422 CIRCULAR FLOW

In: Kernig, Claus D.: Marxism, communism and Western society. - 1. Absolutism - Civil war.­New York: Herder, 1972. - S. 422-435

CIRCULAR FLOW

A. The Concept of Circular FlowI. The Meaning of the TermIl. Forms of Presentation of Economic Circular Flow

B. Development and Applications of Circulation TheoryI. Quesnay's Tableau ecoflomiquen. Marx's Schemata of ReproductionIll. Developments up to KeynesIV. Keynesian Circular-Flow AnalysisV. Modern Structural Analysis in Peter and LeontiefVI. Circular-Flow Analysis and Modern Economic TheorVII. The Application of Circulation Theory

C. Critical Appraisal and Outlook for the Future

CIRCULAR FLOW 4D

A. The Concept of Circular Flow

I. THE MEANING OF THE TERM

The term "circular flow" is associated in economicscience with the idea that the economic and -- moregenerally - the social activities of individuals or groupsare mutually related. These interactions consist prima­rily in transfers of goods and services on the one handand of money or monetary claims on the other. Suchtransactions between individual economic entities arefrequently purchases of goods and services (flow ofgoods or services from vendor to purchaser) in returnfor a sum of money dependent upon the price of atransaction unit (money flow from purchaser to ven­dor). All transactions in economically relevant objectsduring a particular period of time within a specifiedeconomic sphere (e.g. a national economy) - theexternal sphere as a whole being regarded as a separateeconomic entity - may be seen as a circular-flowpattern.

There are usually too many economic entItIesoperating at once for the construction of a clear and,above all, computable model of circular flow to bepossible without combining several similar entitiesthrough aggregation into higher entities (sectors orpoles). Moreover, the lack of suitable statisticalmaterial (e.g. on the transactions of a single house­hold) or the particular purpose of an investigation(e.g. economic trend analysis) frequently entails a highdegree of aggregation of economic entitites into a fewsectors (poles). Economic entities with as many com­mon features as possible are combined into sectorsaccording to the criterion ofsimilarity in each particularcase - e.g. according to functions (recipients of wagesand salaries, profits, pensions, etc.), institutions (pri­vate households, manufacturing enterprises, publiccorporations, etc.) and regions (economic entities inregions A, B, C, etc.). A more detailed grouping isarrived at by combining these or other criteria (e.g.the private households of wage-earners in region A).

II. FORMS OF PRESENTATION OF ECONOMIC

CIRCULAR FLOW

There are various ways of covering and representingcircular-flow relationships within a given sphere. Inorder to illustrate the specific nature of the differentforms of presentation and their similarities, we shalltake as a basis the Marxian model of the growing capi­talist economy - i.e. of expanding reproduction ­omitting foreign relations and government, but in­cluding saving undertaken by workers.

First, we shall distinguish between three (aggregated)economic entities, viz. the production sector (P), non­entrepreneurial households (W) and entrepreneurialhouseholds (U). The following relationships exist be­tween these economic entities:

I. Production enterprises obtain labour servicesfrom non-entrepreneurial households and the use of

capital from entrepreneurial households. To these realinputs correspond monetary payments in the form ofwages, salaries and profits.

2. The production sector produces consumer goodsand capital goods, and for the supply of consumergoods there exists a corresponding monetary demandby households.

3. The payments of wages. salaries and profits overa period amount to more than the purchases of con­sumer goods within this period by entrepreneurial andnon-entrepreneurial households; also. within a givenperiod, the value of the capital goods produced exceedsthe replacement demand due to depreciation.

These relations can be expressed in a system of linearequations, for which we shall introduce the followingsymbols and abbrevations:

L wages and salaries ( the monetary elJuivalent ofwork done)

G profits ( the monctary clJuivalcnt ofthc yicld ofcapital)C w purchases of consumer goods hy 1I0n-cntreprencurial

householdsC u purchases of consumcr goods hy cntreprcneurial house­

holdsC total consumption

Sw savings of non-cntrepreneurial householdsSu savings of entreprencurial householdsS total savingso depreciationIe replacement of capital assetsIn net investmentI total investment

Y national income ( net value added net nationalproduct)

Y h gross national product (GNP total final output)

The following definitional equations can thus beformed:

Cw Cu j Ie In Yh (i)G L D Yh (ii)Cw Sw L (iii)Cu Su G (iv)Sw Su In (v)

Equation (i) defines the GNP from the produc­tion side: it equals the total of consumer goods andcapital goods produced. Equation (ii) shows thedistribution side. The GNP is divided between en­trepreneurs' and workers' incomes and depreciationcharges for the period. The division of wages, sala­ries and profits for purposes of consumption andsaving is shown by equations (iii) and (iv), whileequation (v) shows the ex post identity betweennational savings and net investment for the period(but not necessarily any correspondence betweensavings and investment plans at the beginning of theperiod). From these definitional equations furtheridentical equations can be derived but these likewisedo not suggest any conclusions about any kind ofequilibrium; they only confirm the accounting equal­ity - following from the definitions -- of certain cir­culation parameters at the end of a period.

From equations (ii), (iii) and (iv) the third definitionof the GNP, from the utilization side, is obtained:

C w Swi- Cu • Sur 0 Y b (vi)

424 CIRCULAR FLOW

p W U v

ITe -~

I

~Cw I LG I Cu Cu Go Ie

~Su In 0

ctIn

------;-

:v b' ~-"- ~_IL_ ~ I I'

Another frequently employed method of presentingcircular flow is the explanatory diagram, each of whosenodes (rings) represents an (aggregate) economic entityand whose arrows represent the transactions takingplace between the economic entities as an inter­relation between two poles in each case. The circular­flow model selected yields the following diagram forthe respective flows of payment:

The directions of all the arrows in the diagram arereversed if, instead of the flows of payments (for con­sumer goods, labour services, etc.), tl!-e correspondingflows of priced goods and services (supplies of con­sumer goods, work done, etc.) are considered. As a

equation Ie t In I. Partial aggregation of this kindadds to the simplicity and clarity of presentation, ofcourse, but it involves at the same time a considerableloss of information, since we are now no longer toldanything about the value of net investments, whichis relevant in the long term. In practical applicationsof circular-flow analysis it is therefore generally veryimportant to maintain a proper balance betweenclarity of presentation on the one hand and informationcontent on the other.

The situation is different if - as in the determina­tion of business operating efficiency - transactionsbetween economic entities are to be recorded in aschedule of accounts on the double-entry system, itthen being possible to keep the number of sectorssmaller because the number of individual items,enteredin an account is in principle a matter of choice. If wetake production sector P, non-entrepreneurial house­holds W, entrepreneurial households U and the changein net assets V as the balancing account, we obtain thefollowing system of accounting entries:

) w U E I F Total

-- w Cu Ie In Yb

- - - L~....._--~--

J - - G

- - 0

Sw Su - In'

b L' G' In

to

fromp

W L

U GE 0

F

Total Y

It states that the GNP conceived as gross nationalincome is used to cover depreciation charges,for the purchases of consumer goods by householdsand for their savings. From equations (i), (v) and(vi) we obtain the identity between the replacementof capital assets and the wear and tear on capitalgoods for the period, recorded in depreciation chargesin monetary terms. Finally, if the GNP Yb is adjustedby the value of the sums set aside for depreciation, adefinition of the net national product is arrived atwhich in our model must exactly equal the total ofwages, salaries and profits, i.e. the national income.

The equation form of presentation can now be easilytranslated into a circular-flow matrix in which thesupplies of the different sectors to one another are ex­pressed in such a way that, for example, only themonetary values corresponding to the actual flowsare recorded (thus wages and salaries are to be under­stood as the value of work included in budgets, forinstance) :

This example also shows that it was not possible tomake do with the original number of sectors (P, W, U)with the circular flow closed, i.e. if the total of thegoods and services supplied by a sector is supposed toequal the total of the goods and services received by it.In the matrix this means that the total of each row ofthe square matrix must equal the corresponding columntotal (i.e. Yb c= Vb"~ etc.). Without the additional sec­tors introduced into the above table (the replacementsector E and the expansion sector F), the total of thegoods and services (represented by the monetaryquantities L and G) received by the production sectorwould have been greater than the goods and servicessupplied to households (consumer goods), and thecircular flow would not have been closed. The aboveexample thus shows that any given open circular flowcan easily be transformed into a closed circular flowthrough the introduction of one or more fictitioussectors (poles). The necessary equality in a closedcircular flow between the value of the flows of goodsand services received by a sector (also called "pole­width") and the value of the flows supplied by thatsector is frequently called the "circular-flow axiom".

In the above example it would also of course havebeen possible to have introduced only one additionalsector in order to cover the changes in net assets (V);then the number of sectors (rows and columns) wouldalso have been one less, and investments would havebeen summed up in aggregate form according to the

CIRCULAR FLOW 425

result of the introduction of the balancing pole (changein net assets), we obtain here also a closed circularflow where the value of all the flows to a pole (thetotal of the different flow widths) equals the value ofthe flows away from it.

In addition to these four main forms of presentationother kinds of circular-flow models have also beenelaborated, such as analogue models in the form ofpiping systems, electrical wiring networks, controlcircuits, etc. With the further development ofcyberneticand other systematic theoretical processes, the con­ception of circular flow as a control circuit - i.e. as adynamically self-regulating system with specific controlmechanisms (e.g. markets, planning authorities, etc.)­will become important. This conception was introducedby W. G. Waffenschmidt(Anschauliche Eirifiihrung ... )in the form of block-circuit diagrams and has since beendeveloped further on the lines of control-system engi­neering (cf. Tustin, The Mechanism of Economic Sys­tems, ... ; Allen, The Engineer's Approach to Econom­ic Models, ... ; Id., Mathematical Economics . .. ; Fohl,Geldschopfung u. Wirtschaftskreislauf, ... ; and others).By analogy with the theory of chemical chain reactionsB. I. Pliukhin and R. N. Nazarova (Upravliaemaiatsepnaia reaktsiia rasshirennogo vosproizvodstva vodnosektornoi i dvusektornoi modeliakh. In Nem­chinov, ed. Primenenie matematiki ... , t. 2, p.342-428)have worked out one-sector and two-sector circular­flow models of the controlled chain reaction of econom­ic growth.

B. Development and Applications ofCirculation Theory

I. QUESNAY'S TABLEAU ECONOMIQUE

The first exact statement of the notion of circulationis to be found in Fran<;ois Quesnay's Tableau econom­ique (1758), which distinguishes between three socialclasses: the tenant farmers (classe productive), thelandowners (classe des proprietaires) and the - in thephysiocratic view unproductive - class comprising allthose who do not work in agriculture (classe sterile).By generalizing the numerical examples given bythe table, the following circular-flow model is arrivedat:

(a) The total gross product of society is produced bythe productive class and distributed, via paymentsof rent to the landowners and purchases by the sterileclass, between all three sectors of the economy.

(b) In successive periods circulation remains un­changed (stationary), since no new production capaci­ties are created nor does technological progress takeeffect.

A clear picture of Quesnay's idea of circulation isconveyed if the flows of goods and services and ofmoney are each represented in respective diagramsslightly modified from Benard, Marx et Ques-

nay ...):

In order to close the circular flow of goods in thisdiagram the credit pole "land" and a fictitious flow ofgoods and services (the "social net product") fromL to Ld are introduced, thus providing a clear illustra­tion of the physiocratic view that land is the source ofall wealth and that landownership is tantamount to thefact that the entire social net product rests in the handsof the landowning class. This is seen even more clearlyin the circulation of money, which can be expressed invalues that have already been balanced and thereforedoes not need "land" as an additional pole to balance it.

Foodstuffs purchases from L

Social net income

Taking into account what has been said above aboutthe correspondence between the various forms of re­presentation and also the circular-flow axiom, we caneasily obtain from these diagrams the general systemof definitional and identical equations underlying theTableau economique and the matrix of macro-economicrelationships. In the third edition of the TableauQuesnay deduced in his "zig-zag diagram" a preliminaryform of the multiplier theorem which is fundamental tothe work ofKeynes and Leontief(see below: B, IV & V).

II. MARX'S SCHEMATA OF REPRODUCTION

Quesnay's pioneering work in the field of circulationtheory remained unnoticed in classical political economyfor more than a century - except by Adam Smith andSimonde de Sism6ndi who made unsuccessful attemptsto develop it further (Smith, An Inquiry ... ; Simondede Sismondi, Nouveaux principes ...) - until Marxtook up the idea of circulation again and elaborated itfurther in both form and content (in Volume 2 of

CIRCULAR FLOW

Both equations reduce to a single condition forsimple reproduction:

Since in the stationary economy no net investmentsare made, capitalists and workers must spend their in­comes entirely on purchasing consumer goods fromDepartment II:

which requires a surplus of value added in the producergoods department exceeding the consumption of themeans ofproduction in the consumer goods department.We can convert (iii a) into an equation by taking netinvestment explicitly into account in both departments(.~Cl and ~C2):

.1c .1c l .1C2 VI m l C2 (iii b)

This relation ~. which according to V. S. Nemchinovamounts to the growth potential (see INVESTMENT) ~shows that the level of overall net investment (~c)

varies positively with the level of value added in theproducer goods department and negatively with theconsumption of producer goods in the consumer goodsdepartment. The modern theories of economic growthand technological progress (qq.v.) take this equation astheir starting-point (see below: B, VI).

only entrepreneurs' savings are available for the for­mation of real capital.) Thus the totality of wages andprofits must not be spent entirely on consumer goods,and the value of the producer goods produced within aparticular period must exceed the depreciation of theexisting productive equipment. Both requirements nowreduce to a single condition for expanded reproduction:

C2 < VI m, (iii a)

III. DEVELOPMENTS UP TO KEYNES

Shortly after the publication of Marx's schemata ofcirculation in Volume 2 of Capita! in 1885 Eugen vonBohm-Bawerk, the Austrian theorist on capital, putforward his "ring schema" conception of circulation(The Positive Theory of Capita!, ...) in which he ar­ranged national capital in "annual rings" (Jahresringe)according to "categories of maturity (Reifek!assen)He regarded capital goods as a kind of intermediateproduct which, when the greatest "maturity of enjoy­ment" (Genussreife) is achieved, become consumergoods. Thus production, as it is conceived by Bohm­Bawerk, is a kind of one-way street, each process­ing stage bringing the products nearer to their ultimatedestination (consumption). However, supplies of prod­ucts "ready for consumption" to production stagesapparently further back (e.g. incandescent lamps readyfor consumption supplied to the mining industry)cannot be explained by way of this theory. This er­roneous view of the structure of production in theeconomy was not rectified until Leontief made hispioneering studies (see below: B, V).

Although in his analysis of expanded reproductionBohm-Bawerk took over from Marx the importanceof positive net investment, by stressing the saving func­tion of entrepreneurs, which is crucial for the growthof the economy, he reinterpreted the inexorable lawsof accumulation to which, according to Marx, all capi­talists are subject. He thus abstracted from the con­ditions for the realization of profit, without whichsaving by entrepreneurs is impossible, and explainedthe lalter's economic function as promoters ofeconomicgrowth in terms of their saving activity.

(i)

(iii)

w

Total(commodity value)

c v s

Derartment

Derartment I(Producer goods)

Derartment II(Consumer goods)

Total(Value clements)

Since by dclinition W j + W2 = W = C + v + s, thecircular flow is closed. For a stationary circulation(simple reproduction) the following conditions ob­viously apply (see also above: A. II):

The producer goods department (Department 1)

must produce within each period sufficient capitalgoods to enable the fixed capital consumed in bothdepartments to be replaced:

- i.e. the value added (wages and profits) in Depart­ment I must equal the replacement requirements of De­partment II; or in other words, Department I must pro­duce sufficient in excess of its own replacement require­ments to be able to satisfy the replacement demands inDepartment II with appropriate supplies of producergoods. A striking feature of this conclusion is that thefulfilment of either one of conditions (i) and (ii) im­mediately guarantees the validity of the other. Inter­estingly enough, quite similar conclusions can be de­duced from the Keynesian circular-flow analysis (seebelow: B, IV).

With a given state of technical knowledge the ex­pansion of existing production capacities - i.e. netinvestments and corresponding reductions in expendi­ture on consumption by capitalists and/or workers ­is necessary for the growth of an economy (see above:A, II). (In Marx, however, the level of wages is estab­lished in the community at subsistence level, so that

('l/piw!). Marx based his circulation analysis upon adivision of the economy into two departments (theproducer goods department and the consumer goodsdepartment) and two classes (workers and capitalists).(Sec DEPARTMENTS OF SOCIAL PRODUCTION.)

According to Marx the output w in terms of valuein both departments equals the sum of depreciation oncapital goods (constant capital c), the wages paid forlabour (variable capital v) and the profits realized(surplus value s). This is illustrated in the followingtransaction matrix:

CIRCULAR FLOW 427

A more important contribution was made in 1910 byWilhelm Lexis (Allgemeine Volkswirtschajislehre, ...)with his attempt to elaborate further Marx's theory ofreproduction in conjunction with Smith's theory ofvalue and Bohm-Bawerk's ring concept and in par­ticular with his efforts to ascertain the dynamicstructure of the process of reproduction. But the funda­mental defects of Smith's theory of value (especially hisfailure to take into account those producer goods whichare consumed) and of Bohm-Bawerk's ring schemataprecluded the construction of a useful circular-flowmodel; and his attempt at cyclical analysis also remainsof purely formal interest.

Other theorists before World War I who used theconcept of circular flow include J.J.O. Lahn (i.e. N.A. L. J. Johannsen, Der Kreislauf des Geldes . .. ) andJoseph Schumpeter (The Theory of Economic Develop­ment . .. ). On the other han(i. both the historical schoolin Germany and the neo-classicists (particularly thevarious marginal utility schools) practically ignoredthis aspect completely. An exception in this respect wasthe Lausanne school founded by Leon Walras (Ele­ments o{ Pure Economics . .. ), whose conception ofmicro-economic equilibrium (see below: V and VI) isbased upon the idea of a stationary circular flow.

Of the Marxist economists Lenin (On the So-CalledMarket Question, ...) was the first in 1893 to take upthe Marxian schema of expanded reproduction, and inhis numerical example he took into account changes inthe organic composition of capital (~) and the effectof technological progress. Divergences from the ex­ponential" trend underlying Marx's numerical exampleare bound up with this in the trend of the circulationparameters over time; but these differences can be com­puted in the form of an elementary chain index. Theconclusion which can be drawn from Lenin's dynamic

model, which includes Marx's schema as a special case,is that it is necessary to observe certain relationshipsbetween the trend of the difrerent circulation parame­ters and that there is a change in the structure of pro­duction in favour of the capital goods sectors underthe conditions of technological progress. These findingswere later realized in practice at the time of the firstfive-year plans -- i.e. during the extensive phase ofSoviet economic planning --- when heavy industry wasaccorded priority (see INDlJ STRIALIZATION, B).

In Marxist discussion of "realization" crises and ofthe theory ofeconomic collapse considerable differencesof opinion arose out of the question of the correctinterpretation of Marx's schema of reproduction,especially between M. I. Tugan-Baranovsky (Theoreti­sche Grundlagen des Marxismus, ... ) and Rosa Luxem­burg (The Accumulation of Capital, ...), the latterdeducing that a closed circular flow of expanded repro­duction was impossible under the implicit assumptionsof simple reproduction. She thus attempted to explain- erroneously - the need for the imperialistic opening­up of foreign markets as a means of ensuring adequatedemand and also the ultimate collapse of the capitalistsystem (see IMPERIALISM).

The considerable increase in the state's share of thenational product during and after World War I gaverise to a new stage in the development of circulationtheory; the quantitative effects of the growing stateactivity in the economy as a whole -- observed alreadyin 1883 by Adolph Wagner (Finanzwissenschaji, ...)~­now had to be investigated. The overall balance of theUSSR for 1923/24 represents the first attempt to recordstatistically the quantitative relations between the dif­ferent sectors and to set them out in a "chessboardbalance", the general features of which are illustratedin the following specimen balance:

Balance of Productive Equipment and Buildings in the USSR for 1923/24

Uses

Sources

A. Productive EquipmentAgricultureIndustry

Stocks Total

Total

B. BuildingsAgriculture

Total

Total A + B

C. ImportsDistribution CostsInitial Stocks

Total A + 8

42X CIRCULAR FLOW

Although the collection of statistically and economi­cally important data is incomplete and the classificationof economic sectors is very rough, this Soviet balancesheet represents a significant step towards the produc­tion of the modern inter-industry transaction matrices(see INPUT-OUTPUT ANALYSIS), which are very impor­tant as instruments both ofcentralized economic plann­ing in the socialist countries and of general macro­economic control in the advanced capitalist countries(see below: VI).

In the 1920s and 1930s German researchers into cir­cular flow devoted especial attention to the monetaryaspect and that of business-cycle policy (cf. Lederer,Der Zirkulationsprozess als zentrales Problem derokonomischen Theorie, ... ; Lowe, Wie ist die Kon­junkturtheorie Uberhaupt moglich? ... ; Neisser, DerKreislauf des Geldes, ... ; GrUnig, Del' Wirtsclwfis-Kreis/auf; ;and Fohl, Ge!dH'/u)[J./il1lR unci Wirtsc!lafis-kreis/auf; ). The Great Depression and later national-socialist economic policy in particular acted as con­crete incentives to research and also determined thetheoretical problems. That circular-flow research andthe national-socialist armaments and wartime industryco-operated is apparent from the "Control Panel of theGerman War Economy 1942/43", which was intendedto show the most important inter-industry relationshipsfor the war effort (i.e. the production prerequisites).

IV. KEYNESIAN CIRCULAR-FLOW ANALYSIS

The impact of world-wide mass unemployment andthe decline in production during the Great Depressionof 1929-33 caused political economy in the non-Marxistcountries to take a new line. Liberal and vulgar-econo­mic concepts in neo-c1assical theory ~ particularlythe notion of a kind of "prestabilized harmony" of thecapitalist economic system (e.g. Say's Law, quantitytheory, the need for "purging crises")~ were abandon­ed and now concepts of circulation theory were againused to explain and eliminate macro-economic im­balances. This came about in part through deliberaterecourse to Marx's theory of reproduction ~ e.g. byMichal Kalecki (Essays in the Theory of EconomicFluctuations, ... )~ and in part through the rediscoveryof circulation theory, in particular by John MaynardKeynes (The Genera/ Theory of Emp/oyment, Interestand Money, ... ).

In Keynes's work a more or less automatic balancingof the supply of savings and the demand for investmentfunds through the market rate of interest (q.v.) isreplaced by a system of more realistic behaviouralhypotheses (functions). Both micro-economically andmacro-economically the consumption (q.v.) and sav­ings of private households depend primarily upon thelevel of the national income, while the volume of fixedinvestment planned by entrepreneurs is also determinedby the newly formulated principle of the "marginalefficiency of capital". The demand for money of thoseactive in the economy is no longer determined solely

by the transaction motive - the need to be able todefray current payments - but also by the speculationmotive, the desire to have money available for lucrativepurchases of securities. Thus in the Keynesian systemmacro-economic equilibrium is characterized by acombination of real national income, price levels andrates of interest which harmonize saving and investmenton the one hand and the supply of and demand formoney on the other.

The above-mentioned conditions however by nomeans ensure full employment in the capitalist marketeconomy. According to Keynes an inadequate volumeof total demand can lead to involuntary unemployment,even if the workers or the trade unions accept a "mar­ginal productivity wage" (see PRODUCTIVITY; ECONOM­IC THEORY). But real wages and prices (qq.v.) that arerigid downward are also for Keynes important causesof a state of underemployment, which could possiblybe an equilibrium state. This, he maintains, is generally.associated with insufficient readiness on the part ofentrepreneurs to make further investment equal to thesupply of savings, thus causing an open deflationarygap. Many other disturbing factors, such as a lowinterest-elasticity of investment and savings or a floorto interest on the money market (i.e. the liquidity trap)make heavy demands on the stability of the system,which in reality cannot always be met.

From this detailed macro-economic analysis it ispossible to deduce in principle those changes in macro­economic data which must be untertaken if the stateis to eliminate economic disequilibrium. This mayinitially take the form of indirect macro-economic inter­vention: influencing the rate of interest by changingthe volume of money or the bank rate, influencinghousehold consumption and saving behaviour byvarying taxation or encouraging various ways of utiliz­ing income or changing income distribution (see below:VI), damping down or stimulating business investmentdemand by varying the rates at which assets are writtenoff, etc. The most important consequence of the Keyne­sian conception of the circular flow is however the com­pensation for the lack of private investment activity bymeans ofdirect intervention in the form ofcompensato­ry government expenditure, especially for public invest­ment projects, with the acceptance of deficit spending.An important role is played here by the multiplier theo­rem, which is based on the simple observation that pur­chases in one part of the economy generate further pur­chases in other sectors -- the more so, the less is saved orhoarded of the monetary payment received in each caseat each economic stage. This can be deduced in theelementary circulation system (see above: A, II) forthe so-called investment multiplier, for example, whichin the simple model equals the reciprocal of the marginalpropensity to save (dS/dY) and gives the total increasein national income (~Y) when investment expenditureis increased by the amount ~I.

In order to make clear the dependence of consump­tion, and hence of saving, upon the level of the national

CIRCULAR FLOW 429

v. MODERN STRUCTURAL ANALYSIS INPETER AND LEONTIEF

By differentiating (i) for Y and (ii) for L we get onthe one hand the definitional equation

pricesfinal productslabour supplied

PI' P2YI. Y211,1 2

wage-ratesshipments from i to j(i, j 1,2)

tative description of types of structure. Particularlyimportant here is the formation of structure coeflicients.which relate the individual llow-widths of a pole to itsoverall pole-width. Owing to the high degree of general­ity of formal circulation theory. the importance ofthese quotients can vary considerably in practice. Inmacro-economic models (see ahove: A. II) they re­present. for example. the ratio of savings or of wagesand salaries to the national product: hut if individualbranches of industry are chosen as poles. then weobtain the corresponding input-output coefficients.Thus to this extent Peter's formal circulation theoryincludes Leontief's input-output analysis (see below)as a special case. Like Leontief. Peter also made hiscircular-flow system dynamic hy comhining the ac­counting equation system with certain growth hypo­theses (e.g. exponential functions). in order to ascertainthe process of expanded reproduction too in terms ofcirculation theory.

The studies of Wassily W. Leontief (e.g. HarvardUniversity. Harvard Economic Research Project. Stud­

ies in the Structure of'the American Econolllr, ... ) focuson the investigation of inter-industry relationships,while the economic entities outside industry (e.g.households, the government. etc.) form the reSidualsector as final consumer demand in the open Leontiefmodel. Thus, to begin with, market transactions are leftout of account or clearing of the market is assumed,and flows of investments are likewise recorded in aseparate investment account (see INPUT-OUTPUT ANAL­YSIS). The total of the supplies Xj priced at the price Pj

and delivered by each sector j (j I. .... m) to itself(self-consumption), to the other m I branches of in-dustry and to the residual sector (final consumer de­mand) must, according to the circular-flow axiom,equal the value of shipments received plus the factoroutputs received from the residual sector. If the residualsector is identified with the "households" sector andits factor outputs with "labour", then the followingdiagram is obtained for two industries:

(i)

(ii)

(iii)

(ia)

(iia)

~Y. dSdY

or

~Y

~I

dSdY

Y C (Y) S (Y) (consumption side)Y C (Y) I (production side)

which still applies approximately, iffinite but sufficient­ly small changes in income and investment (~Y and ~I)

are considered. Then follows the multiplier effect, al­ready formulated, of increases in investment upon thelevel of the national income:

between the marginal propensity to consume and themarginal propensity to save, and on the other hand theequation

dY (I dC)dI dY

The most important contribution towards the for­mulation of a complete formal theory of circulationwas made by the German economist Hans Peter (Struk­turlehre des Wirtschaftskreislaufes, ... ; Mathemati­sche Strukturlehre des Wirtschaftskreislaufes, ...) who,using methods of topological group theory - graphtheory in particular - evolved a system for the exactrecording and presentation of circulation relationships(e.g. pole-width, the circular-flow axiom) and a quanti-

M . .--Ld~

dY

Multipliers for changes in government expenditure,exports and imports, the volume of money, etc., canbe deduced from systems of circular-flow equations inthe same way; but the extent of the actual multipliereffect usualiy depends upon other factors which arenot taken into account in the circular-flow model, sothat the isolated consideration of the various multipliereffects "renders absolute a state of affairs which inobjective reality appears only as a tendency" (MUlier,Die hurgerliche Kreislaufiheorie, ... p.I44).

Keynesian circular-flow theory has had an extremelystimulating effect upon non-Marxist economics andmodern economic policy (the controlled market econo­my, state monopoly capitalism), and is being acceptedincreasingly in the socialist countries as well (despitemany reservations) for the analysis of short-run prob­lems (see below: B, VI & VII).

income, as postulated by Keynes, consumption andsaving are represented as functions of national income.We thus obtain these macro-economic equations:

430 CIRCULAR FLOW

(ii c)

This inverse matrix can - by formal analogy with themacro-economic consumption ratios of Keynesiananalysis - be interpreted as a multiplier: the increaseof final consumer demand by t1.y leads initially to anincrease of the same amount in gross production. Butsince additional inter-industry supply commitmentsnow arise in proportion to the coefficients aij, we obtainthe total change in the output as the sum of the differentrounding-off adjustments

goods and dynamic production structures are takeninto account different results are obtained.

However, the open Leontief system need not be re­garded as a special form of a micro-economic totalequilibrium only (see below: VI); it acquires much morepractical importance through the underlying assump­tion of linearity, which makes it possible to use themodel with the methods of linear algebra for the widestpossible variety of computations. At the same time theinput-output coefficients vary widely in significancedepending in each case upon the object of the investiga­tion. In ex post studies they characterize, for example,inter-industrial relationships shown in statistical form,or the actual expenditure of materials and labour;ex ante, the model is suitable both for making forecastsand for general economic programming (in the lattercase the coefficients have the significance of prescribedexpenditure norms). Comparison of the coefficientsover time enables technological progress to be ascer­tained, while the comparison of ex ante and ex postcoefficients can be used as a measure of plan fulfilmentfor the different sectors of the economy and as a basisfor the future programming, revision and checking ofplans (see below: VII).

Particularly important is the computation of thegross industrial production and labour input requiredfor the desired - or planned - final consumer demand(components of the national product [q.v.]), which canbe calculated from (ii) by solving the set ofequations or,more generally, by forming the inverse matrix (i.e. theLeontief inverse):

(i)

(ii)

(iii)

(iia)= national

pro~uct

totalwages

~

r 2 ' X2 W = w(r\x l + r2x2) = w(11 . 12 )

+

(l aldx 1 aI2x2-~YI

a 2 I X 1 -I (I a 2 2) X 2 y 2

national product total wages and salaries (c-= nationalincome), which not only represents a simple identitybut can also be conceived as a condition for equilibriumif prices are regarded as variables and the assumptionof proportionality (linearity) underlying the Leontiefsystem is made with respect to inter-industrial rela­tionships and for factor input, and the input-outputcoefficients are accordingly defined aij = ~ and thelabour input coefficients rj ~. Then fro~ the ac­counting equation of the poles we obtain the set ofnon-homogeneous linear equations

or more generally, in matrix notation:

(E A) x y E unit matrixA = (ai) i =~ I, , m

j I, ,m

and the condition that the labour input must not exceedthe total available labour force 10 :

By adding the accounting equations for the two in­dustrial branches, we obtain the equation

Now we must define the price system which for agiven wage rate w brings the circular flow into balancein accordance with condition (i). In this simple casethese equilibrium prices - also called shadow prices(Dantzig), multipliers (Peter), solution multipliers(Lur'e), objectively conditioned valuations (Kantoro­vich) - are obtained by multiplying system (ii) line byline by the corresponding prices. Since according tocondition (i) the total wage sum must equal the pricednational product, we obtain the schema

and directly from this, vertically, the equilibrium pricesystem

(I - a\l)p\ - a21 . P2 = wr l more generally: A' = (aj;)(E -- A ')p = w . r

- al2 . PI + (I - all) P2 = wr2 (ii b)

This optimal price system satisfies the Marxianlabour theory of value both in the sense that prices(q.v.) are proportional to the average labour expendedand also in that "labour" appears as the sole value­creating productive force - in the optimum case pro­fits according to schema (iia) are zero - and finallylabour is conceived as a homogeneous factor; thus thevarious activities of labour are reduced, as in Marx, to"simple labour". This conclusion is based however onthe unrealistic assumption that only labour and currentinput are included in social production (q.v.). If capital

~X == ~x(1) + ~X(2) + ~X(3) ••• - ~x(nl

= ~y + A ~y + A 2 ~y + ... -j- An - I ~y

= (E + A + A 2 + ... + An - I) ~y

which in the case of production (final demand produc­ing) systems converge towards 0, and we thereforehave for n -- 00 the total (E-Af 16.y = 6.x analogousto (iic).

Leontief and many other writers after him haveamplified and modified the model for practical andtheoretical purposes. Instead of clearly defined produc­tion processes factor substitution in the form of alter­native input-output coefficients is for example intro­duced - in a similar way to neo-classical theory ­so that the choice of the processes to be used is deter­mined by the basic object being aimed at and/or byadditional constraints (thus giving rise to an optimiza­tion problem). The time dimension of circular flow

CIRCULAR FLOW 431

embraces, amongst other things, changes in technicalcoefficients over time (technological progress andcapital formation) and the introduction of time lagsbetween the beginning and the end of individual pro­duction processes; investments are also taken intoaccount. In order to explain the demand for capitalgoods Leontief used the acceleration principle, accord­ing to which sectoral investments should be propor­tional to the change in gross output. For the purposesof forecasting the computation model is often expandedin that assumptions of proportionality are also madefor exogenous variables such as wages and salaries,profits, tax yields, etc., so that not only the attainablefinal consumer demand, but also the total of wages,profits and similar values are related to the level ofgross output - i.e. the overall turnover of the economy.In Peter's more generalized circular-flow model on theother hand the variables are determined endogenouslyand not - as in the Leontief model- subsequently in­corporated by means of additional hypotheses oflinearity. By contrast, the great importance of theLeontief system lies in the possibility that it affords ofapplying efficient mathematical procedures (e.g. linearalgebra, optimization theory) in order to arrive atresults which are directly usable in practice (see below:VI).

VI. CIRCULAR-FLOW ANALYSIS ANDMODERN ECONOMIC THEORY

Formal circular-flow theory in the sense of Peter'ssystem describes only the necessary conditions forbuilding economic models and is thus a component ofevery theory but not in itself a complete theory. Signi­ficant conclusions can be drawn from circular flowmodels only when certain specifications which definethe system are incorporated. These include the choiceof poles, the circular flows postulated in each particularcase and the assumed mechanism of interactions(behaviour equations, production coefficients, changesover time, etc.). The important heuristic value offormaltheory for developing meaningful economic theorieslies in the obligation it places upon the economist todefine concepts exactly and to make consistent modelassumptions as to the behaviour of poles, but it cannotensure the practical relevance of the models. For this,however, the possibility of empirical verification orrefutation by means of the data of economic statisticsexists (see below: VII). This constitutes an importantfield of econometrics (q.v.) and has already been highlyformalized through the application of modem mathe­matical methods, e.g. regression analysis and the sta­tistics of probability (see STATISTICS). Historically, thisassessment of circular-flow theory is confirmed forinstance by the revolutionary effect of the theories ofMarx and Keynes (see below: C). The fruitfulness ofthe flow-theory approach is apparent today in a widevariety of practical and theoretical economic fields.Here we give no more than a brief outline of the mostimportant stages in its development.

The contradiction in Keynesian theory between sta­tionary total demand and long-term expanding produc­tion capacities has made dynamic analysis necessary.The demand of entrepreneurs for capital goods hastherefore been explained - more realistically than inthe Keynesian system - in terms of the change indemand for consumer goods or in the national income(the acceleration principle). The trend of the nationalincome over time is then always a function of the ratiobetween the multiplier and the accelerator. From themodels of Paul Samuelson (Interactions ...) and JohnHicks (A Contribution . .. ), both (increasing, decreas­ing or constant) cyclical fluctuations and also the con­tinual growth (or continual decline) of the nationalincome can be deduced. The circular-flow models ofRoy Harrod (An Essay in Dynamic Theory, ... ; Id.,Towards a Dynamic Economics, ... ) and Evsey D.Domar (Capital Expansion ... ; Id., Essays . .. ) areclearly concerned with a question of growth theory ­viz. the determination of an equilibrium rate of nationalincome growth which accords with the sales expecta­tions of entrepreneurs, the growth of the gainfully­employed population and the linear production struc­ture of the economy ("growth on the knife-edge").

The circular-flow model

(average)y = C(y) + S(V) .~ CrY) r sV marginal

saving ratio (i)Y = C(V) + I( l1V) C(Y)r a l1Y a accelerator (ii)

gives this growth condition direct by equation: theequilibrium growth rate gy is accordingly

(iii)

- i.e. equal to the ratio between the (marginal) savingratio and the accelerator. The accelerator need not beregarded here as a decision parameter of business in­vestment activity. Ex post, it can be identified with the

marginal capital output ratio ~~ - a technical variable

-if investments for replacement are left out of account.Finally, if a linear production structure is assumed thenthe marginal capital output ratio equals the averagecapital output ratio KjY, and the accelerator is deter­mined purely by technical factors.

The demand-orientated, post-Keynesian approach,as further elaborate.d by Joan R0binson (The Accumu­lation ofCapital, ...), Luigi Pasinetti (Rate ofProfit ...)and others, is opposed by the supply-orientated,neo-classical growth hypothesis. This postulates agree­ment between investment plans and saving plansand its basic model (Solow, A Contribution tothe Theory of Economic Growth, ...) defines thegrowth path' of the national product - assuming con­tinual substitutability of capital and labour (qq.v.),perfect competition and neutral technological progress- towards which, with a given savings ratio and aconstant rate of technological progress and populationgrowth, the system converges. The interdependencepostulated in the post-Keynesian approach of the

4J2 CI RCLJ tAR FLOW

national income and the national product on the onehand and the savings ratio and the capital output ratioon the other is of particular interest for socialist growth­pbnning. hecause undn the conditions of socialisteconomic plannin~ agreement hetween savings andinvestment. and thus also hctween the savings ratioand the investment ratio, can be achie\ed more easilythan in a state-influenced capitalist market economy(duhhed "state monopoly capitalism"). hH' this veryreason. however. the orientation of' the nco-classicalmodels towards supply is more in keeping with a cen­trally planned socialist economy than with the situationin the advanced industrial countries of the West. wherethe main prohlem is usually to increase ctfective demandthrough advertising, leisure activitIes. etc. I.e., inMarxist terminology. to realize the total social productand the surplus (cL Baran. Sweezy. ,1;foIlOfJOfr COfJiwl,

... ). Both these approaches to growth have been takenup and developed further by Manist economic theoristsas well (e.g, by Dobb, AI/ Lssa.l" o///:'CII//olllic Groll,th . .. ;and Kornai. il,lat!wl1Iatical Plal/I/illg ... ). Similarstudiesby Marxist economists (e.g. Volkonskii. V. A. Skhemaoptimal'nogo per~pektivnogo planirovaniia i otsenkiresursov.ln Nemchinov.ed. PrimellCllienwt£'llwtiki . .. ,t. J, p. 88· 106; Kornai. Mathematical Plalll/i//g ... ) havebeen prompted by thedetermination ofoptimum growthpaths of consumption (Phelps, The Golden Rule of' Ac­cumulation.... ; von Weizsiicker. Wac!zstlllll . .. ).

A particularly fruitful application of circular-flowanalysis lies in dynamic distribution theory. On thebasis of a Keynesian two-class model including savingby workers' households and constant class-specificrates of saving Nicholas Kaldor (Alternative Theoriesof Distribution.... ) has demonstrated the dependenccof income distribution upon the level of investment.According to his thcory, which bears a striking resem­blance to corresponding Marxist ideas, a low share ofwages in national income i.e. a high rate of surplusvalue and a high rate of consumption by workers arebound up with high rates of protlt and growth; thisadmittedly assumes that the goods produced can all besold -- possibly on foreign markets (see IMPERIALISM).One weakness of the model is however the lack ofclarity of the causal relationships postulated.

With appropriate interpretations in terms ofcircular­flow theory (e.g. Stobbe, Ullfers/lc!zwIRel/ ... ; WoIC'itet­tel', Verteilungstheorie ... ), Kalecki's theory of thedegree of monopoly also provides valuable insightsinto the process of income distribution under conditionsof monopolistic competition (Kalecki, The Deter­minants of Distribution ... ). If the supply situation isalso incorporated into Kaldor"s model (Preiser, Wachs­tum ... ) -- especially with regard to the privileged

position of the capitalist owners of the means of produc­tion as compared with the workers dependent upontheir wages ~ then the superiority of the classical andMarxist approach over neo-c1assical theory in the ex­planation of the capitalist process of reproduction interms of distribution theory is clear.

The development of the micro-economic thcory ofeconomic cquilibrium (q.v.) is also closely bound upwith the economic theory or socialism (Langc. Taylor.011 t!le t:collolllic 7J/eorl' o!.",·ocia!islll . ... ) and with themathematical theory of optimization. which in turn isalso a constituent of formal circular-flow theory (seeabovc: B, V). The formal further development or micro­economic equilibrium theory (Dorfman et al. LinearProgrwnming , .. : Debreu. Theory of Value, ... ) inparticular shows that the subjectivist explanation ofequilibrium price ratios by means of the relationshipbetween the corresponding marginal utilities is nolongcr necessary. cvcn though to date certain conceptsof utility theory are still involved in the explanationofdemand (see Eco:--,:oMlcs). With thc stock of resourcesand the production coeftlcients taken into accounL theequilibrium price system assumes on the supply side thesignificance of a schema of objectivcly conditionedvaluations (shadow prices). the socio-political charac­ter of which is explained by the constraints and objec­tives of socia: reproduction (e.g. the property systcm,external effects, political objectives). Many of thecharacteristics of the formal micro-economic equilibri­um models are more reminiscent of actual representa­tive socialist economies than of capitalist economies;the market model of perfect competition on which theyare based can easily be structurally convertec.i into aplanning model of perfect computation (Altvater,Rationalisierung ... ; Kornai, Mat/wl/Ultical Planning... ; Ward, The Socialist ECOII0111.1", ... ). This situationclearly refutes the opinions voiced by liberal economistssuch as Ludwig von Mises, Friedrich August von Hayekand Lee Robbins on the possibility in principle orpractice of efficient allocation in economies planned onsocialist lines.

Finally. the theoretical and practical importance ofthe Marxian analysis of expanded reproduction is con­firmed in modern sectoral growth theory, which fre­quently adopts the division of th~ economy into a con­sumer goods department and a capital goods depart­ment. But in content, too, interesting connections withcirculation theory are often found- for example, be­tween .I. von Neumann's growth model. the Wah-asmodel and the schema of expanded reproduction.Michio Morishima (Economic Expansion ... ) pointedthis out. Altogether, there is an unmistakable tendencyin both Western and socialist economic theory todevelop interdependent and dynamic systems insteadof static and partial models.

VII. THE ApPLICATIO'.; OF CIRCLLATIO!\ THEORY

As a constituent of economic theory, circular-flowanalysis is not dependent upon empirical verification.Its importance for both Western and Marxist economicsis due precisely to the fact that it does not belongto the sphere of praxeological rules of behaviour beyondempirical demonstrability, which predominate espe­cially in the micro-economic theory of the household.

CIRCULAR FLOW 433

Although circular-flow analysis can also be used forpurely theoretical questions for studying the func­tional principlcs of the capitalist economy as in the caseof Marx and Keynes, for instance, or for abstractdidactic description e.g. for depicting the free econo­my as an object of economic cognition as Ellinghausdid extensive formal development is due chieOy toits relevance as an instrument for the study (diagnosis),forecasting and active creation (planning) ofempiricallyobserved or observable interactions and transactionsbetween the various economic entities.

Thc totality ofconcepts, data and procedures involvcdin the statistical calculation of these processes are sub­sumed in the Western countries under the term "nationalaccounting" (q.v.). Both the requirements as to theprecision and detail of the national accounts and theconclusions to be drawn from them for state activityvary according to the quantitativc extent and qualitativefunction of state economic activity.

The methods of statistical data collection, classifica­tion and systems of accounting applied today in theWestern industrial countries are largely standard (e.g.the OEEC standard system of 1952 and 1960), so thatthe value of the accounts for international comparisonsis increasing all the time. However this is also leadingto the development of a more or less standard set ofinstruments of economic policy, although their actualuse depends in part upon the socia-political situationin each case. In principle this should also make for thebetter international co-ordination of measures ofeconomic policy, but because of the different socialsystems (e.g. the state as comprising the productionsector), statistical concepts (e.g. the national productseen not as a flow of goods, but as a stock of physicalend products) and demarcations (particularly theseparation of the productive and unproductive spheres),complete comparison with the corresponding data ofcommunist countries is not possible.

The essential difference between Western economicstatistics and those of communist countries arises inparticular out of the varied nature of the functions ofthe authorities responsible for economic policy. In thecase of the macro-economic overall control principallyapplied in capitalist economies the identifying and in­fluencing of macro-economic aggregates such as privateand public demand, investment, etc., are particularlyimportant. Control of this nature serves primarily tostabilize the economic trend, and not so much to planthe economy's structure and growth. In the Westerncountries detailed input-output analysis has thereforenot yet acquired the importance which it has in thesocialist economies, where input-output coefficientsare used not only descriptively but also prescriptively,as indices of economic efficiency, norms of direct andfull expenditure or as price standards. Moreover, theelaboration of binding perspective and annual plansrequires a higher degree of precision than the drawing­up of forecasts and guide-lines which, after all, carry noobligations with them. Finally, certain problems such

as the computation or macro-economically optimalproduction methods and product prices (macro-eco­nomic and sectoral optimization) or the co-ordinationof various economic plans (plan consistency) havehitherto been excluded from the functions of govcrn­ment economic policy in the Western countries andhave been left to the market i.e. in clfcct to co­operation between monopolies ano interest groups.But it was precisely because of these and similar prob­lems on the other hand that economists in the commu­nist countries came to develop special methods ofmathematical economics to bciiitate the drawing-upof overall regional and sectoral plans (planomctrics).

Progress in the use or mathematical methods inoverall economic planning is evident in, for example,the changeover from pure computation models of theLeontieftype to models of optimaI perspective planning(Kantorovich, L. V., and Makarov. V. L. Optimal'nyemodeli perspektivnogo planirovaniia. In Nemchinov,ed. Pril1lenenic IIw!cl11lJfiki ... , 1. 3. p.7 X7), growthplanning (Gavrilets, ihid.) and interregional planning(Gladyshevskii, Gavrilets, ihid.), as well as in the in­creasing usc of methods of non-linear optimization andin the building of multilevel planning models (Kornai,Liptak. Ketszinti tervezes, ... ; id., Liptak. Two-LevelPlanning, ... ). In view of the conflicting tendenciesamongst communist economists and politicians it isstill not possible to foretell with any great accuracy howfar the decentralization of decision-making functionsin the "New Economic System of Planning and Manag­ing the Economy" will cause market relations andeconomic incentives to takc the place of centralizedplanning computations (see PLA!\iNISG). Nevertheless,socialist economic researchers are showing increasinginterest in the development of macro-economic circular­flow models - e.g. for determining optimal investmentratios or efficient consumption paths (cf. Mine, 8., andPrzelaskowski, W. Uravneniia sotsialisticheskogo na­kopleniia. In Nemchinov, ed. Pril11enenie nwtel11atiki

... , 1. 3, p. 303-11). In a certain sense symmetrical withthis is the increasing development in the Western coun­tries of sectoral models for forecasting and planningpurposes (e.g. Ragnar Frisch's "Oslo" model; cf.Frisch, Osnovnye cherty promezhutochnoi modeli"Oslo". Ihid., t. 2, p.429-531 ) and their growing use forstatistical computation (e.g. the French regional plann­ing models).

The combination of sectoral and macro-economiccircular-flow analysis has proved particularly useful,as shown in the expanded input-output table (inter­industry transaction matrix) overleaf.

Quadrant I contains the input-output coefficients(or also only inter-industry sales) in money terms;Quadrant II breaks down the various sectors of finalconsumer demand (households, the state, etc.). InQuadrant III the primary inputs (labour, use of capital,etc.) are entered in money terms. Finally, the redistri­bution of the national product can be represented inQuadrant IV. That this circular-flow table is useful and

434 CIRCULAR FLOW

BIBLIOGRAPHY

valuation, it does not expound an e\.:onomic theory. Atthe present stage of economic knowledge '"neo-classicalsynthesis" (Samuelson) can therefore also only beregarded as an attempt, with no objective foundation,to reconcile the praxeological rules of behaviour ofmicro-economic theory with the phenotypical descrip­tion and regulation of the capitalist process of reproduc­tion (see ECONOMIC THEORY).

The practical utility of circular-flow analysis in thetwo social systems should not however lead to the falseassessment of circulation theory as an a priori objectivetheory, independent of socio-political conditions. In sofar as it actually exists as formal circulation theory (asin the case of Peter), it does not yet constitute an econ­omic theory. But the identification of the abstract modelwith spezific objects of reality and, still more, thespecification of functional relationships - the forma­tion of economic models and theories - themselvesindicate that a dialectical relationship has been estab­lished between social reality and economic theory. Thisrelationship becomes explicitly political through thepractical application of circulation theory. which withregard to both its possible applications and goals isdetermined directly by political and social conditions.

The scientific structure of circulation theory leads tosurprising ambivalences in its application. Thus thecircular-flow analysis of the liberal economist Keynesfound its most logical application in fascist social sys­tems, and formal circular-flow research received power­ful stimuli from the wartime economy and armamentsindustry, especially in Germany. And for its partKeynesian circular-flow analysis, which demonstratedthe possibilities of eliminating macro-economic dispar­ities and thus (for a time. at least) maintaining capital­ism, again took up Marx's criticism of the contradic­tions inherent in the capitalist process of reproduction.

Many problems remain to be solved by circulationtheory. Special importance attaches here to the con­sistent imputation ofobjectively conditioned valuations(shadow prices) in dynamic systems of the social repro­duction process, which must replace subjectivist marketor planned prices and which as indicators of the socialextent of the scarcity of goods make possible the reallyefficient allocation of material and intellectual re­sources. The growing importance of circular-flow ana­lysis, which in all its basic categories can be traced backto Marx, seems at all events today to confirm JoanRobinson's observation at the end of her Essay onMarxian Economics: " ... and if there is any hope ofprogress in economics at all, it must be in using aca­demic methods to solve the problems posed by Marx"(Robinson, An EssaJ' ..., p.IIS).

Allen, R. G. D. The Engineer's Approach to EconomicModels. In Economica, n. s. 22, 1955, no. 86: 158-68. - Id.Mathematical Economics. 2nd ed. Ldn, N. Y., 1959. 1st ed.1956. - Altvater, E. Rationalisierung u. Demokratisierung.In Argument 8, 1966, Nr 4: 265-89. - Baran, P. A., andSweezy, P. M. Monopoly Capital. N. Y., 1966. - Benard, J.

lin~d Ih:mand

mel I S Export-< " • lalL' Import

I

III

Uses Industry

Industry I

2

Sources

Depreciation

Wages Salaries

Profits(Surplus)

Indirect Taxes

(jross SocialProduct

m

Productive( 'onsumption--------t-------+--+---+-----+---+----

to a very large extent independent of any politicalsystem is indicated by the fact that it is used frequentlyin both the West for national accounting and in com­munist countries for the overall balancing of accounts.In view of this the efforts of some socialist economiststo demonstrate a fundamental difference between theresearch into circular flow being done in the East andin the West do at times seem rather ridiculous; they areprobably due largely to ignorance or to the misinterpre­tation of Western research in this field.

C. Critical Appraisal and Outlook for the Future

The importance of circular-flow analysis in bothsocial systems is due chiefly to the possibility which itaffords of using the computation of micro-economicand macro-economic interaction mechanisms (such aselasticities of demand, multipliers and capital outputratios) to estimate the effects of state intervention inthe economic structure. There is hardly any doubtabout the need for and practical advantages of suchcomputations in the communist economies. The ques­tion ofwhether government intervention in the capitalistmarket economy is compatible with the laUer's func­tional principles cannot, however, be resolved by themethods of circular-flow analysis. But the experiencesof the Great Depression and recent economic trends inthe USA and West Germany show clearly that govern­ment intervention is essential if this system is to functionproperly. This gives rise to prognoses about the futureof the capitalist system, but these cannot be said tobelong to the realm of theoretical economic analysis ofthe capitalist reproduction process. And the designationof this mixed private-enterprise and state-controlledsystem as a "social market economy" (Miiller-Annack),"organized capitalism" (Hilferding) or "state monopolycapitalism" - which themselves each express a parti­cular view of the role of the state in capitalist society(see CAPITALISM; STATE) - only implies a political

CIVIL WAR 4J5

'tan et Quesnay. [n Frl/ll\,ois QI/""W\' "I la pln·.liIICl"lIli".\'ul I. Paris,195X: Ill5 30, Biihm Vlln Bawerk (von Biihm­Bawakl. E. 7il" Posilivl' 7hl'oryofCllpi(al, Ldn, N.Y" IX91,Ger.: P/lsi(iv" Th"urie d", /\api(als, Innsbruck, IX89. (/o.:lI{Ji(III1/ /\lIpiwl::ill.l. Abt. 2.) Bombach, G. Kreislauftheorie u.\<llkswirtschaftlichc Gcsamtrechnung. [n Jh, SO::illll1'is.l, 11.1%0' 21742 & 33150. Debreu. G, nl'ory of lalt"":\n a,\lOmatic analysis of cconomic c4uilibrium, N, Y., IlJ5lJ.Dobb, M. H. All Essllr 011 Ecollolllic (;rol\,(h lIlUI Plallllillg. Ldn,!\<. Y .. IlJ60. Domar, E. D. Capital Expansion. Rate ofGrowth and Employplent. :n ECll//lII/1I'(rica !-i, IlJ46. no, 2:137 47. [d. t"sSIII'S ill (hI' Thl'/lrI' /If EClill/llllic GI"I'I\III. N. Y.,Il/S 7. Dorfman, ·R., et al. Lil/I'a~ Progralll/1/illK lIlUI EC[lIIOIlli<AlIlIhsis, N,Y.,Toronto, Ldn,1l/5X hihl.C (jddschiip/lmg11. li'irrschafi.lhrl'i.I/l/I/( Munchen, 1l/.17. Grlilllg. F. /)1'/'

Wil'!.Hht!fis-Kreis!lIuf [And] Da.l· Wirrsc/wlf.l-i\!odd/. Mun­chen, il/33.- Harrod, R. F. An Essay in Dynamic Thcory.In EC/lI/. J. 49. 1l/3lJ: 14 33. [d. T[I\I'arlls a Drlll/lIlic £C[lI/O/1/­in. Ldn. N. Y" 194X. Harvard University. Harvard Econom­ic Research Project. Swdies ill (he S(rUClUre of (he Al/lt'I'icallI:'co 11 [I1/1.1'. Theoretical and empirical explorations in input­output analysis. By W. W. Leontiefet oIl. N. Y"llJ53. Hicks,J. R, A Colllrihwi[lll (0 (he TIII'[lI")' o!lhe Tradl' CI'ele. Oxford.Il/SO. Horowitz. D., cd. M;lrx lIl1d ."'[lder;t EWl/oll1ics.Ldn. 1%8. Johannsen, N. A. L. J (j, J, O. Lahn, pseud.) Dl'I'/\rl'i,l!al/r des GI'Itlcs 1/. i\!echlll/isllll/S dcs .';O::illl-Lchl'lI.l, Bin,1903. Kaldor, N, Altcrnative Theories of Distribution. InR. Ecul/ Sll/d. 23. 195f,: X3 100. [d. A Model of Economic(jrowth. [n Ecoll. J. 67, 1957, no. 26/\: 5lJ1624. Kalccki.M. The Determinants of Distribution of the NatIOnal '[n<.:ome,In EC/lllo/1/efrica 6, 193/\, no. 2: lJ7 -112. Id, Essay.l· ill Ih"Th"[lr\, of EcollolI/ic FII/Cll/llliolls. Ldn. 1939. 1'01.: P/'{jhll (I'/lriik/lllil/;tkll/ry. Warszawa, 1933. Keynes, J. M. The (;elll'l"lI!Theorr /If Employmell!, IlI!e/'eSIl/l/d A1/1l/ey. Ldn. 1936, Kor­nai. l., and Liptak, T. Kctszintli tervezcs. In Magyar Tudo­m~IIlYos Akadcmia Matematikai Kutat6 Intczet. A lI!agyurTudol/uil/Yos Akadhllia A!all'lIIa(ikai KWa((j 1//(cZI'I(;//ck k(i::II'­III('/lll'i. VO\. 7. Budapest, 1962: 577-621. [Planning on TwoLevels.] Id. Marhemarical Plal/I/i/lg 01' S(ruclllral Deeisio//s.Amsterdam, Budapest. 1967. Hung.: A KlEdllsligi ,1::erke::cll1la(cl/1iI1ikai le!'Vezcsc, Budapest. 1965. [d .• and Liptak, T.Two-Level Planning. [n £col/oml'lrica 33, 1965: 14169.K relle, W. Volksll'ir(.I'chafilichc Cl'slIflllr<'Ch//III1K ei//sch/iess!ichillpw-o/lfpw-A/lalyse mi( Zahlelljiir die BRD. 2. Aufl. Bin, 1967.1st ed. 1959. Lange, 0., and Taylor. F. M, O/l (hc Eco//omicTheory of Socialism: Minncapolis, Minn.,1938. Ledercr. E.Der Zirkulationsprozess als zentralcs Prohlem del' okonomi·schen Theorie. [n Archil' Soz.- Wi.l'.I'. So::.-Po/i(. N. F. 56, 1926:I 25. -- Lenin, V.1. On the So-Called Market Question. [nLellill CW, VOI.I.: 75125. Leontief. W. W. TlleS(rll('/l/reor Ihe Al11aicall Eco/lOII/Y, 19/9 1<)3<). An cmrmical applica­tion of equilibrium analysis. 2nd cd. Cambridge, Mass., 1951.1st cd. 1941. [d. Die Wirtschaft OIls K reislauf. [n Archil'So::.-Wi.l's.Soz.-Po!i( N.F.60, 1l/28: 577623. Lexis, W.Altgemei//I' Vl/lkslI'irf.l'chafi.l/ehre, 3. Aufl. Lpz., 1926. 1st cd.1910. Li.)we, A. Wie ist die KonJunkturtheorie uberhauptmiiglich'.' In Wellwirlsch. Archil' 24, 1926: 165 97. L.uxem­bu~g. R. Thl' Acculllullllil//l (It' Capilal. N. Y., 1964. I st cd.New Haven. Conn., 1951. Ger.: Die AkkulI/1I1ario// des Ka{Ji­(als. Bln,1913. -- Morishima. M. Economic Expansion andthe Interest Rate in Generalized von Neuman Models. InECI///ollle(rica 28, 1960, 352-63. - MUlier. K.O. W. Dicbiirga/ichc Kreislaujihcoric. Bin [OJ, 1968. Neisser. H. Del'Kreislauf des Geldes. [n Welrwir(.I'ch. Archil' 33, 193 I : 365 40X.--- Nemchinov, V, S. Eko/lomiko-IIw(cnw(ichl'skic 11/1'(I/d\· iIIIl/dl'li. 2.izd. M., 1965. 1st cd. 1962. [Economico-Mathemati­cal Methods and Models.] - [d., ed. Prillle/le/lie II/a(emll(ikiI' cko/lomichl'.I'kikh isslct1lmmiiakh. 1'.2 & 3. M., 1961-65.[The Application of Mathematics in Economic Resear<.:h,]Pasinetli, L. L. Rate of Profit and Income Distribution inRelation to the Rate of Economic Growth. In R. ECOII, Swd.1<), 1961.62: 26779. Peter. H. Ala(helllalischc S(rllkwr­lehrc des Wir(schafiskreislal/fes. Gottingen, 1954. - [d. S(ruk­(IIr(cllre de" Wirl;cha[r.l'kreislallrcs. BIn, 1943. --- Phelps, E. S,The Golden Rule of Accumulation; a Fable for Growthmen,[n All/cr. EculI. R. 51, 1961: 638-43. Preiser. E. WachswlII11. FiI/!,olllll/('fI.Hcrleilll//g. 2nd ed. Heidelberg, 1964. I st cd.I 'i(, I, ()uesnay, F. Analyse du Tableau economique. In

J. aKric//11 "Oll/lIIl'1'U' 11I/lllIcel 17 ()() . .Iuin. [d. /'lhl"IIII(;('o/loll/i'llle a ...·c .1011 e.lpliea(io//, .1. cd, \ ersaillcs. 1759, Ist cd175X. Fng. Ill: Mlrabeau. \. de Riquettl. 1-"1"111I(0/.1 (jlle.\I/II1ilte Fcol/olllielll Tah!e ! Ta/>I"ol/ I'COII'I/I/i'/'II' j N. Y., (llhS.

Rohinsnn. J. The .-lecl/l/lI/laliol/ 0/ Cllpi(ol. 1I,'mev\oo,1. Ill ..Ldn.1956. [d .. 111 1:".1.1'1.1' 011 .\f"rli"lI Fec.llomin, Ldn. 1942.2nd cd. 1900. S.ll11l1elslln, 1'.,.\, (nter<tdions hetv.ecn theMultiplier Analysis and the l'rlllciplc "I' A,:celcr'ltion, InR. bOI/. S,ali.lf. ~/. 1931J. no 2: 75 7~ Schneider. F. I-:ill­liihrl/lIg ill die 11 'irrsc/IlI/(Sfh,.orr,.. T. I, TI"'orie d,..1 Il'il'!sch"/Is­",n,l/all/s, 13, :\ull. Tllbmgen.1967, 1st ed. 1947. Sdllllll­petcr. .1.:\, Th,. Th"ol"l ot" Lcol/ol/lic J)('\"/0I'II/('I/(. 111/ !l1l1//inilllo P/'(Ifi(s, Capital, Cr"dil, 111("1',.11. alld Ih,' HI/,Iilless ('I cl,..('amhndge. Mass .. 1l/.~4. (ier l'Il,'ori .. "('I' "'/rfsclllll(!ich"1IEII(lI'ick!/I/IK' Fine lInterslldlUng libel' l 'llIcrnehl1lergewinn,Kapltal. Kredlt. Zins Lt. den Klllllllnktllllykllls. 1.1'/ .. 191~.

Simonde de Sismondi, J. C. l.. .\"0111'''''1/.1 I',il/cip"s d',;ecll/oll/iepolili,//I<'. 3,cd. Ciencve, 1951. 1st cd. Paris. IXI9. Sllllth.:\./111 !l1l/l/ir\" illlo (h" .\'''1/11'1' alld C'/ilI,'S of Ihe Wl'lII(h "I .\'lllio/lI,~ vol. Rci'll'. N. Y .. 1966, Ist cd. 1776. Slllov., R, A ('"ntn­bution to the Theory of Economic Growth. In QI/all, 1. L"'I/.70, 1950, no. I: 65 94. Stohhe. ,\. l·lIler.I·/lch,lllgell ::/11'lI/akr(}(ikl/1I0mischell Thcorie tiel Eillko/IIIII"II.I'I'('l'!eilrlll.!:, Tll­bingen, 196~. Swec/Y, 1'. M, Th,' Th,.o/'l' 01 ('apiflllil{D('\'"Iopme/ll: PIillcipln of :\fur.lioll Polr(i( 01 Lco/l0 111,1" , 4thprint. N. Y"llJ56. 1st cd, 1942. Tugan-Baranovskii ITllgan­Baranowsky), M.1. TheIJreliscllt' Gl'lllId(uKell dn ,\fal'Xi.IIIII/I.Lpz .. 1905. Tustin. A. The Afedwl/i,I/1I o/F'cl//lomic S\'SI,'IIII.An approa<.:h to the problem of e<.:onOI11I<.: stabih/atlon fromthe point of view of control-system engllleering. 2nd cd. I.dn,1957. 1st cd, 1953. von BortkiewiL'l, L. Wertrechnllng 1I.

Preisrechnung im Marxschen System, In Archil' So::.-II'iS'.S·oz,-PIJ!i(. 23, 1906: I 50; 25, 1907: 10 51 & 445 HH, [d.Zur Berichtigung del' grundlegenden theorelis<.:hen Konstrllk­tion von Marx im dritlen Band des "Kapital". In .Ihr .\"a(.­Gkl/II. Sta(i,I/. 8<), 1907: 319 35. von Weizsiicker, (·.C·

Wac!lslllll/, Zills 1/. ofJlimalc IIlI'"s(i(il/llsql/olc. Basel. Tublllgen,1962, Wal'fenschmidt, W. (j. AlI.lc/ltlllliche I:'illfii!ll'llllg ill dieIII1KCII/cille u. (hel/rc(ischc Na(iollaltikol/I//l/ie, Meisenheim'Glan.Il/50. Id. Wirfsch,(fl.I/l/ed/(/llik. StUtlgart,Ii)57.Wagncr, A.H.G. Filllll/::"'is,1 ,,11.1 cho/I. Bd I. 3. Allll. l.p)'.,I !I/\3. Walras, L. £'11'/1/1'11I.1 of Pur" LCIJllollliCl; or Tlreo/Tof'Socia! Wellllh. Repr. N. Y.,I96lJ. 1st cd. Ldn, 1954. h'::tlhllell(S (ren!llomie po/iliqlle 1'/11''' 0/1 I/":orie de la richessesl/ciale. 2 t. Lausanne,l/\74 77. Ward. H. The Sociali,\(ECIIIWIIIY; II SII/dy olOrKwli::aliollal AI(ert/lI(i!'e.l'. N. Y.,1967.

Wolfstetter, E. Verteilungstheoric, Lohnpolitik u. gewerk­schaftI i<.:he Strategic. [n f!eidelherKer Hltif(('I' 5. 196/\, Nr 12 '13:33-60.

HaIlS Gotl(ried NUl::ingcr