American Economic Association

Keynesian : The Search for First Principles Author(s): Alan Coddington Reviewed work(s): Source: Journal of Economic Literature, Vol. 14, No. 4 (Dec., 1976), pp. 1258-1273 Published by: American Economic Association Stable URL: http://www.jstor.org/stable/2722548 . Accessed: 19/03/2012 09:33

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to Journal of Economic Literature.

http://www.jstor.org Keynesian Economics: The Search for First Principles

ByALAN CODDINGTON Queen Mary College Universityof London

I would like to acknowledge various forms of indebtedness in con- nection with this paper: to, among others, Thanos Skouras, Alan Peacock, David Currie,and an anonymous refereeof thisjournalfor helpful commentson an earlier draft; to Michael Kennedy and Co- lette Bowe for indispensable guidance at various stages; and to the Universityof Manchesterfor the Hallsworth Fellowship in , which made this work possible.

I. Introduction

N The General Theoryof Employment, methodconsisted of analyzing markets on Interestand [13, 1936] and else- thebasis of the choices made byindividual where, Keynes attacked a body of theory traders.Thus, the resultingtheory oper- that he designated "Classical." In posing ates at two distinctlevels-that of in- a threat to the "Classical" system-or at dividual choice, and that of least to a recognizable caricature of it- phenomena-even thoughthe connection Keynes also called into question the between the two levels may be provided method of analysis by which this system onlyby an analysisof the choices of a "rep- was constructed.The purpose of this arti- resentative"trader. Moreover, in orderto cle is to inquire into the various ways in providea basis fora manageableanalysis which methods of economic analysishave of marketphenomena, the analysisof in- come to terms with this threat,either by dividualchoice has to be of a particularly responding to it or by reinforcingit with stereotyped and artificialkind. This furtherthreats; it seeks to ask the ques- methodof analysis,using markettheory tion: "What has to be changed or sacrificed based on choice theoryof a type thatal- in order to accommodate Keynesian ideas lowsthe twolevels to be connected,I will within standard methods of analysis?" Its referto as "reductionism,"on thegrounds theme will be the varietyof ways in which that the centralidea is the reductionof this may be done: three broad types will market phenomena to (stylized) in- be presented and the contrasts between dividualchoices. them explored. Considerationsof tractabilityimpose The firsttask accordingly is to charac- restrictionson the kind of choice theory terize the method of analysis of that body on whichthe market theory can be based: of theory in opposition to which Keynes the theorycannot deal withchoices in all presented his own. Very broadly, this the idiosyncraticdetail in which actors 1258 Coddingtonon Keynesianism 1259

conceive of them, nor in terms of the elu- gram, I will refer to as "fundamentalist sive and wayward manner in which actors Keynesians." It is the purpose of this sec- make up theirminds; stable objectives and tion to expound such an interpretation,to well-defined constraints are needed to consider some of the difficultiesin sustain- provide a firmenough foundationfor mar- ing it, and brieflyto discuss its significance ket theory.And just as the choice theory for economic theorizing. has to be restricted in the interests of Like the interpretationof the work of building up to market theory,so the mar- any active mind, the interpretation of ket theoryhas to be restrictedin the inter- Keynes's writingsrequires the use of se- ests of working back to choice theory. lection and emphasis: it requires a view as Overwhelmingly,reductionist theorizing to what is central and what merely periph- has confined its attention to situationsof eral, what is essential and what merely in- market equilibrium; for these situationsa cidental, in his writings;in thisway appar- choice theory basis is relatively straight- ent inconsistencies and obscurities may forward. There may be, in accordance readily be resolved, at least to the satisfac- with the standard schedules, a gap be- tion of those adhering to that interpreta- tween market demand and market sup- tion. For fundamentalists,what is central ply, but the choice theory from which and essential in Keynes's writingis to be each of these schedules is derived sup- found primarilyin his article "The Gen- poses that all choices are realizable; ac- eral Theory of Employment" [14,1937] in cordingly,the standard schedules can tell the Quarterly Journal of Economics of us nothing about what will happen when 1937, an article concisely restatingthe ar- the tradersattempt to do what, in the ag- gument of the General Theoryin response gregate, is impossible; nor are the to various critics; in the General Theory schedules likely to persist as the traders itself,the essence is said to lie in chapter become aware of the difficultyof doing 12, "The State of Long-Term Expecta- what they had regarded, in making their tions," and, to a lesser extent, in chapter (intended) choices, as straightforward. 17, "The Essential Properties of Interest Overwhelmingly, therefore, reductionist and Money." The kind of considerationsto theoryhas been concerned with the con- be found in these places can be traced nection between equilibrium states of back at least to the work of ten years ear- market phenomena and the choice logic lier in The End of Laissez-Faire [12, 1926] from which these states could be gener- and, with hindsight,still furtherback. ated. It should be noted, however, that An early statementof the fundamental- this concern with market equilibrium is ist position was provided by Hugh Town- not a definingcharacteristic of reduction- send [28, 1937]. He argued that the kind ism: it is rather a way in which reduction- of considerations raised by Keynes in his ist theorizinghas been rendered manage- theory of liquidity preference have quite able. devastating consequences forreductionist price theory if they are allowed to apply II. FundamentalistKeynesianism to all assets. Once all prices are seen as If Keynes's ideas are to be seen as a money prices, and all assets as bearing a threat to the reductionist program, the liquidity premium, price theory becomes question naturallyarises of how serious a enmeshed in the same tangle of expecta- threat they are: of how fundamental the tional and conventional elements that aspects are that are threatened. Those characterizes Keynes's theory of the rate who have seen Keynes's work as a frontal of interest. On this view, the hope of ex- assault on the whole reductionist pro- tracting"real" (relative) prices fromtheir 1260 Journalof EconomicLiterature

monetary context looks bleak; although the clearly specified and stable objectives we should not, as Hicks [9, 1946, chap. 12] and constraintsrequired by reductionist has pointed out, confuse Keynes's innova- theorizing, Keynes emphasises that the tion in analytical procedure (in dealing basis of choice lies in vague, uncertain, with the rate ofinterest in associationwith and shiftingexpectations of futureevents money-holdingdecisions rather than with and circumstances:expectations that have borrowing and lending) with his sub- no firmfoundation in circumstances, but stantive contributions.Nevertheless, the that take theircues fromthe beliefs of oth- threat to the reductionist program does, ers, and that will be sustained by hopes, on thisview, indeed appear to be a funda- undermined by fears and continuallybuf- mental one. feted by "the news." He was drawing at- Perhaps the most uncompromising,and tention to both the importance and the certainlythe most tirelesslyeloquent, ex- elusiveness of the state of business confi- ponent of fundamentalistKeynesianism is dence, and the way it unfolds. Keynes G. L. S. Shackle [25, 1967; 26, 1972; 27, focused on the conventional element in 1974]. His own work has centered on the valuation: the way in which valuations irreducibly creative element in human may persist to the extent that they are choice: its basis in constructsof the choos- shared, but are therebyrendered sustaina- ing mind. His appreciation of Keynes's ble in the face both of minor events and contributionsto economic theoryhas, ac- of changes in circumstances,but vulnera- cordingly,centered around thissame con- ble to anythingthat threatensthis conven- cern, and naturallyhe sees these matters tional basis. In the course of a riot,for ex- of expectation,uncertainty, and ignorance ample, the moods and feelings of the -matters of the provision of knowledge- rioters may be widely shared until, at a surrogatesin the face of knowledge defi- later stage when the riot has lost its force, ciency-as of the essence. A most succinct the moods and feeling may generally and distillationof Shackle's reading of Keynes rapidly revert to normal. The coordina- has recentlybeen provided by B. J.Loasby tion of such crowd behavior and its charac- [17, 1976]. teristicdynamics arise from the fact that One furtherline of thought must be the participants are taking their cues di- mentioned in the present context: this is rectly from one another. Reductionist one that has attempted to use the funda- choice theory as it has been developed mentalistaspect of Keynesianism as a way does not shed any light on decisions in- of clearing the ground to permit a return volving such immediate and stronginter- to a certain cluster of doctrines and con- dependence as this. cerns that are variously referred to as Once its choice-theoretic foundations "classical" (as distinctfrom "neoclassical") are threatened, the whole reductionist or "Neo-Ricardian." The objective of this program is called into question; for with- school, whose most distinguished practi- out them the market theory would have tioner is Joan Robinson, is to produce a nothing on which to stand, nothing to hybridof Keynesianismwith those aspects which it could be reduced. The concept of of Ricardo's work that were appropriated market equilibrium is in this way left ex- by Marx: Ricardo minus Say's law and the posed to attack. For without a clearly- quantity theory of money. specified and stable basis in choice logic, Keynes's QJE paper of 1937, to which the idea of market equilibrium is no fundamentalistsattach such great impor- longer connected to the realizabilityof in- tance, is, firstand foremost,an attack on dividuals' intentionsin the aggregate. This the kind of choice theory that is required does not mean that market equilibrium for the reductionist program. As against cannot be rehabilitated; what it means is Coddingtonon Keynesianism 1261 that the sustainabilityof equilibrium must fromthe restrictionsthat it imposes on our depend on conditions that are confinedto thinking.This, however, is not so much a the level of the market. For the funda- matter of what Keynes said, as of what we mentalist, however, Keynes's ideas re- are led to if we followhis line of thought, quire the rethinkingand reconstructionof taking the QJE article as the definitive the whole body of reductionisttheory: its guide to its direction. choice-theoreticbasis and the equilibrium Where we are led by a line of thought theory of markets that rests on it. depends a great deal, of course, on where The objections to equilibrium theoriz- we are disposed to go. Fundamentalists ing have been elaborated by fundamental- have, correspondingly,contributed freely ist Keynesians. Joan Robinson has shown of their own preoccupations in arrivingat that if the idea of equilibrium is pursued interpretations of Keynes's thought. At relentlessly,then as the concept becomes their most uninhibited, fundamentalist all-embracingit becomes paralyzed by its Keynesians have presented Keynes's ideas own logic: equilibrium becomes a state of as an escape fromthe essential "timeless- affairsthat is, strictly,unapproachable: un- ness" of the modes of thoughthe attacked. less it already exists, there is no way of More concretely,they have presented his attaining it [20, 1953]. Similarly,in the central message regardingemployment as work of G. L. S. Shackle, the idea of gen- concerning the existence of a liquid asset eral equilibrium is shown to require the in a world of uncertainty,thus providing pre-reconciliation,one with another,of all a retreat from the holding of real assets present and futurechoices of all economic and the associated commitment to (em- actors [26, 1972]. On either ground it ployment-generating)production of a par- would follow that the standard use of the ticular output. This theme has been much method of comparative statics (or, better, elaborated by Shackle and is concisely ex- "comparative equilibria") to analyze the pounded by Loasby; in Joan Robinson's effects of changes in circumstances, is work,however, we findits place taken by strictlyunwarranted and illegitimate.' Of a preoccupation with the heterogeneityof course, thisline of thoughtwould have ni- capital goods: the fact that individual hilisticconsequences forthe entire corpus items of the capital stock that historybe- of economic theory and in particular for queaths to us cannot be costlessly trans- its applicability;in this respect, the line of formed into one another, but exist in thought reaches a purist and impractical particular forms, embodying particular conclusion that is in marked contrast to techniques, reflectingthe superseded ex- Keynes's own highlyeclectic approach to pectations of the past. The problems economic theory. raised by the existence of liquid assets and The concept of equilibrium is accord- durable, functionallyspecific capital as- ingly seen by fundamentalistsnot as a sets, are not, however, unrelated; the na- useful simplificationFor economic theo- ture of capital goods means that holding rists,but as a distraction.2The essence of them involves a kind of commitment, Keynes's thoughtis seen as the liberation while the nature of liquidity allows an es- fromequilibrium theorizing,as an escape cape from that particular commitment. Fundamentalist Keynesianism, in see- ' This argumentis elaboratedin Coddington[3, ing Keynes's ideas as a wholesale on- 1975] and Loasby[17, 1976,chap. 3]. 2Thus: "The argumentstops when. . . the equi- slaught on the reductionistprogram, does libriumlullaby hushes further inquiry" [21, Robin- not see those ideas as providing a substi- son, 1964, p. 80]. But thissoporific effect is never tute for that program. Rather, it sees reconciledwith the concurrentlyheld view that "The conceptof equilibrium,of course,is an indis- Keynes's own ideas as a firststep in a thor- pensabletool of analysis"[21, 1964,p. 78]. ough-going revision of economic theory. 1262 Journalof EconomicLiterature

Accordingly, it sees what Keynes did spondingto fullemployment as nearlyas is constructivelyas merely a makeshift,an practicable,the classicaltheory comes intoits ownfrom this point improvisation, a stop-gap. To take onwards.If we supposethe the volumeof outputto be given,i.e. to be deter- constructivepart of Keynes's work (in de- minedby forces outside the classical scheme of veloping the consumption function,the thought,then there is no objectionto be raised marginal efficiencyof capital schedule, againstthe classicalanalysis of the mannerin etc.) as being the substance or result of whichprivate self-interest will determine what in particularis produced,in whatproportions "the Keynesian Revolution" would there- the factorsof productionwill be combinedto fore betoken a failure of nerve, a betrayal produceit, and howthe value of the final prod- of fundamentalistprinciples.3 uct willbe distributedbetween them. In order to sustain the fundamentalist interpretation,it is necessary to postulate This is abundantly clear, and in obvious that Keynes himselfhad occasional lapses. conflictwith the fundamentalistview of Thus, Joan Robinson [23, 1973, p. 3] Keynes's thoughtbeing subversive of the writes: whole classical ("reductionist") scheme. . . .there were momentswhen we had some Accordingly,we findJoan Robinson writ- troublein gettingMaynard to see what the ing [21, 1964, p. 92], in connection with pointof his revolution really was, but when he this passage, of the "fallacy" that Keynes came to sumit up afterthe book was published fell into, and remarking sadly that, "He he got it intofocus. was himselfpartly to blame forthe perver- Here she refers,of course, to the QJEarti- sion of his ideas" and "Keynes himselfbe- cle of 1937. gan the reconstruction of the orthodox Again, she writes [21, 1964, p. 75]: scheme that he had shattered" [22, 1971, The GeneralTheory broke through the unnatu- p. ix]. ral barrierand broughthistory and theory A further embarrassment for funda- togetheragain. But fortheorists the descent mentalists is that Keynes indicated quite intotime has not been easy.After twenty years clearly that he found nothing to object to the awakened Princess is still dazed and groggy. in Hicks's distillation[8, 1937] of the Gen- eral Theoryinto the IS/LM framework,or Keyneshimself was not quite steadyon his feet.. what has come to be known as "the in- come-expenditure model," quite devoid She then goes on to refer [21, 1964, p. 75] of any fundamentalist characteristics.4 to Keynes's ("highly suspicious") remark This again must be seen as some kind of about the timeless multiplier [13, 1936, p. momentary lapse on Keynes's part if the 122]. fundamentalistinterpretation is to be sus- A major embarrassment for fundamen- tained, at any rate if Keynes himselfis to talistsis to be found in the finalchapter of be allowed to be a fundamentalistKeynes- the General Theory.Here we findKeynes ian. arguing as follows [13, 1936, pp. 378-79]: What, then, does fundamentalism add to? It does not . . .if ourcentral controls succeed in establish- up provide any sort of de- ing an aggregate volume of output corre- terminate theory or model of how the economy functionsat the aggregate level; 3An immediatedifficulty for fundamentalists is the it does not enable one to make any definite factthat the QJEarticle of 1937, afterhaving ad- predictions about the likely effectsof al- vancedthe argumentsalready discussed, goes on to ternative policies or circumstances. On stressthe importanceof the consumptionfunction, whichis thendeployed (anticipating terminology I will introduceat a laterstage) in a thoroughlyhy- 4See Keynes'sletter of March31, 1937, to Hicks draulicfashion. in Hicks [10, 1973,pp. 9-10]. Coddingtonon Keynesianism 1263 the contrary,it is a viewing point from inhibitingeffects on alternative modes of which such constructionswould appear as economic regulation. More broadly, the rather desperate makeshiftsof transient comparison also arises among the alterna- applicability. Fundamentalist Keynesian- tive effectsof investment decisions taken ism is concerned with the texture rather within alternative institutional frame- than the direction,as it were, of the eco- works (various powers and responsibilities nomic process. having been given to agencies of the To stress the basis of all economic ac- State), whose regulative capacities then tivityin more or less uncertain expecta- also become a part of the appropriate tions is precisely to emphasize the open- comparison. ness and incompleteness of economic In summary,we can say that fundamen- theorizingand explanation. It does not it- talist Keynesians are united in seeing self provide any kind of fixedmechanism Keynesian ideas as posing a threat to the according to which the unfolding of whole reductionist program; and that events takes place; but it does show how their primary concern has been to rein- one would set about constructinga narra- force this threat with further threats. tive of events. It is a view about where the When it comes to providingan alternative gaps are in the causal chains that can be to the reductionist program, however, identified in the economy: the points at mattersare less unified.There is a marked which the economic process is susceptible contrast, for example, between the pro- to influence.We can accordinglybegin to spectus offeredby Joan Robinson for the appreciate the deep ambivalence of this completion of the Keynesian revolution standpoint towards . On and the insightoffered by Shackle into its the one hand, it sees potentialityfor enor- integrityand essence. And when we move mous leverage, the whole economic proc- from the critical to the constructive as- ess moving in response to changing states pects of fundamentalism, not only are of mind and consciousness; on the other mattersless unified,they are also less defi- hand, the very precariousness of this vi- nite. In Loasby's work, this indefiniteness sion leads very naturally to thorough- is transformedinto a methodological prin- going scepticism about the predictability ciple [17, 1976, p. 167]: of the effectsof deliberate attemptsto ap- If one can summarisein one sentencethe the- ply leverage in pursuit of political objec- oryof employmentset forthby Keynesin his tives. The point of view in itselfprovides [QJE]article of 1937,it is this:unemployment no guidance on whether the precarious- in a marketeconomy is theresult of ignorance ness is so pervasive as to undermine the too greatto be borne.The fully-specifiedmac- potential for political leverage. That is to roeconomicmodels miss the point-which is preciselythat no modelof this situation can be say: the wayward and unrulycharacter of fullyspecified. individual choices-and in particular in- vestment decisions-is seen as an impedi- III. HydraulicKeynesianism ment to economic functioning;but the During the nineteen fortiesand fifties, question that must be faced froma policy there appeared a number of expositionsof point of view is whether it is a greater "Keynesian economics," attempting to impediment to the self-regulationof the make the ideas accessible to students,and economy than it is to the workingsof dis- even to intelligent laymen. What these cretionaryfiscal and monetarypolicy. This works had in common, quite apart from matter would involve not just the consid- mattersof substance,was an unmistakable eration of an impediment to economic enthusiasm for (what were taken to be) functioning,but a comparison between its Keynes's ideas. This enthusiasm was at 1264 Journalof EconomicLiterature timesunrestrained to the pointof excite- aggregatelevel.7 Indeed, the esteem in ment; it was the authorsof these works whichthe twoaspects have been held has who spoke without reservation of a tendedto movein oppositedirections, the "KeynesianRevolution," one ofthe books periodwhen "fiscalist"policy enthusiasm in facthaving this title [15, Lawrence R. was at itsheight being a timeat whichthe Klein, (1949) 1966]. It is some indication intellectualinterest in theunderlying the- of the level of enthusiasmreached by oryhad become moribund.Again, the de- these expositorsand popularizersthat mise of "fiscalism"in the late sixtiesand one,Jan Pen, wrote a booksetting out and earlyseventies was accompaniedby a rea- discussinga particularspecification of a wakeningof interestin the underlying static"Keynesian" model of relationships theoreticalconceptions. (We shall have between a small number of macroeco- more to say about thisrevival in the next nomic aggregatesand gave it the title section.) ModernEconomics [19, 1965].It is notmy Allthis should not be allowedto givethe intentionhere, however, to attemptto impression,which would be quite mis- chart the process of the diffusionand taken,that the fiscal enthusiasm stemming popularizationof Keynes'sideas.5 fromKeynes's ideas did not include, or The period of Keynesian enthusiasm could not provide,a theoryin supportof was reallythe post-warperiod: the ideas its policy doctrine.It could and it did. went canteringbriskly through the fifties What,then, we are led to ask,is the theo- and earlysixties; faltered sometime in the reticalbasis forfiscalist enthusiasm? How middlesixties and stumbledinto the sev- is it to be characterizedas one of the enties.6This, at any rate,is the pictureas strandsin the developmentof Keynesian its emergesat the level of popularinflu- thought?It is to these questionswe now ence,at thelevel of widely and influential- turn. ly-heldviews on macroeconomicpolicy; at The theoreticalcontent of the body of thelevel, that is, of Keynesianism as a doc- ideas that has been propagatedthrough trine about how a largelydecentralized the educationalsystem in the West since economymay be subjectto broad (as op- WorldWar II as "KeynesianEconomics" posed to detailed)central control or influ- (by,for example, 'speda- ence through the instrumentof the gogically authoritative textbook [24, budget.It is temptingto adopt the prac- 1973]) I shall proceed to referto as "hy- tice ofreferring to thisdoctrine as "fiscal- draulic Keynesianism."This designation ism" to showthat it is a particularvariant reflectsthe view that the natural and obvi- (and perhaps a corruptionor vulgariza- ous way to regard elementarytextbook tion) of Keynes'sideas. At any rate,it is Keynesianismis as conceivingof the econ- importantto keep distinctthe ups and omy at the aggregatelevel in termsof downs of Keynesianismas a policydoc- disembodiedand homogeneousflows. Of trine fromthose of Keynesianismas an course,conceiving of the macro-economy academicallyrespectable theoryof the functioningof a capitalisteconomy at the 7 Reflectingon the fragmentationof Keynesian thought,Axel Leijonhufvud makes the following ob- 5 But see JohnKenneth Galbraith's "How Keynes servation:"For some time now, contentmentwith came to America"for some interestinginsights into thisstate of the artshas restedon the motto'The the way Keynesianideas made theirentry into the TheoreticallyTrivial is thePractically Important and U.S. academiceconomics establishment [5, 1971]. the PracticallyImportant is the Theoretically 6 For an attemptat intellectualstock-taking at that Trivial.'It is a disturbingformula which can hardly time, see my "RethinkingEconomic Policy" [4, be a permanentbasis forthe furtherdevelopment 1974]. of the field"[16, 1968]. Coddingtonon Keynesianism 1265 in thisway willbe fruitfulonly to the ex- is onlyone agencymaking deliberate acts tent that there exist stable relationships ofchoice; that one agencyis "the govern- betweenthese overall flows. And it is my ment."And it is the beliefthat there are contentionthat the centralcharacteristic indeed stablerelations among the various of "hydraulicKeynesianism" is the belief overallflows in theeconomy that provides that such stable relationshipsdo existat a basisfor "the government"to pursueits the aggregatelevel. It is thisbelief that policygoals regardingthe overalllevel of givessome point to the hydraulicconcep- economicactivity and hence,relatedly, of tion;without such a beliefthe conception the level ofemployment. It is the stability would simplybe a matterof nationalin- of these aggregaterelationships that pro- come accounting,not of economic theory. vides"the government" with the leverage It should be noted that the flowsin- it needs to influencethose flows that are volved in thisconception are flowsof ex- not under its directcontrol. By making penditure,income or output.That is to deliberate choices for the flowsit does say, neither prices nor quantities per control(via the budget),and bearing in periodmake a separateappearance: they mind the (allegedly)stable relationships appear inextricablyin the contribution betweenthis and the otherflows that are each makesto the overallflows of spend- objects of concern for economic policy, ing and receipts.It shouldnow be appar- "the government"can, in principle,exer- ent whythe belief in the existenceof, and cise an indirectcontrol on theoverall level the attemptto establish,stable relation- (althoughnot the composition) of the flows shipsbetween the overall flows is radically thatare not the objectsof anyone'sdelib- inconsistentwith reductionism.For any eratechoice. That is thestory. On theface reductionistprogram must give a crucial ofit, it mayappear a majortriumph in the role in itstheorizing to pricesas such(not march of human reason:a dramaticand to the contributionthey make to overall irreversibleextension of the boundaries of spendingflows). The groundsfor this view politicalresponsibility. Instead of unem- are thatit is pricesas suchthat provide the ploymentand depressionbeing seen and incentivesthat individuals face in making accepted passively,like the weather,they the choiceson whichthe wholescheme is are to be seen as mattersfor human will to rest.This does not mean thathydraulic and design, something that human Keynesianismcan allow no part at all to agency,through the instrument of central be played by prices; when we come to government,could actually resist and thinkof such prices as embracingwage remedy.8As an idea it lookedboth simple rates and interestrates, we can see that and good; accordingly,it was, at the end thiscannot be so. Correspondingly,it does ofthe war,rapidly assimilated to boththe not mean thatreductionism is incapable policystatements and rhetoricof all major of allowingoverall flows to play any part politicalparties.9 in its scheme.Since these are alternative 8 This changed attitudedid not come easily or programsfor theorizing, rather than alter- quickly,and fundamentalattitudes had been under- nativetheories, they revolve around mat- goinga processof erosionfor some decades by the ters of emphasis.They do not concern time Keynescame on the scene. For a painstaking documentationof thisprocess in Britain,see Jose whatcan or cannotplay a partin a theory, Harris[7, 1972]. butwhat can or cannotplay a centralpart. 9 The majorbridge in Britainbetween Keynesian In fact,contrary to the standpointas- doctrinesand politicalplatforms was WilliamHenry Beveridge[1, 1944]. The ideas were given official sociated with reductionism,hydraulic recognitionin the WhitePaper EmploymentPolicy Keynesianismis a schemein whichthere [6, 1944]. 1266 Journalof EconomicLiterature

In summary,it can be seen thatthe hy- ever reason,in itsattempts to respondto draulicapproach is in conflictwith reduc- changesin expenditure. tionistmarket theory. The hydraulicap- Since the IS-LM apparatuswas put for- proach shows how things would work ward by Hicks, however,we have had when marketprices (and wages) willnot, somethinglike 30 yearsexperience of de- or will not quicklyenough, or willnot be mand managementpolicies based on the allowed to, performtheir allocative role; assumptionthat the economy exhibits it analyzesa situationin whichprices are marked hydrauliccharacteristics in the failing,both as disseminatorsof informa- short-run;and the questionof why these tion about relativescarcities and in the policies have been less effectiveat some provisionof incentivesto act on the basis times than others naturallyraises in a of thatinformation. practicalway the questionof the scope of If the centralmessage of the General thehydraulic conception. It is thereforeof Theoryis thatoverall employment is more considerableinterest that Hicks, in a revi- a matterof the demand for output than of sion of Keynesianeconomics in the light real wages, except when "full employ- ofthis experience, does notadopt his own ment"already obtains, then that message IS-LM apparatus for the purpose [11, is certainlyembodied in the hydraulicap- 1974, chap. 1]. Rather,he providesan al- proach.As such,it is an audacioussimplifi- ternativeframework in whichthe possibil- cation,which is, on the face of it,in con- ityof an expansionin demandbeing trans- flict with the corpus of reductionist lated intoan expansionof output depends theorizing.Furthermore, as a way of cruciallyon thestructure of inventories at thinkingabout macroeconomicpolicy, it the outsetof the process.In particular,it seemsto workto some extent,sometimes. dependscrucially on therebeing plentiful The intellectualproblem that it raises, stocksof materialsto sustaininvestment however,is thatof its own scope.What we projectsuntil decisions to increasethe out- need to know are the circumstancesin put of thesematerials are taken;or what which,and theextent to which,the opera- amountsto the same thing,it depends on tionof an economymay be conceivedof there being ample foreignexchange re- in hydraulicterms. There are variousap- serves representingcommand over for- proachesto thisquestion. A familiarone eign inventoriesof materials.Unless this is provided by the IS-LM apparatus, conditionis met,the attempted expansion withinwhich it can be readilyshown that can easilyrun up againstbottlenecks and the economyexhibits the characteristics dissipate itself in the diversion of re- of the hydraulicmodel to the extentthat sourcesfrom other uses and, notoriously, theinterest elasticity of expenditure is low in creatingbalance of payments problems. and ofthe demand for money is high;with Of course,if the increase in investment a zero interestelasticity of expenditure expenditureis translatedinto a net in- and an indefinitelylarge interest elasticity crease in real investment,the multiplier ofdemand for money, the operation of the processcan set in,and adequate stocksof economywould be exactlyin accordance consumergoods will thenbe requiredto withthe hydraulicmodel: changesin ex- avoid bottlenecksat thisstage and sustain penditureflows would lead to changesin the exDansion.'I outputflows without any repercussions on 10This analysis can readilybe transformedfrom a the rateof interest. In sum,it followsthat "fixprice"basis to a "flexprice"one; in whichcase the the economymay exhibit the characteris- preconditionfor a successfulexpansion becomes that theprices of the various goods needed to sustainthe tics of the hydraulicmodel to the extent expansionwhile changesin expenditureare being thatthe interest rate is impeded,for what- translatedinto changesin outputare significantly Coddingtonon Keynesianism 1267

There are, of course,other approaches ity of the concept of equilibrium:the to the questionof the scope of hydraulic Clower-Leijonhufvudposition being that theorizing.Indeed, the Monetaristargu- the concept of equilibriumshould be mentsagainst Keynesian conclusions may abandonedin theinterests of a morethor- be seen as one possible answer to this ough-goingreduction of Keynesianideas question:namely, that the scope of hy- to choice logic. The thesis is that once draulictheorizing is practicallynonexist- equilibriumhas been abandonedand one ent. In these argumentsthe Keynesian focuseson a processof tradingat disequi- conclusionsare underminedby the rein- libriumprices, then one has a framework troductionof a choice-theoreticbasis of that is entirelycongenial to Keynesian thestandard reductionist type. As we shall ideas, unlike the frameworkof equilib- see in thenext section, the work of Clower rium theorizingwhich, on this view, and Leijohhufvudmay also be seen as con- leaves roomfor them in onlythe mostat- tributingto this question of scope, al- tenuatedand ad hoc form.The problem thoughthis is nothow eitherof them pre- thenbecomes one ofproviding a moreso- sentedhis work. phisticatedspecification of the constraints on individualchoices, opening up the pos- IV. Reconstituted Reductionism sibilitiesfor theoretically novel and chal- During the 1960's there emerged a lengingforms of market interdependence school of thought,associated primarily arisingfrom a schematizationof the proc- withthe names of RobertW. Clower [2, ess of disequilibriumtilading. 1969] and Axel Leijonhufvud[16, 1968], In orderto lead up to my characteriza- concerned with reappraisingKeynes's tionof the workof Clower and Leijonhuf- contributionto economics.These writers vud, it is appropriateto begin by discuss- presentedtheir work as concernedwith ing each writer'sown characterizationof reestablishingand reassertingthe discon- his work:how each of themconceived of tinuitybetween Keynesian economics and the task he had set himself.I will argue its alternatives,a discontinuitythat they that their own characterizationsare in saw as havingbeen blurredand finallylost variousrespects unsatisfactory, and that to view by the variousactivities of inter- myalternative is nottherefore gratuitous. pretation,condensation, and reconstruc- I shallnot, however, attempt to substanti- tionthat came in the wake ofthe General ate the designationof the work of these Theory.It is withinthis perspectiveac- writersas reductionist.I shall take it that cordinglythat the contributionof Clower once the idea of what is involvedin the and Leijonhufvudto ourunderstanding of reductionistprogram is appreciated, it Keyneshas been discussedand appraised. shouldbe clear thatthis work falls within Mypurpose here, however, will be to pre- the program. sentthe dispute between Clower and Lei- Let us take Clower first.Having ad- jonhufvud,on the one hand, and those vanced the "dual decisionhypothesis" as whoseviews they were combating,on the a basis forexpecting consumer spending other,as a familyquarrel withinthe re- to depend on currentincome, Clower ductionistprogram. Most fundamentally, goes on to speak unguardedlyof Keynes thefamily quarrel is aboutthe expendabil- having had this theory of household behavior"at theback of his mind when he belownormal, so thatas theexpansion proceeds trad- wrote the General Theory"[2, 1969, p. erswill release their stocks onto the market as prices 290]. Clower goes on immediatelyto ad- rise.If this condition is notsatisfied, the expansionary impetuswill be whollydissipated in priceincreases mitthat "I can findno directevidence in [11, Hicks,1974, pp. 23-30]. any of his writingsto show that he ever 1268 Journalof EconomicLiterature

thoughtexplicitly in these terms."After new perspectiveis to be apportionedbe- advancingwhat he takes to be "indirect tween Keynesand Leijonhufvud.Keynes evidence" for this, he concludes that maywell have providedthe inspiration for "Keynes either had a dual-decisionhy- the task,but ifthe productof the distilla- pothesisat the back of his mind,or most tionis to be presentedas a (purified)"Eco- of the GeneralTheory is theoreticalnon- nomicsof Keynes"to be contrastedwith sense."The picturehere seems to be one the (corrupted)"Keynesian Economics," of Keyneswith a mindfull of ideas, some then we are back in the realmsof mind- ofwhich he gotonto the pages ofthe Gen- reading,especially as this"Economics of eral Theory,the task being to work out Keynes"can be readinto the GeneralThe- whatthe remainder must have been. This oryonly with what seems to me to be a is a problemof readingnot so much be- greatdeal ofingenuity and determination. tweenthe lines as offthe edge ofthe page. So althoughLeijonhufvud at firstseems to In his conclusion,however, Clower main- be concernedwith the rather modest task tains,rather more soberly,that his pur- of findinga freshperspective from which pose has been "simplyto clarifythe for- thedevelopment of Keynesian Economics mal basisof the Keynesian revolution and can be surveyedor appraised,it turnsout its relationto orthodoxthought" (empha- thathe is in searchof theone perspective sis added) [2, 1969, p. 295]. This then fromwhich the Keynesianness of these de- leaves the taskquite up in the air,for it is velopmentscan be judged. Whatlooks at not explainedto the reader how thisre- firstlike a searchfor new anglesturns out lates to the previousconcern with what to be a searchfor authenticity. Keyneshad "at the back of his mind." Butit is notjust a matterof authenticity, Turningto Leijonhufvud,we findthat forthe fundamental presumption that un- he is at somepains to tryto makeclear the derliesthe work of Clower and Leijonhuf- taskhe has set himself.First, he makesit vud is thatKeynes said somethingimpor- plain thatthe doctrine-historicalquestion tant,not only for economic policy, but for of "what Keynesreally said" is a strictly economictheory. They are saying:"Let us secondarymatter for his purposes [16, read the General Theoryin a search for 1968, p. 9]. "The primarypurpose," he theoreticalinnovation." In other words, explains, "remains . . . to provide a fresh far frombeing engaged in disinterested perspectivefrom which the income- exegesis(as the concernfor authenticity expendituretheory may be reconsidered" mightsuggest), they were concernedwith [16, 1968,pp. 9-10]. (The "income-expen- reworkingwith a viewto rejuvenating(by dituretheory" is Leijonhufvud'slabel for which standardsthey must be judged to the "conceptual frameworkwhich has have had some success). crystallizedout of the debate triggeredby How, then,is the taskthat Clower and the GeneralTheory" [16,1968, p. 6].) This Leijonhufvudset themselvesto be ex- seems straightforwardenough. The dif- pressedand understood?The view I want ficultyarises because whatwas presented to advanceis thatthey were settingthem- was not just "Leijonhufvud'sfresh per- selves the task of constructinga frame- spective,"but rather the fresh perspective work that would provideroom or scope that Leijonhufvudclaimed to have dis- for Keynesian ideas. This quite rightly tilledfrom the General Theoryitself. On takesit forgranted that we alreadyhave the faceof it, the taskappears to be to get a good roughidea what Keynesianideas a perspectiveon the whole debate by go- are: ofwhat the GeneralTheory was driv- ing back to the originsof it. But the ques- ing at. Whatwas wantedwas a theoretical tionarises of how the responsibility for this niche in which what were taken to be Coddingtonon Keynesianism 1269

Keynes's insightscould take root and somethingthat Clower and Leijonhufvud thrive.The motivefor this search was evi- simplytake forgranted. dentlythe recognition that the framework The whole questionof whetherKeyne- of general equilibriumtheory that had sian ideas should be accommodatedby been widelyadopted forattempts at pre- abandoningequilibrium theorizing rather cise expressionof Keynesianideas leaves naturallyraises the questionof what use practicallyno room or scope for them: Keynes himselfmade of the-concept of they may appear in only the most at- equilibrium." It is certainlytrue that tenuatedand ad hoc form. Keynes made use of the term "equilib- On its own terms,then, the essence of rium." But before we conclude that if the Clower-Leijonhufvudposition is this: Keynes could expresshis ideas in these thatin orderto accommodateKeynesian termsthen theymust be perfectlycom- ideas, we have to abandon equilibrium patible with equilibriumtheorizing, we theorizingand addressourselves to an un- must pause to considerthe meaning of derstandingof the process of disequilib- equilibriumand theuses to whichan equi- riumtrading. In my terms,however, it is libriumconcept mightbe put. We must not just equilibriumtheorizing that has bear in mindthat it is entirelyin keeping been shownto be uncongenialto Keyne- with Keynes'seclecticism that his use of sianideas, but ratherequilibrium theoriz- the termequilibrium could have been a ing withinthe reductionistprogram. And rather desperate improvisationat one one can see whythis should be so without stage in the "long struggleof escape." even takingany detailed view about the An equilibriumis a configurationwhich, workingsof the economy.For withinre- once attained,will be maintainedpro- ductionismeverything boils down to acts vided the underlyingcircumstances (for- ofchoice within a well-specifiedsystem of mally,the parameters and exogenousvari- objectives,constraints and formsof inter- ables) remain unchanged. Accordingly, dependence; and in equilibriumtheoriz- the interestand usefulnessof an equilib- ing we confineour attentionto situations riumconstruction, as an end in itself,de- in whichall the independentlyarrived at pends on a questionwhich is, in principle, choicescan be simultaneouslyrealised. It an empirical one, namely: what is the thenfollows rather naturally, irrespective rangeof variabilityof the underlyingcir- of any detailsof marketforms or institu- cumstancesover the orderof magnitude tional arrangements,that such a system of the time involvedin adjusting(near leaves no roomfor the "unintended"and enough) to its equilibrium configura- "involuntary":for malfunctioning and dis- tion?'2That is to say,if the underlying cir- order. It follows, however, from my cumstancesare fairlystable relative to the characterizationof such theorizingthat speed of adjustmentof the endogenous thereare two distinctpossibilities for the variables,the equilibrium configuration of accommodationof Keynesian ideas: (i) the the systembecomes a matterof some in- abandonmentof equilibriumand (ii) the terestin itselfand mayprovide a reasona- abandonmentof reductionism.Clower bly useful substitutefor becoming in- and Leijonhufvud consider only the volved in the complexitiesof the ad- formerpossibility. We can see, however, justmentprocess. It is somethingto know thatthe claim that equilibrium theorizing II For a detailedexegesis of thispoint, see Don mustbe abandonedin orderto accommo- Patinkin[18, 1976,pp. 113-19]. date Keynesianideas postulatesthat theo- 12 We are here avoidingthe large question of whetherthe systemmay approachan equilibrium rizingmust be carriedout in accordance configurationwithout shifting the equilibriumthat withthe reductionistprogram; but thisis is beingapproached. 1270 Journalof EconomicLiterature wherewe are heading,provided we have to theproblem of attaining it.13 It asksthe some groundsfor believing that we will question how a decentralized market get mostof the way therebefore we start economy might,with some degree of headingsomewhere else. effectiveness,perform the task that the It is in the lightof theseconsiderations Walrasian auctioneer would perform thatwe can saywhy Keynes's use ofequi- smoothly.To ask thisquestion, one needs libriumconstructions was a peculiarone: a constructionin whichprices adjust less He was concernedwith discussing, among thaninstantaneously to economiccircum- otherthings, the instability of the underly- stances,so that at any point in time the ing circumstancesof his construction. pricesmay be effectivelyproviding incen- That is, one of his focusesof interestwas tivesto act, but the informationthey re- preciselythe failureof his equilibrium flectwill not be appropriatefor the equi- constructionto satisfythe conditionsfor libriumthat is being approached. the routineusefulness of an equilibrium Now itmay well be thatformulating this construction.Therefore, in arrivingat an questionraises some of the most profound appreciationof Keynes'smethod, it is not questionsin macro-and monetaryeco- enoughto ask the natureof his construc- nomics;but we are stillin need, forthe tion;we mustenquire also into its mode practicaldeployment of Keynesianideas, ofanimation. Whenwe have reasonto ex- of a usable simplificationsuch as the hy- pect relativelystable underlyingcircum- draulicapproach provides. And the use of stances,the construction may be animated such a simplificationwill require an accordingto the methodof comparative awareness of the circumstancesunder statics.When the animationis endemic, whichit maybe expectedto worktolera- when one is concerned,as it were, with blywell: an awarenessof its scope.This is the restlessnessof the underlyingcircum- where a reconstitutedreductionism may stances, the use of the construction play a part. For in order to examine the becomes less straightforward,and cer- scope ofa theoryin whichprices fail alto- tainlyless mechanical.Whether, in this getherto play their(ideal) allocativerole, case, there is anythingmuch leftof the one needs a theoryin which there is a concept of equilibriumis a matterof no partial failurein thisrespect. This latter particularimportance. What is important theorycould thenbe used to interpretthe is to see that,just as one does not expect practicalsuccesses and failuresof the hy- to quell a riotby takinga photographof draulicapproach: as a wayof trying to dis- it, neitherdid Keynes'smakeshift use of tinguishthe circumstancesconducive to the equilibriumconcept involve the ex- its being an adequate simplification.Ac- pectationthat he could freeze the econ- cordingly,we should see Leijonhufvud's omyin a particularstate. Shackle has ex- book as not so muchabout the economics pressed this idea with characteristic of Keynes as about the scope of the eco- elegance [25, 1967, p. 182]: nomicsof Keynes.Clower and Leijonhuf- vud claimto have shownthat Keynes was At each curtainrise the GeneralTheory shows us, not the dramaticmoment of inevitableac- tryingto adapt thereductionist method to tion,but a tableauof posed figures.It is only theexpression of his own ideas byrefocus- afterthe curtainhas descendedagain thatwe ing it on situationsof marketdisequilib- hear the clatterof violentscene-shifting. We have seen thatClower and Leijon- 13 In orderto do this,Clower and Leijonhufvud avoidJoan Robinson's ultra-strict logic of equilibrium hufvud'sversion of Keynesianismis a accordingto whichthe equilibriumstate is unap- reconstitutedreductionism: it addresses proachableand hence the problemof attainingit itselfnot to the state of equilibrium,but insoluble. Coddingtonon Keynesianism 1271 rium. But in displayingthe analytical istprogram: the fundamentalistapproach unmanageabilityof such a program,they by its rejectionof the choice theorythat make it clear that,insofar as Keyneswas is essentialto and the (equilibrium)mar- able to come to any definiteconclusions ket theorythat is typicalof reductionist abouteconomic functioning, he musthave theorizing;the hydraulicapproach by its short-circuitedsuch problems. short-circuitingof reductionistmarket Withinthe hydraulic approach, employ- theoryand its eschewal of formalchoice mentproblems are quite distinctfrom al- theoryfoundations; and the reconstituted locationproblems; they arise at theaggre- reductionistapproach by its attemptto gate level, and they are independentof makeroom for Keynesian ideas withinthe relativeprices and the compositionof de- reductionistprogram by refocusingthe mand or output.The thrustof the recon- market theoryon disequilibriumstates stitutedreductionist approach, however, whilstretaining the standardchoice-theo- is to present unemploymentas a by- reticfoundations. product or even a species of allocation It remainsonly to make some comments problem.But if thisformulation does not on the relationshipof the approachesto set any definitelimits on the scope of the one another;the thrust of these comments hydraulicsimplification, all it can suggest will be that the variousapproaches are, is a generalscepticism regarding the ap- in their contributionto understanding, propriatenessof aggregatetools to deal largelycomplementary. withproblems that are seen as involving The fundamentalistapproach provides the internalcomposition of those aggre- a verygeneral critique of the methodsof gates;this, however, adds nothingto what reductionismwith regard to bothits style we alreadyknow, namely that the hydrau- of choice theoryand the equilibriumthe- lic approach is a simplificationand ab- ory of marketswith which it is typically stracts from allocation problems. The associated.As suchit clearsthe ground for questionthat still remains is essentiallya theintroduction of Keynesian ideas; at the questionof decomposability. It is theques- same time it formsa kind of backdrop tion of the separabilityof employment against which hydraulic thinkingcan problemsfrom the allocationproblems on thrive,and, as it turnsout, reductionism Whichthey are, in practice,superimposed. can reappearin a modifiedform. Hydrau- To whatextent may we disregardthe al- lic thinkingcan thrivebecause, in the ab- locative structureof macroeconomicag- sence ofstandard reductionist results, one gregates?Just how bluntan instrumentis needs some drasticsimplification in order demand management? If the recon- to say anythingat all definiteregarding stitutedreductionist approach could be forecastingor policy.(The alternativecan- made tractablewithout collapsing into the didate is the drasticsimplification pro- Monetaristsimplification, it could be ex- vided by the quantitytheory of money pectedto shedsome light on thesematters and its modern variants.)Reductionism (as indeedthe Monetarist simplification it- can thenreappear because, in makinguse selfhas done). ofa drasticsimplification, one is led to ask questionsabout its scope and limits;these V. Conclusion questionswill concernwhy the economy In thispaper we have consideredthree may not work in the way that standard varietiesof Keynesianism:the fundamen- reductionisttheory indicates and are talist,the hydraulic, and thereconstituted questionsthat could be formulatedin a reductionistapproaches. Each one has modified and expanded reductionist been locatedin relationto the reduction- framework. 1272 Journalof EconomicLiterature

Thus, the fundamentalistapproach 6. GREAT BRITAIN PARLIAMENT. Em- clearsthe ground for Keynesian ideas, the ploymentPolicy. Cmd. [6527], Lon- hydraulicapproach provides the danger- don: H.M.S.O., May 1944. ous simplificationthat makes themat all 7. HARRIS, JOSE. Unemploymentand definiteand manageable,and a loosened politics: A study in English social reductionismprovides the reservations policy, 1886-1914. London: Oxford, and qualificationsthat provide guidance ClarendonPress, 1972. on the scope of this simplification.The 8. HICKS, JOHNR. "Mr. Keynes and the mattermay be expressedcryptically in 'Classics': A Suggested Interpreta- termsof Keynes's "long struggleof es- tion,"Econometrica, April 1937, 5, pp. cape." We may say thatwhat he escaped 147-59. from was (unreconstituted)reductionism; 9. . Valueand capital.Second edi- whathe escaped to was the hydraulicap- tion. London: Oxford, Clarendon proach;and whathe went throughin the Press,[1939] 1946. processof strugglehas been preservedin 10. . "Recollections and Docu- thefundamentalist approach. For a gener- ments,"Economica, Feb. 1973, 40(1), ationbrought up on Keynesianideas, how- pp. 2-11. ever,a senseof intellectual liberation is far 11. . The crisis in Keynesian eco- morelikely in the struggleof escape from nomics.New York:Basic Books; Ox- hydraulicthinking into a reconstituted ford:Basil Blackwell,1974. formof reductionism. In treadingthis par- 12. KEYNES,JOHN MAYNARD. The end of ticular path, Clower and Leijonhufvud laissez-faire.London: Woulf, Hogarth were quite rightto identifytheir work Press,1926. withthat of Keynes;they differ from him 13. . Thegeneral theory of employ- onlyin theirdirection of travel. ment,interest and money.London: Macmillan,1936. REFERENCES 14. . "The General Theory of Em- 1. BEVERIDGE, WILLIAM HENRY [SIR]. ployment,"Quart. J Econ.,Feb. 1937, Full employmentin a free society. 51(2),pp. 209-23. London:Allen and Unwin,1944. 15. KLEIN, LAWRENCER. TheKeynesian 2. CLOWER,ROBERT W. "The Keynesian revolution.London: Macmillan, 1949; Counter-Revolution:ATheoreticalAp- Second edition,1966. praisal,"in MonetaryTheory. Edited 16. LEIJONHUFVUD,AXEL. On Keynesian by ROBERT W. CLOWER. Harmonds- economics and the economics of worth:Penguin Books, 1969. Keynes.New York:Oxford University 3. CODDINGTON,ALAN. "Creaking Sem- Press,1968. aphore and Beyond:A Consideration 17. LOASBY, BRIANJ. Choice,complexity of Shackle's 'Epistemics and Econom- and ignorance. Cambridge: Cam- ics,'" British Journal for the Phi- bridgeUniversity Press, 1976. losophyof Science, 1975, 26(2), pp. 18. PATINKIN, DON. "Keynes' Monetary 151-63. Thought: A Study of its Develop- 4. . "Re-thinking Economic Pol- ment,"Hist. Polit. Econ., Spring 1976, icy,"Polit. Quart., Oct.-Dec. 1974,45 8(1), pp. 1-150. (4), pp. 426-38. 19. PEN, JAN. Moderneconomics. Trans- 5. GALBRAITH, JOHN KENNETH. A con- lated from the Dutch by TREVOR S. temporaryguide to economics,peace PRESTON. Harmondsworth: Penguin and laughter.Boston: Houghton Mif- Books,1965. flin;London: AndreDeutsch, 1971. 20. ROBINSON, JOAN. "The Production Coddingtonon Keynesianism 1273

Functionand the Theoryof Capital," ductoryanalysis. New York:McGraw- Rev. Econ. Stud., 1953-54, 21(2), pp. Hill, [1948] 1973. 81-106. 25. SHACKLE, G. L. S. The yearsof high 21. . Economic philosophy. Chi- theory.Cambridge: Cambridge Uni- cago: Aldine,1962; Harmondsworth: versityPress, 1967. PenguinBooks, 1964. 26. . Epistemics and economics. 22. . Economic heresies.New York: Cambridge: Cambridge University Basic Books; London: Macmillan, Press,1972. 1971. 27. . Keynesian kaleidics. Edin- 23. . "What has become of the burgh: EdinburghUniversity Press, Keynesian Revolution?" in After 1974. Keynes. Edited by JOAN ROBINSON. 28. TOWNSHEND, HUGH. "Liquidity-Pre- Oxford:Basil Blackwell,1973. mium and the Theory of Value," 24. SAMUELSON, PAUL A. Economics. Econ. J, March 1937, 47(1), pp. Ninth edition of Economics: An intro- 157-69.