics inev itably had wide impacl bblh in Ca nada a nd a b road. Desp ile lhe development of subst ilUtes. 'th e book con ti n ued to enj oy a demand three decades afte r its publicat io n. Professo r Timlin 's in flu e nce on the d isse m ina ti o n an d evaluation of Key nes' id eas went we ll beyond publ ica tion. howeve r. At the time he r book was pub lished. it was still poss ible to take a degree with specia liz.a tio n in econo mics at a number of Ca nadian universities without a systematic a na lys is of Key nesian econo mics. That situation was cha ngi ng. of course. particularly because of the influence in o ur unive rs it ies of a few theorists such as Pro fesso r Timlin who fu lly understood th e revoluti o n in eco nom ic theory which was und er way. She wa s fortunate to have a great love for leaching. the opportunity to work in se min ars with sm a ll n umbe rs of able slUdenlS. and a cl ose relation ship with a com pleme nta ry grou p of colleagues. A se minar o r lecture class wit h Professor Tim lin W,IS a n experience which few students forgot. Fo r a significant number. their life's work was sc I. She was an enthusiast ic but demanding teache r who. through htr own st ro ng background in ecbnomics. force of character. and intensely persona l approach. made her stu de nts stretch th eir ho ri zons both in terms of econo mics and in le rms of th e ir view of the mse lves. Those students who measured up to the high standard s she se l were made to feel they had been admitted 10 a select society of pe rsons worthy of pursuing economics in the world at large. Indeed. a steady stream of ab le students from Saskatchewan were accepted by major graduate schools and went on to dist inguished economics careers in Canada and abroad . A number of these thought in

Biographical Note xi la ter life that much of their sense of profess ional standards and of commit men t to a caree r in economics had been acquired in working wit h Professo r Tim lin . And so me believed th at th at applied to thei r understandi ng of economic theory as we ll.

A. E. Safarian,

Professor of Economics. Department of Poijlical Economy. University of Toronto, May, 1976

POSTSCRIPT: The Editors of th e Carleton Library regret to say, as a postscript to Professor Safarian's biographical no te, that Professor Timlin died in Saska toon on 19 September, 1976.

Foreword The re issue of Dr. Mabel Timlin's Keynesian Economics mu st of course be recognized as a well·deserved tribute to an extraordinary Canadian economist, but it is morc than that. At a lime when interest in the Economics of Keynes-to employ Leijonhufvud 's term for dis~n g u i s hing it from " Keynesian Economics"has again quickened, having the book once morc readily available should prove to be a real convenience for economists. Hopefu lly. it wiJI stimul ate further work in the field, for sure ly it contains enough not-yet-explored ideas to provide many valua ble leads. 1 do n Ol mean lo- indeed I cannot- trespass into the territory that Dean Saracian cultivates in his Biographica l Note to accompany this volume, but I should point out the remarkable accompli.shm ent of the auth or. At the time she wrote , pro Timlin had never been in th e Cambridge which was the birthplace of Th e Gelleral Theor),. and so far as I know she had had no contact with Ca mbridge economists. She had not even se t foot in the other Cambridge which, 1. K. G albraith assures us, was the source of a ll that North Anferica ever learned of Th e General Theory, but she made up for these deficiencies by a read iness to read Keynes' book with an open mind , and with a critical and imagin ative intelligence. The result is a book which, while it does not pretend to provide an Easy Guide to Keynes. nonetheless throws a penetra ting light on The General Theory, here and there ca rryi ng it further than Keynes had d one ; and quite naturally in other sections, saying nothing and occasiona lly eve n misleadin g any reader who would approach it as though it were the New Testament. Before considering its positive merits and describing its cont ributions. it would be wise to note what the book is 1I0f. First, it is not al all a Guide to Keynes. It is incomplete, ra ther quirky and in many places. a student wou ld fmd The General Theory far easier to read than the Timlin volume itseIr Secondly. the author had no interest in showing that Keynes was really a Classica l Economist-though not a good one, as many economists have pointed out, although not always with justifi cation-who got his strange resul ts simply by virtue of the strange curves he postulated; or th at, if on ly he had been a better theorist in the C lass ica l or-Wal ras ian traditi on, his resu lts would ha ve been far more xii

Foreword xiii general. and far more significant. Nor was she conce rned to prove Lhat Keynes was either a primitive Marx ist. o r that he would have been a beller Keynesia n if only he had had a beller understandin g o f " Das Ca pit al". Dr. Timlin o b vio us ly a pproached The Genera l Theory with an open mind : an essentia l . too l for one who seeks to understand - but not '3 comm on one. Indeed. her manifest concern was to underst and wha t Keynes meant in The General Theory . and not to search out in th e pages of that volume wh at she wanted to find th ere o r what she wanted him to have meant. Such an approach. looking back on the histo ry of d octrine over the last two gen era tions. has surely been rare. This would. in itse lf, be a pers uasive reason for paying specia l attenti on to her boo k. But, obvio usly. her book has many oth er virtu es ove r and above its in tegrity. For one. it puts very much more stress on ce nain elements of The General Theory than d o most of the commentators of th e time. It stresses th e dynamic fea tures of th e Theory: th at th e eq uili b rium of \Vee k.J)ne is alm ost ce rt ain to create new conditi ons fo r Wee k Two. which will in thei r turn imply a modifica ti on of the original position of equilibrium . Instead. then . of an equilibrium which is'stal;c over time we are shown th at we mu st grapple with a moving or shiftin g equilibrium. When the Economy is seen to move through tim e in this way. we have no more justificatio n fo r neglectin g th e directi on of movemen t from yeste rday, to today. to tomorrow th a n we would have fo r o ve rlooking the facts th at th e prod ucin g units of our economy are profi t-see king enterp rises. and that the economy itself is not peo pl ed by in divid ually self-s uffic ien t Robinson-C rusoes working in a money-less and class less society. Dr. Tim lin's noti on of the sh ifting or movin g equilib riu m is richer than that co mmonly prese nted. It may be illustra ted by . the follow ing ex tra ct from her chapter on "The Rate of Interest and the De mand fo r Money" :" T here is a seco nd requisite for equilibriu m ove r time if th e eme rgin g Wee ks a re ch a ract e ri zed by positi ve in ves tm ent. Someone must be willin g to hold ti tles. d irectly or indirectly. to th e increases in the q uantu m o f capita l rights on terms which kee p the rutes of money in tcrest related to the induce me nts to in vest at th e eq uil ibri um leve l of inco me. Bu t the additio n o f new un its of capital goods changes the co nd itions of supp ly time· unit will be determinate. 'M r. Keynes himself tcnds to attach the term "psychological" to the three sets of functional relationships (see General Th eory, pp. 246-7). Some persons have impugned Mr. Keynes' analysis because they interpret his use of the term "psychologica l" as implying a cond ition stable over time or only slowly modified. (Note this assumpt ion in Mrs. E. W . Gilboy's article. "The Propensity to Consume," Quarterly Journal of economics. L111 ( 1938-9), pp. 120-40. and Mr. Keynes' re ply in a letter included by her in another note in the same volume. pp.633ff.) In this study, th e term " psychologica l-inst itutional complex" is used to describe the totality of th e effects determ ining the shapes of these functions per unit of time in order that we may escape such imputations of meaning. Between units of time. changes in the sha pes o f the functions and in their positions will be imputed to changes in this complex. It is hoped a lso that th e method used here or holding lhe fun ct ions st able for defmed units of time and describing by degrees the effects of various forms o f c ha nge will have the virtue of making meanings clearer.

10

KEYNESIAN ECONOMICS

Limitations Set on 'he .Analysis. It has been stated in the pre· ceding paragraphs that the Keynes-Lange sy!tem is determinate if "the effective q uantity of money" is given . What we mean by the "effective" quantity of money can be plain only after we have defined the units of t he system. We wish in this place. however, to enter the caveat that we shall not in this study undertake the analysis of the whole system of economi c relations. The specific problems involved in the determination of specific prices (and hence of specific sectional price levels) will in general be ignored.' The in stitution al a rrangements wh ich serve to determine the elasticity of the supply of money ace likewise ord inarily passed over. We omit these important areas in order to explore more thorough ly the partial system we have set out above. We shall make it determinate under the assumption that the effective quantify of money is given. In the next section, we proceed to define the money-unit, the wage-unit, and the time-unit. In connection with the latter we shall explore the concept of temporary eq'uilibrium and describe the models with which we shall work. (2) THE UNITS AND MODELS OF THE SYSTEM

The Mrmey-Unit. One unit for the measurement of our system will obviously be the money-unit, understood as the standard unit for the bank-deposit. Since we shall carry out this' analysis fundamentall y in terms of a closed system, there are no practical compliCations in the understanding of the meaning of this unit. But since the effects of current or anticipated changes in the quantity of money wiU constitute one of ou r chief problems for analysis, it is necessary to find a second unit of measurement which will have none of the characteristics of money. We shall adopt for this purpose the unit which Mr. Keynes has devised, namely the wage-unit.' The Wage- Unit. For purposes of defining the wage-unit, an hour 's employment of common labour is taken as basic and is called the labour-unit. The remuneration for the labour-unit is the wageunit. ThuS" if common labour receives fifty cents an hour, fifty cents is the wage-unit. Other types of labour are reduced to a common denominator with common labour upon the bases of t heir relative remunerations. If plasterers are receiving three dollars an hour when the wage-unit is fifty cents, one hour of plasterer's labour is rated as six labour-units. A sum of three dollars will be equal to 'For the theory of price [ormation, see Book V of th.e General Theory.

'See Gel/erot Theory, pp. 41 -5.

CHARAaER OF KEYNESIAN SYSTEM

11

six wage· units if we are measuring sums of money in wage·units on this basis. The expecta tion will be that rises in wages will be rela tively the same. If the wage of common labour ri ses to seventy-five cents an hour, it will be c?,pected t hat plasterers will be earning four dollars and fifty cents an hour. If this happens, the wage-unit will be seventy-fi ve cents, and four dollars a nd fifty cents will be equ al to six wage-un its as t hree dollars was before. If, in fact, the condition with respect to the sta bili ty of the relative remunerations is not 'fulfilled, t his si tuation is to be regarded as "a rapid liabili ty to change in the supply of labour and the shape of the aggregate supply fun ction. "e A furth er qualifica tion of the defin iti on requires to be made. It is clearl y understood that em ployers, so far as they can, wi ll tend to employ the most able workmen fi rst, and that the last labourers coming into employment at the same rate of pay are likely to differ in efficiency from the earlier ones, who secure more stable employment as a reward for thei r superior efficiency. Of this we dispose in our definition by subsumi ng the decline under the decli ne in the efficiency of the equipment due to the operation of the principle of di mi nishi ng returns. 7 The wage-uni t is a much less objective concept tha n is t he concept of the money-unit, but I t hink at least one advantage may be urged for its use. Th is is the one set out by !vl r. Hugh Townshend in a note ent itled "Liquidity Premium and the Theory of Value .. " M r. Townshend points out that "labour carries no liqu id ity premium";' that is , since labour cannot be used as a store of value, it cannot be used as a substitute for money as a store of value: therefore there can be no varying premium as a price for giving up possession of a stock held to satisfy a desire for certainty. Sin ce the wage-unit sha res none of the functions of money, the two uni ts a re logica ll y independent of each other. T he value of th is independence in our unit will be much more apparent when we have consid ered some of the problems wh ich Ilbid., p . 43. ' [ Md., p. 42. 'Economic Journal, XLVII (1937), pp. 157-69. This.a rticle trea.t l the storeof-va lue or liq uidity-preference attri b utes or dura ble a.ssets and money-cla ims much more extensively t han we shall do here and is an import an t a rt icle in this litera ture. Ilb id., p. 166. l"lr. Townshl'ud a lso add s (p. 165) " the l'normous rl'sistance to either rises or fa lls in mone),-wagl's" a s a reason for makinG the a ssumpt ion o f wage stabilit y thl' bl'st basis for a convention of price stabi lity.

12

KEYNESIAN ECONOMICS

arise out of a change in t he q uantity of money in terms of moneyunits. Speculation in many commodities as stores of value is a familiar phenomenon of cyclical change in the world we know. Where other goods in an exchange system can be used as stores of value to varying degrees ("have liQuidi ty-preferen ce schedules"), the system is imperfectly monetized and the money in use is not a "perfect money" but shares at least one of its functions wit h other goods in the system. A change in the attitude toward money as a store of value ("a change in the liquidity-preference schedule (or money") will be accompanied by a change in the attitude toward goods able to perform the same function and may result in a pervasive change throughout the value system. IO Therefore when we wish to measure quantities of money,. employ men t, consumption, and investment in a very limited field of relations over periods of time, the use of a unit which entirely escapes liquidity·preference cha nges will enable us better to account for change and will relieve us of a very baffling complication of the whole problem, at least temporarily. We have now established two units for ou r system , namely the money·unit and the wage·unit. Since chan ges in the wage-unit will tend to be relatively discontinuous, the period ic revaluation of the wage·uni t in terms of money will not present a great problem. The wage·un it will be the fund amental unit we use in presenting the system. In our system of equations set out in the previous section, a ll symbols will be referred to measurements in wage-units unl ess otherwise stated. Thus for us, within the system of the equ ations, "the effective quantity. of money" will be the quantity measured in wage-units. Through the use of the wage-unit and the money -unit we may set up parameters for the system of functions, but the limits of usc UP. N. Roscnstein-Rodan, "The Co-ordination of the General Theories of Money and Price," E,onomica , III n.s. (1 036) , pp. 257-69. Mr. Keynes thinks that an ideal cost-unit which would be found by taking "3 weighted average of the rewards of factors entering in to marginal pri me·cosu'· might be a better unit than the wage.unit (Gelleral T hrory, pp. 302·3). I have some reservat ions on this matter . . Such :t unit would perhaps bring this parameter of our system into closer unity wit h changes in the nationa l di vidend ; but there is to be urged a gainst it the (acl t hat the concept of uSUCOSt which ~1r. Keynes includes among prime costs has very definite subjective elements based upon t he st ~ t e of expc.·ctat ion wit h respect to future changes in prices. intc re~l rates. quantiti~5 ~..lIable, and upon the probabi lities of C\'ent s (war, peace, weat her, etc.) occurring to affect t hese. The unit would therefore appear to have less independence of the !'it u· ation which it is presumed to measure than does thc w3ge-unit.

CHARACTER OF KEYNESIAN SYSTEM

13

for a given system of fun ctional relations will not, however, be apparent until we have defined our unit of t ime. The Unit of Time andtlte Fundamental Model. We seek to establish our unit of t ime in such a way t hat we may assume the funct ions L, q" and F to be stable for a given t ime-unit. For the flowing rivers of change in the real world we must therefore substitute some' type of convention which will give us moments of rest in which we may study the system of economic relationships to see how they determine each other and how they are in turn affected by changes in the psychological-institutional complex. To do this, we shall Set up three model worlds. The first we shall call our "Fundamental Model" and to this model we shall give practically all ou r attention. In t his model we shall study the world in terms of a series of temporary equilibria, a convention which is implicit in the General TheoryLl itself and which has been made expli cit and expanded by both Mr. Lerner '2 and Professor Hicks.l~ Like t he latter we shall call our t im e·unit the Week. The other two :nodel worlds, to be called t he First and Second Supplementary Models, will be characterized by t ime· lags to be described in their place. Many of the attributes of t he Fundamental Model are inherent in the'nature of t!'te concept of the \Veek. Because Professor Hicks' description fits very conveniently into this model, t hough perhaps it is more precisely developed than we require (sin ce we shall be dea ling with quite macroscopic quali ties of the economic system), we can perhaps set out the qualities of t his time·unit no better than by quoting very extensively from Professor Hicks' own description ofit: 14 . . . I shall define a week as that period of time during which variations in prices can be neglected. For theoretical purposes this means that prices will be supposed to change, not continuously, but at short intervals. T he calendar length of the week is of cou rse qui te arbitrarr; by tak ing it to be nPage 00. "A. P. Lerner, "Saving and Investment: Definitions, Assumptions, Ob.jectives" (Note), Quarterly J ournal of Economics. LIB 0938-9). pp. 6U-19. U) . R. Hicks, Value and Capital (Oxford. 1939), chap. IX . Professor Hicks devises his concept for the purposes of his own analysis hut owns that the " latter half of this book t Value a'id Capital] would have been very different if I had not had the General Theory at my disposal when writing. The fina l chapters of Part IV, in particular, are very Keynesia n" (p. 4). It is ror this reason probably that the Hicksian concept of the time-un it , which" we shall take over wholly for ou r Fundamental Model and mooify only a little ror the Supplementary Models, suits our purposes so well. UAbridgcd from pages 122-'l inclusive of Valli e and Capital.

14

KEYNESIAN ECONOM ICS

very short, our theoretical scheme can be fitted as closely as we li ke to that ceaseless oscillation which is a characteristic of p rices in certain markets. 1 think we shall find, however, that when the week is supposed to be very short oue theory beCti ng the development of the d the demand function Y•. The marginal liquidity preferences which will be satisfied will give liquidity premium ill for the system of multiple margins. As long as the current set of relations giv(>n by Figure XXII I holds t rue,

THE SYSTEM O F THE SHIFTING EQUILIBRIUM

157

the complex of interest·rates connected with this":o rate will form the bases of capitali zation in the debts and equ ities markets. In the production market,' this comp lex of rates will be the system of io and connected r' rates for wh ich investors Il).ust make a llowances before they reckon their own net returns over cost upon expenditures of fun ds for maintenance or new investment. At t.he new temporary equ ili brium, the levels of in come and consumption will therefore be Y. and e•. and the level of investment for the new Week will be I .. a ll as shown on Figure XXII. We must make some qualifications of ou r reason ing for t he Supplementary IVlodels. Under the conditions of t he Fundamental l\Jodel, t he sy~tem will p~ed smoothly to the. new solu tion because of t he degree of market perfection assumed. But under the .cond itions of the Supplementary Models there may be an in terva l of ad justment due to incorrect expectations of .entrepren eu rs. It may take a se ries of Weeks to reach t he a djustment. Over t hat series of Weeks the price- and income-distribution effects of the errors in judgment may somewhat disturb the shapes of the fun ctions L, q" a nd F which would be associated with " correct" expectations. It is also probable that t he errors themselves will become t he bases of revisions in expectations wh ich will bring cyclical changes into operation. In the case we set o ut here of a n in crease in the quantity of money , a cyclical expansion may be se t off', which must, of course, be supported by further e.xpansions of the supply of money if price-expectations and the size of the wage- unit are not to be changed. It is evident that whi le a cha nge in the quantity of money may ra ise the level of in come and employment, the effects of such a change can not be exactly predicted even under situations so highly simplified as are those of our model systems. Moreover, when we consider the effects of· cha nges in the quantity of money measured in money-u nits, we must not lose sight of t he limi ts set upon the effects by the existence of a minimum conventional io rate of interest and by the possibilities of revision in the size of the. wage-u nit. The structure of our analysis applies, as we have used it, essentially to the world of under-equilibrium in the orthodox sense of equilibrium ana lysis. Its precise relationship to conditions of full eq uilibrium we have not undertaken to develop in t his study. Before we leave the nature of t he structure as a whole, we should , however, devote sollle space to an ana lysis of t he nature of repercussions within the system.

158

KEYNESIAN ECONOMICS (3) REPERCUSSIONS

We shall take up first in our discussion of repercussions the effects of an increase in the quantity of money, measured in wageunits. upon the part of the psychological-institutional complex which we have called the "state of expectations." The effects upon expectations in terms of the functions L, 4>. and F may serve either to increase or to decrease the effects of monetary action as these have been set out in terms of the change in the supply of money from AfT to AI. in the preceding section. We shall examine the effects first under cond ition s where the outcome is to reinforce the ' effects of an increase in the quantity of money . An increase in the quantity of money through open-market operations of'the central monetary system, if it comes at certa in periods of the business cycle, may cause the Liquidity Functions associated with the various potential levels of income to shift downward. To have such an effect, the action of the monetary authorities must increase the optim ism of the comm,unity with respect to one or more of the future levels of economic activity. prices or interest·rates. Because of this optimism, " bears" may become "bulls" at lower average rates of interest than formerly. This means that there will be a tendency to release money from the L2 balances for the purchase of debts at lower rates of interest. Release of money from the L2 balances wi ll help to provide new funds for the Ll balances and the downwa rd sh ift of L, will thus act as a deterrent on the rise in the function L and in the rates of interest otherwise accompanying the rising level of activity . The interest·rate complex would fall to a lower point than would be possible if the original functions held their position. The effect of the initi al increase in the quantity of money upon the level of economic activity will be intensified. Alternatively , if changes in the other functions wo uld, even with the increased quantity of money, bring a tendency to raise interest· rates , the change in the position of the Liquidity Functions will offset part if not all of this potential rise in ra tes. Simila rly, if the ini t ial increase in the quantity of money and the resultant fall . in the interest·rate complex actually sta rt an upward expansion in the system, as the expansion gathers force, better prospects for employment may shift upward th e shea f of fun ctions representing the pro pensity to consume. Cyclical plans for expenditure may be revised. The new car may be bo ught this year rather than next. The l\Iultiplier a ssociated with given ia

~

THE SYSTEM OF THE SHIFTING EQUILIBRIUM

159

rates of interest a nd levels of output may be raised and this will intensify the operation of the increase in the Qu antity of money. On the Investment Fun ctions, ~e initiation of an expansion may change the nature of expectations with respect to the Cuture yields of existing capital goods and hence support larger expenditures of funds for maintenance and replacement and new investment activity for a given level of consumption in the presen-t . Thus the In vestment Functions wi ll sh ift to the right. If the process of expansion stacts an expectation of rising prices, that factor a lso may raise the In vestment Functions. These shifts of expectations in tum will tend to increase the effects of the original increase in the Quantity of money. What we get in effec t is a new set of fun ctions. Let us take our structure for the sh ifting eq uilibriu m as shown on Figure XXII, and on Figure XXIV show a shift in the positions of th e functions a rising out ma rc optimistic expectations. The effects of the shift are obvious. Ri sing expectations have prevented the rate of interest from fallin g but have supported an increase in the levels of em ployment and output. We get interest- rate ic in this cast as we did for Figure XX I with the origin a l amount of money M,. but we get income Y t • 'consumption Ct. and investment 1 t instead of Y., C .. and I,. the levels formerly connected with quantity of meney MI' If the impulses towa rd expansion in the system had been stron g enough, we should have had the rise in the interest-rate which is so characteristic of the expa nsion phases of th~ cycl~ as to have been given a name of it s own, namely the "Gibson Pa radox ."l If the centra l b'a nk should endeavour to bring about a deRation by reducin g the qua ntity of money. t he effects may be reversed. We might in such a case regard a reduction of the quantity of mQ.ll ey from _~{, to Af, as bringin g the level of income fr om Y t to Y, because of an upward shiftin g of th e function L and downward shiftings of q, an d F. Effects of changes in t he quantity of ritoney may not cause an actual movement of the fun ctions of the system bu t they may arrest or slow down a movement which ex ists for other reasons. Thus active central ban k operations may provide funds which keep the "secondary" recessions of a cycle from developing beca use of support given member ba nks in extending loans at relatively

oc.

IOn p.1ges 198-208 of volume II of the Trealise 0'1 }.{orl~Y. Mr. Keynes gave a description of Ihe Gibson Paradox and an expianation in terms of the "st ickiness" of the "market rate of interest." The explanation .l iven above in terms of the st ructure of the General Theo ry. appears to be morl! logica l.

,60

KEYNESIAN ECONOMICS

low rates. But whatever the nature of the changes, for any definite values for the function s there will again be a determinate solution for the equations of the system.

Y, Y,

Y.~ ~:I====~~'~~.:~ Y: Y'

yl

I

O'L-----------M~.---M~.~-M

(1)

c,

c

.~----------~--~----~Y, Y (1) F IGURE

XXIV

We have no t exh a usted t he s ubj ect o f repercussions in the system when we have considered changes in th e e ffecti ve quan t it y of mon ey. Whatever t he na ture of th e ch ange , a utonomous or otherwi se. if it is interpreted o pti mistically, t he effects will be to lower t he Liq uidi ty Fu nct ions. or to ra ise the Multi plier o r Invest-

THE SYSTEM OF THE SHIFTING EQUILIBRIUM

161

mcnt Functions, or some combin ation of these effects. If a cha nge is interpreted pessimistically, the effects upon the positions of t he fun ctions will be reversed. An impact upon one set of functions which does not affect the other initi ally may be transmitted to them over a series or Wee ks as the favou rab le or unfavourable solutions of the system become pa rt of its history. Brief consideration will be given in the last chapter of this. study to some of these impl ications in a number of fields of socia l policy. In the meantime, we turn in the following chapter to some of the relations between t he structure we have se t out and general orthodox equilibrium theory.

11

CHAPTER XIII THE FIELD OF THE SHIFTING EQUlLffiRlUM

I N the fo regoing twelve chapters we have completed t he ma in st ruct ure of t he a nalysis which it was inten ded to present . The one rema ining task is to ind icate somewhat more cl earl y than we have yet done the connection of t he fun ctional relationships wit h orthodox equ ili bri um ana lysis an d in particulaJ: wit h th e theoretical level of employment. In pe rform ing t hi s task , we must go back to t he analysis of Chapter I V. In the first section of the cu rrent chapter we shall consider t he determ inants of t he level of employment in a society which has reached fu ll stationary equilibrium under the conditions of F igure 1. 1 In the succeedirli section we shall discuss t he dete rminants of the level of employment in a n econom ic society which has reached stationariness under condit ions where t here is a long-period inconsistency between t he ma rginal propensity to save an d t he inducements to invest at t he level of full employment. We retu rn here, ~ h ere fore, to t he condit ions ana lysed in connection wit h F igu res III a nd V.2 In t he third and last section we shall attempt to move t h rough t he cond it ions of the F irst Supplementary Model to a place from which we may get a glim pse of t he condi tions of t he real world. (1) EM PLOY M ENT I N THE ST ATIONAR Y STATE (FIRST TYPE)

The Description oj Full Equilibrium in the Stationary Siale (First T ype). We begin the definit ion oJ the relations of the shifting equil ibrium to t he level of employment under the conditions of the stationary state wit h a positive rate of interest by reviewing t he conditions of such an equilibrium as expressed or implied in Chapter I V. The stock of capital is assumed to be ON, as shown on Figure I. "The" interest-rate is NN' as shown on the same figure; that is to say, all rates of interest in the interest st ructure wi ll be equa l to each other and to NN'. The level of saving per Week is zero. T he quantum of property tides is ad justed to the demand for these ISuprD, p. 39. lSupra, pp. 45, 49.

,62

THE FIELD QF THE SHIFTING EQUILIBRIUM

163

in such a way that t he saving by new savers per Wee k is eq ual to t he dis-saving by old savers or thei r heirs. Activity in t he investment-goods indust ries is only sufficient to main ta in t he qua nt um of capita l in t he system intact. Activit y in t he consumption-goods indu stries is equa l to consum ptioQ fo r t he Week . If goods of some kind s a re wi thdrawn from hoards of liqu id capi ta l for t he Week, equi valent amoun ts of consum ption goods are added to other hoards of li quid capital. There are no hoards of money as such. The stationa ry sta te is assum ed to have lasted long eno ugh for the L , balances to have been absorbed in to the L t balances and the only demands for money are those of the types analysed in Chapter v. All enterprises are operati ng under conditions where ma rgina l costs are equal to ma rgin al revenues with no unemployed factors of production. If a single enterprise a ttempts to push outpu t beyond the level associated wi t h condition s of equilibrium or if the attempt is made to establish more than the equilibrium number of enterprises, factors can be secured for such activities only by bidding them away fr0111 other enterprises, a proceeding whi ch will not be worth while because it will increase costs more t ha n it wi ll increase revenues. On Figure XXV we establish t hese relationships. On th e vertical axis of t hi s figure we measure t he rea l wage-rate. On t he horizon ta l ax is we measure units of employment in la bo ur-uni ts. The supply curve fo r labou r for a Week uncler t he condi tions of t he stationary state we set out upon the basis of Figure I on page 166 of Mrs. Robinson's Essays i n the Theo ry of Employment. T his curve assumes t hat at very low wage- rates , la bo ur su pply will increase as real wages are raised but that at some level of t he rea l wage-rate t he supply of labour will be at a maxim um and that at ,. higher real wage-rates t here will be a d isposition to prefer increased leisure so t hat as the rea l wage- rate rises, the fun ction shows negati ve elast icity. The fu nction is related to responses of labour to real wage-rates und er conditions of full employment. The demand curve, DqD' , is t he curve wh ich represents t he demand for labour which will be associated with capital stock ON of Fig ure I , wit h al\ fo rms of capita l presumed instantaneously, adjustable to t he supply of labour. Cha nges in the rates of interest wh ich will be associated with each level of employment are subsumed in the form of t he fu nction. The point q gives the leve l of effective demand for t he system when t he in terest-rate is NN' (Fi gure I) a nd a ll firm s ha ve adjusted out pu t to "correct" expec ta tions in such a way tha t for each Wee k's ope rations ma rgina l costs

KEYNESIAN ECONOMICS

16.

are equal to marginal revenues and both prime and supplementary costs arc covered for all finns. Because WC"'&fC assuming "correct" expectations, we may also assume that the form of the demand curve will be dominated by the operation of the principle of diminishing returns in such fashion that "in general, an increase in employment can only occur to the accompan iment of a decl ine in the rate of rea l wages. "S At a later place in th is chapter we shall set for th some reasons for thinking that the reasoning of the

UI

D

UI,~----------------------~~q

"

" ....

'D'

O~----------------~-------N N, FIGURE XXV

Keynesian system itscH gives ground for challenging the complete validity of th is gcnt;!ralization except under stationary conditions. The pertinent dema nd for labour in this system under these assumptions is thercCore Dq. The function s of the shifting equi'General TJuory, p. 17. Dr. Jacob Viner challengw this position in his review article which was included in the symposium upon the General Throry in vol. LI of the Quarterly Journal 0/ Economiu. Sec "Mr. Keym::s o n the Causes of Unemployment," pp. 149·51. A considerable volume of eriticalliteratu rc has grown up about the problem of the relation of money wa~es to economic fluctuation s. A list of some of the main articles, other than the art ide by Dr. Viner named above, is given below: R. B. Bangs, "Wage Reductions and Employment," J ournal 0/ Political Economy, L (1942), pp. 251.71; J. T. Dunlop, "The Movement of Real and Money Wages," Economic Journal, XLV III (l938). pp. 413-34, "Trends in th,.

THE FIELD OF THE SHIFTING EQUILIBRIUM

165

librium will find a sol ul ion which will bring th is demand into being under the cond itions outl in ed. The val ues established by th e solution will agree wit h the val ues of a set of general equilibrium equations of the \oValrasian type. The equilibrium achieved will be onc built upon a ba lance of opposing (orces and it will be a stable equi librium only as long as these opposing forces arc in perfect bala nce with each other. The ,Movemertt toward Full Equilibrium under Conditions of Full Employment. We have described the situation only after full equilibrium has been reached. We must take a brief look at the changes involved as such a system moves toward a point of rcst . If we go back in lhe system to where the Quantum o f capital is, say,OA as show n on Fig~ re I, and hold to a ll other assumptions made , t he demand cu rve DqD' will lie much lower down . In th e movement toward capital level ON (Figure I), with the interest stru ctu re perfectly mall eable, we may expect the price- and wagerate structu res to accommodate themselves to the changing si tuation in such a way that employment is always full. T he point q will travel the path shown by the supply curve SS' (by assumption supposed to remain in the same position) as the system moves over time a nd t he quantum of capital grows. For each Week of the period it takes t o increase capita l to ON, there will be certain va lues for t he structure of the shifting equi librium . The values of thi s system will determine how fast the system moves toward the ' Rigidity' of English Wage Rates," Rntiew of Economic; Sludies, VI (1938-9), pp. 189-99 ; N. Ka ldor, "Professor Pigou on Money Wages in Relation to Unemplo)'ment," Economic J ournol, XLVII (1937), pp. 745-53, "Stability and Full Employment," &onomi, Journal, XLVII I (1938), pp. 642-57, "Money Wage Cuts in Relation to Unemployment: A Reply to Mr. Somers," Review of &onomic Siudits, VI (I 938-9}, pp. 232-5; M. Kalecki, " The Determinants of Distribution of the Nat ional Income," EconornetriUJ , VI (1938), pp. 97-112; J. M. Keynes, "Professor Pigou on Money Wages in Relation to Unemployment " (Note). Economic Journal, XLVII (1937), pp. 743-5, " Relative MovemenlS of Real Wages and Output," Economic Journal, XLIX (lea9), pp. 34.-51; A. P. Lerner, "Ex-Ante Analrsis and Wage Theory," Economica , VI n.s. (1939), pp. 436-49. "The Relation of Wage Policies and Price Policies," American Economic; RmttlJ, XX IX (1939) , Supplement. pp. 158-69; A. C. Pigou, " Real and Money \Va~es in Relalion to Unemplorment," Econo mic Journal, XLVII (1937), pp. 0105 ·22, "Real and ~I oner Wages in Rela tion to Unemployment" (Note), Economic Jourl1al. XLViii (1938), pp. 134-8; J. H. Richardson, "Real Wage Movements," &onomic JOllrnal, X LI X (1939), pp. 425-4 1; H. M. Somen, "Money Wage Cu ts in Relation to Unem ployment," Rl!t!irnl of Ecollomic Studies, VI (1938-9), pp. 161-3; L Tarshis, "Changes in Real and Money Wages" (Note), Ecollomic JOllrllol, XLIX (1939), pp. 150-4; J. Tobin, " A Note on the Money Wage Problem," Q rlG rler~'Y Jollrnal of Ec;o/romics, LV (1940-1), pp. 508·16.

166

KEYNESIAN ECONOMICS -

stationary state. The influences of time-preference elements wi11 show th emselves in the position, arrangement, and dispersion of the Multiplier Functions. The influences of the laws of cost as the consumption and investment industries come into competition with each other will show themselves in the shape, arrangement, and dispersion of the Investmen t Functions. The requirements the system for quantities of money, measured in wage-units , will show themselves in the Liquidity Functions. The successive solu tions which eme rge from the functions of the shifting equil ibrium Week by Week will tend to make the io rate of interest agree with the current marginal prcxluct of capital, but the related r' rates will fall below that rate in accordance with the expected fa ll s in the marginal product over the life of the secuntles. But under the assumptions made, the whole complex of rates will drift downward an d there wi ll be a connected upward trend in rea l wages, relatively accelerated when the supply function of labour begins to show negative elasticity toward increases in rates. Progress toward the ideal quantity of capital for the system will, however, tend to be slower than it would'-be if the supply curve were still positively inclined. . If we postulate certain minimum frictions connected with the adjustment of capital to new forms as the quantity of capital to be allied with the labour supply increases, a nd connected a lso with moving labou r from job to job or retraining it for new forms of capital made available, the values of the system will depart somewhat from those described above. Employment may be always a little below the ideal; the rates of interest may always be a little higher than the ideal rates because of an artificial scarcity of capital due to lack of adjustment to its ideal forms. But the character of the relationship will be substantially the same and the system wi11 move toward a set of relationships of the character of those shown in Figure I. In terms of the shifting equilibrium. in such a system the Investment Functions will be falling as the channels of investment available under the existing state of technical knowledge an d supplies or physical factors a re progressively fi ll ed , up to the level justified by the equi librium margin of substitution, NN' (Figure I). In term s of the Multiplier Functions, increases in real income will be bringing a tendency for the marginal propensity to consume to fall. offset to somc degrce by the fall in the io rate of interest. In terms of Figure I, when the system has reached the place where the maintenance rate is equal to the current marginal product of

of

THE FIELD OF THE SHIFTING EQUILIBRIUM

161

capical, or io rate, it will come to the full pojnt of rest. Since no further fa ll is expected in the marginal product, a ll inte rest·rates will be equal to t he io rate a nd equal to NN' of Figure I. All of the multipl e ma rgin s of substitu tion throughout the syste m will be in equili brium .

w

s

D

w•.I--- - - ------">i-- - ----">;.'

w·~------------~~~

O L-----------~_+.-----+~~ N Nq • n, F IGURE

XXVI

(1) EMPLOYMENT IN THE STATIONARY STATE (SECOND TYPE)

The Conditions of Equilibrium in a System with Chronic Un· employment. We return in this section to consider the level of employment and othe r economic relations in a system in which the i o rate is too inflexible to permit the economic community to come to full stationary equilibrium at the leve l of full employment. We go back, t herefore , to th e cond itions of Fig ures III and V. We sha ll in what rollows use Figure V and its symbols for points or reference .an d we show in Figure XXVI the cha racter of the relationships which wi ll apply. The supply curve for labour of this figure is set .up under t he same conditions as the supr1y curve of Fig ure XXV. It is in connection with [he demand curve fo r labour that we must analyse t he di fferences in the sit uation .

,68

KEYN ESIAN ECONOMICS

As before, the demand curve Dg'r'D' illustrates the demand for labour under conditions .where changes in the rate of interest a re subsumed under the funct ional forms. Th e dcmana curve is shown lower do~n than the demand curve in Figure XXV upon th e hypothesis t hat with the lower quantity of capita l (OAf of Figure V instead of ON of Figure I) demand prices in real wages will be lower for given quantities of labour-units because of lower physical margi na l prod uctivity. What we have in the demand curve is essentially a series of partial equilibria, based upon certa in definite assumptions. If the real wage-rate and the amount of capita l a rc given , and if the system operates under conditions where it must reach temporary equilibrium under e.xpectations of stability each Monday, t hen the amount of employment to be associated with that quantity of capital and rea l wage·rate wi ll be a definite quan tity also. Any point on the demand cu rve shows the rate of rea l wages and the qua ntity of employment which wi ll be associated with each other under these assumptions. The complex of the rates of in terest to be associated with each wage·rate and each associated q uantity of employment is as before subsumed under the functional form. Th e essential difference between the case illustrated in F igure XXVI and that illustrated by Figure XXV is t hat in Figure XXVI we assume that the interest-rate actually does lose Aexibili ty when employment strikes level Nfl" associated with q' on the demand curve of Figure XXVI. This will mca,n in terms of Figure V that the io ra tc has reached the level Om . Since the stock of capital is assumed to have become stationary at OM, not on ly the current ma rgina l product is equa l to Om but a\1 the future marginal products are expected to equal Om. At interest·rate Om t he Mu ltiplier Functions show the' average propensity to consume to be unity and th e Investment Functions will show the investmE"nt level to be zero when the general level of employment is that associated wit h employment N Q• a nd the real wage-rate is OW",. ]n term s of the graphic systems shown in Cha pter XII, as employment ri ses above N Q• and the associated level of Y, the propensities to save increase faster than the inducements to in vest; as employment fhlls below N v • the propensities to save fall off faster than the inducements to invest. The Equilibri;,m Rale of Real Wages. WE" have yet to show t ha t this solu tion gives stability under the conditions of the Fundamental Model and that th e situ a tion requires that the real wage-rate be

TH E FIELD OF THE SHIFTING EQUILIBRIUM

169

OWq • if equilibrium is to be reached and held. This situation appears to be gue for t he fo ll ow ing reasons: (1) At any h igher leve l o f employment with ca pital stock OM and wage-rate OWl/" operation of the principle of diminishing return s will push the physica l marginal productivity of the existing stock of capital be low the rate Om while the increase in real income will increase the marginal propensity to save. For such a situation 5 would exceed I . For the reverse situation I would exceed S. By the definitions of Sand 1 such inequalities are impossible. (2) At any lower q uantum of capital stock fo r the same level of employment a nd rea l wage-rate, the physica l marginal prod uctivity of capital will be above Om and the ma rgi nal propensity to save wi ll be lowered by the fall in real incom e. I will exceed S . ]n th e reverse situation, S will exceed I. The situations are again impossible. (3) At any lower rea l wage-rate than OW q • wi'th capital OM and employment N q • there will be a degree of inequal ity in the distribution of income which wi ll lowe r the Multiplier Functions and destroy the equality between S and I in such a ma nner that S would exceed I. In t he reverse situation I would exceed S. The situations are again impossiblE.'. We have still to inquire whetht:r or not any combination of changes can bring another solution which may be peananen ti y maintained. This really means , ca n equil ibrium be es tablished with any lower wage-rate than OW q .? Lowe r wage-rates may be combined with (1) the same amount of capital and level of employment as a bove, (2) less employment and less capital, (3) more employment and morc capita l, (4) less employment and more capital, or (5) more employment and less capital. No. (1) is covered by case (3) above . No. (5) is the relevant case, for it is the case where entrepreneu rs under the spur of lower money wages attempt to substitute labour for capital in t heir operations. This is t he case most fa vou rable to the establi shment of anot her equilibrium for the system. If another level of equilibri um for the system does not appea r probab le here, t he probabil ity is th a t stable equ ilibrium can be ach ieved for the system only with the wage-rate OWq •• The effects of a subst it4tion of labour for capital under the spur of a lowe r real wage-rate achieved through a lower money wage-rate may be divided into an income effect and a distribution effect. The distriburion effect is clearly to rai se the ma rgina l propensity to save by throwing real in come to the higher in come classes and thu s inc'reasing inequa lity. The income effects arc uncertain.

KEYNESIAN ECONOMICS

170

The decrease in capital stock and increase in employment wh ich accompanies t he substitution may result in a total real income in goods and Services per Week which is (a) equal. (b) more, or (c) less than with stock O.M" and employment N q •• . If income is equal to, or morc than, income at stock OM and employment N q " distribution effects will tend to make 5> I and no solut ion can be reached under the cond itions of the Fundamental Model for the system of the shifting equil ibriu m. If dis-investment of capital is carri ed to the poin t where rea l income is lowered, it must be lowered enough to make the effects of the fall in income exactly absorb the effects of the greater inequality in distribution. Nevertheless, this situation will not give a new equilibrium. Capital is created a lso by labour. The dis-investment of capital will throw the ma"rginal returns on capital instruments over the rate Om, the lower real wage-rate in itself being a factor in the situation lowering costs of production for capital instruments as well as for direct output. Labour wi ll be usoo for the production of capital goods again until there is no difference at the margin bet,ween t he rewards for applying labour directly and indirectly. If, as real in come increases wit h t he increase in capital, some of the real income increases do not fall to wage-workers, distribution effects again will disrupt temporary equilibrium. I t does not appear there fore that any other wage-rate than OWq' will give a definitive equ ilibrium under the conditions given. If entrepreneurs do not realize this, the system is subject to error and may wander back and forth between i.nvestment and dis-investment, with the rea l wage-rate varying between the limits given. We havc, however, . exhau sted t hc capacitic!:!. of the Fundamental Model, which cannot afford us information respecting such a situation. We shall gain more information regarding the probable behaviour of employment, rea l wage-rates , interest-rates, and prices , if we leave this ::Vlodel and go over to the conditions 'of the First Supplementary i\lodel. (3)

E~1PLOYMENT.,

REAL WACES, AND-THE SHIFTING EOUIl.tDRIUM

The First Supplementary Jlodel atld the Nature oj E,rtreprenet" Errors. We shall still be dea ling with t he cond itions implied in Figure XXVI and shall follow throug h certain processes of change connected with it in a world of imperfect foresight built upon the plan of the First Suppleml'llta ry iVlodel. In the course of that

THE FIELD OF THE SHIFTING EOUILIBRIUM

171

inquiry we shall come to much closer terms with the nature of that Model than we have hitherto done. In addition we shall assume within the ModeJ the degree of separation between the retail and othe r markets and the time-lags in the movem en ts of prices wh ich are customarily present in the rea l world. Since we are working from t he conditions of Figure XXVI, we assum e that t he Quanti ty of capita l to sta rt with is OM and that the inte rest· rate is at its institutiona l minimum of Om. We a lso assume to si mplify exposition that 'the qu antity of money measured in money-units is a datum. These assum ptions mean that it entrepreneurs were to offer employm ent Nil' at real wage-rate OWq • with no expectations present of changes in prices, money wages, or interest·rates , the situation would be stable. S = I "" 0 as long as changes in tastes or technical knowledge, autonomous changes in the physical supplies of factors, changes in population factors . or changes in the aversion to work do not upset the conditions for eq uilibrium . Our first inquiry leads us to ask what types of error can occur in the Mode l and how these conditions may a ffect the equilibrium relation ship. Investigation will show that there will be four basic types o r error with two cases to be listed under each: I. Emp loyment is at the equilibrium level (No,) but the real wage·rate is (1) morc or (2) less than the equilibrium rate (OW,.). II. Th e real wage·rate is at the equi librium level (OWo') but employment is (1) more or (2) less than No,. III. Employment is greate r than No. and the real wage-rate is ( 1) more or (2) less than OW,.. IV. Employment is less than N,. and the real wage·rate is (l) more or (2) less than OWo" Ins pection of t he relationships above will show t hat in cases I (l). II (1), and III (1), there is an apparent condition where mar· ginal costs exceed marginal revenues if we reason that under the influence or the principle of diminishing returns , the marginal product o f labour will have a value equal to o' for each labour· unit expended when employment is No.,and for all levels of employment beyond that will have a lower margin al product. By sim il ar reasoni ng there is an appearance in cases I (2) a nd 11 (2) that mar· gi na l costs wi ll be less than marginal revenues and that in cases III (2) and IV (1) a nd (2) the effects cannot be generalized because there arc offseuing influences present.

aw

172

KEYNESIAN ECONOMICS

These appearances rest upon the possibility of hoards of liquid capital in the system which may be subject~ to di s- investment or furth er investment and upon the degree of sepa ration between t he wage-goods (retail) ma rket, a nd the tendency for pricing in the latter market to ex hibit Jags with the wholesale and other production markets. It will be shown by analysis that we shou ld fall into errors in reasoning if we should think that the appearance that marginal costs a rc exceeding marginal revenues which is present in type cases I (1), II (I) , and III (1) above means necessarily that there are inRuences present tending towa rd immediate contraction of output, as is t he ordin a ry reason ing with respect to the individual business enterprise showing such a n inequality. I n order to investigate the seq uences of error under the technica l '1lpparatus we have developed in the past twelve chapters, we shall cut in on the First Supplementary Model under an assumption t hat entrepreneu rs on Monday make an error of type I (1) in making contracts with the fa ctors of production . We shall not inquire as to the circumstances which lead to the error here. We must start so'm ewhere and we shall find as we go on that the course of errors a ppears to lead nowhere to a clear-cut establi shment of an equilibrium position even in the simple case laid down here. If with employment N q , wage-workers take t he money given them on Monday by ent repreneu rs into the retail market on Tuesday an d are a ble there to bu y ~va ge - goods (in Professor Pigou's phrase, meaning types of goods t hat wage-workers buy) on terms which make their ren.l wages exceed those t hat ent repreneurs thought t hey were ha nding to them on Monday, there wi ll be two possi bl e types of adjustment to t his situation. The first adjustment is that ent repreneurs in manufactu ring and wholesaling shou ld raise 'he prices of wage-goods furni shed to entrepreneurs in the retai l market until the dema nd fo r these goods comes into equilibrium with the ma rgina l costs of production, that is, until wage-workers ca n get in t he market a real wage equal to the current va lue of t he marginal producl. IT that could be done at one sudden co up under conditions where there were no further expec tat ions of price ri ses, the equilibrium levels of employ ment and wage· rates would be establi shed. Th e second type of adjustment would be that money ',mges would be ltroVered and prices left as they were. In view of the pric:!lIlindedncss of entrepreneurs and the money-m indedness of workers, we arc goin g to say that in t he prescnt case, the ad justment is made by pri ce ri ses rather than by cuts in money wages. Since the

THE FIELD OF THE SHiFTiNG EQUILIBRIUM

173

orig in a l error will have resulted in a dis-investm ent in liquid capital hoards. it will influence ent repreneurs optimistically. It will be interpreted as mea nin g that the demand for wage-goods is getting "stronge r." Since ind ividual en trepreneu rs really d o not know very much about the genera l sit uation, it is un likely that the price adjustments wi ll be made by a sudden coup. Rather where the errors are of the type I (1) , entrepreneurs in the rE!tail marke t will tend to ma ke gradua l price adjustments , experimenting to see how much demand ha;; in creased and how much of a price rise in the variou s lines will give t he relationship between price a nd quantity sold which wi ll be most ad va ntageousJrom the retailers' point of view. But a gradual adjustment is likely to affect t he psychologicalin stitutiona l complex in a world of imperfect foresight and in SO doing it wi ll affect the functional forms of the shifting equilibrium. Entrepreneu rs, awa re as individuals only of t hei r own immediate sit uations, may ext rapoiate t his price rise and the expectation may become a part of the sit uat ion determining the fun ctions of the shifting equ ilibrium in a way which will t hrow them farth er from the posit ion which t he situ ation may from a standpoint closer to the underlying economic realities be able to fulfil. The Liquidity Functions attached to defin ite levels of activity will fal l; the Investment Functions will rise. Some aven ues of investment now promise to pay more than the rate Om, part of the expected yields com ing from expectations of price ri ses. There will be an investment in these and a tendency fo r expansion in the consumption-goods indu stries under t he Mu ltiplier inRuence. Under t hese conditions we may expect employment to increase beyon d the equil ibrium level N q• and the system will be thrown from the errOr of type I (1) to that of type III. In the first instance t he enlarging of the wages-bill resulting from in vestment activities may be paralleled by dis-investment of hoards of liquid capital on a scale and at prices which leave real wage-rates st ill higher than t he eq uil ibrium level. Bu t as t he p~ice ad justment ca tches u p, the errors wi ll be t hose of type III (2). If the error is of type I I I (2), it is possible that the effects of the excess employment are counterba lanced by the lower real wage. Under these conditions entrepreneu r illComes take on the appearance of being maximized. But here action of the inte rest-rate will bring disequilibrium again. The rise of the Investment Functions and the fall of the Liquidity Functions had been based upon an expectation developed by t he period of rising pri ces t ha t certain

KEYNESIAN ECONOMICS

new investmen ts would in fact pay more than the institutional minimum ra te Om. Sooner or later the increase of these investmcntgoods will culminate in an increase in the quantity of consumption goods placed on the market. When that happens the price rises will decelerate and then level off with the slackening of investment activity, and finally there will be a fall in employm ent as it becomes evident that the new investments (or some portions of the old ones) are not likely to pay as high even as the rate Om. The fa ll in employment will tend to cause or to accelerate a price fall in the wage-goods market which wi ll tend to raise the rea) wage-rate towa rd t he equi librium rate . Both employmen t a nd the wage-rate will be moving toward the equilibrium levels therefore b.u t t here will be at.least th ree influences tending to carry the system past the equilibrium relationship of employment, wage-rate, and interest-rates: (1) By the period of positive investment the quantum of capital has been carried beyond OM, or the equ ilibrium amount for rate Om. (2) During the period of price falls ; entrepreneurs may g row to expect furth er price falls and t he L t fun ction may rise ca rrying the L fun ctions up, while the Investment Functions will fall. Employment will tend to be thrown below N, .. (3) If money wage-cuts accompany the price-cats (as is likely) another elemen t of instabili ty enters , namely entrepreneurs who would otherwise be planning the expenditure of maintena nce or replacemen t funds may find that it looks profitable to postpone expenditures un t il money wages have fall en furth er. The quantity of capital will fa ll not only back t~ OM bu t below it. Under t hese influences, t he general sit uation may pass over from error of type II I (2) to type IV (1). Thi s implies again that the real wage-rate is receivin g support from the liquidation of stores of liquid capital. That is to say, under these circum stances, entrepreneurs with stores of liquid capital may be atlempting t o sell them before prices fall farther or may be forced to sell them in "distress" sales. But as the liquidation of inventories comes to an cnd , real wage-rates will fall . An error of type 11 (2) may a ppear but there is no reason inherent in t he situation for the wagerate to stop at t he equilibrium level. The situation is likely to carryover into error of type I V (2). Here again effects on ent repreneur aClion a ppear un certai n. I£ the effects of lower real wage-rates exactly ba lance the effect s of a small er level of employment, entrepreneurs may feel that since

THE FIELD OF THE SHIniNG EQUILIBRIUM

115

current marginal costs appear equal to the value of the product in the market. the situation is stahle. Once again the interest-rate comes in to change this situation. Th e period of dis-investment will have brought the marginal product ivi ty of remaining instruments over the level where it agrees with the rate Om under condition s of stable (or rising) prices. As t he pri ce falls and wage fa lls level off , this condition becomes apparent. The Liquidity Functions will fall and the Investment Functions will rise. Re-investmcnt begins. If we cou ld jump straight to the equi librium levels, a ll might be :veil. But it is likely that the level of employmen t N q , may be reached with an error of the type I (2) if expansion of the consumption·goods industries lags behind expansion of the investment industries, waiting for the effects o£ prices transmitted through the retail market. Under such conditions the re will be elements of expansion in the system tending to ca rry it over into errors of type III (2) and then III (1) . as the expansion of consumption goods from the newly produced investment goods begi ns to flow into the market. But as the newly produced goods begin to flow into the market. the price ri se will slacken off, the new investments will prove again to be too optimistica lly und ertaken. and the system is ready for another downward spin. Many different sequences of error may occur and it is hard to see. once the sequence of error has started. any reason why the equ ilibrium levels should ever be establi shed. It is to be noted too that the cyclical movements have been support ed without an expansion of the supply of money, by transfers to and from the L, deposits. I t is perhaps advisable at t his point to consider the rela· tion of the quantity of money to the process ana lysed. The Elasticity of the Supply of ,Maue'y. the bllcrest·Rate St'tldu,e, atlll ImJest11lcnt Activity. If th e stock of money is a datum, in the ri sing phases of the cyclical process, the short rat es may be driven above Om by competition for the available stoc k. If there is an eXJ>Cctation that Om is a "norm