Anwar Shaikh - An Introduction to the History of

Download Anwar Shaikh - An Introduction to the History of

Post on 22-Jan-2016




0 download


Crisis Theories


  • An Introduction to the History of Crisis TheoriesANWAR SHAIKH


    U.S. Capitalism in Crisis, U.R.P.E., New York.

    This paper is about the history of crisistheories. Broadly speaking, the term "crisis" asused here refers to a generalized set of failures inthe economic and political relations of capitalistreproduction. In particular, the crises we seek toexamine are those towards which the system isinternally driven, by its own principles of opera-tion. As we shall see, it is in the nature of capitalistproduction to be constantly exposed to a varietyof internally and externally generated disturbancesand dislocations. But only at certain times do these"shocks" set off general crises. When the system ishealthy, it rapidly revives from all sorts of set-backs; when it is unhealthy, practically anythingcan trigger its collapse. What we seek to examineis different explanations of how and why thesystem periodically becomes unhealthy.

    I Reproduction and Crisis

    Consider how peculiar capitalist society is. Itis a complex, interdependent social network,whose reproduction requires a precise pattern ofcomplementarity among different productiveactivities: and yet these activities are undertakenby hundreds of thousands of individual capitalists

    who are only concerned with their private greedfor profit. It is a. class structure, in which thecontinued existence of the capitalist class requiresthe continued existence of the working class: andyet no blood lines, no tradition, no religiousprinciple announces who is to rule and who is tobe ruled. It is a cooperative human community,and yet it ceaselessly pits each against the other:capitalist against worker, but also capitalistagainst capitalist and worker against worker.

    The truly difficult question about such asociety is not why it ever breaks down, but why itcontinues to function. In this regard, it isimportant to realize that any explanation of howcapitalism reproduces itself is at the same time(implicitly or explicitly) an answer to the questionof how and why non-reproduction occurs, andvice versa: in other words, the analysis of repro-duction and the analysis of crisis are inseperable.This is true whether or not a particular theorymakes this connection explicit.

    In the history of economic thought, we candistinguish three basic lines of analysis aboutcapitalist reproduction. First, and most popular, isthe notion that capitalism is capable of automaticself-reproduction. It may be smooth and efficient(neoclassical theory), or it may be erratic andwasteful (Keynes), but it is self-equilibrating.


  • Above all, there are no necessary limits to thecapitalist system or to its historical existence: if leftto itself (neoclassical theory) or if properlymanaged (Keynes) it can last forever. Naturally,this has always been the dominant conception inbourgeois theory.

    The second position takes the opposite tack:here, it is argued that/ by itself, the capitalist sys-tem is incapable of self-expansion. It must grow tosurvive, but it requires some external source ofdemand (like the non-capitalist world) in order tokeep it growing. This means that its reproductionis ultimately regulated by factors outside of thesystem: the limits to the system are external to it.The different schools of underconsumption, in-cluding Marxist ones, have their origin in this lineof thought.

    Lastly, there is the position that, thoughcapitalism is capable of self-expansion, the ac-cumulation process deepens the internal contra-dictions on which it is based, until they erupt in acrisis: the limits to capitalism are internal to it.This line is almost exclusively Marxist, andincludes both "falling rate of profit" and "profitsqueeze" explanations of crisis.

    Each of the above positions implies a corres-ponding notion of crisis, why they occur and whatthey imply. We will therefore examine each inturn.

    II Capitalism as AutomaticallySelf-Reproducing

    In what follows we will discuss the laissez-faire and Keynesian traditions of orthodox theoryin separate sections.

    A. The Laissez-Faire Tradition

    Unfortunately we are all too familiar with thenotion of capitalism as a self-regulating, smooth,efficient and harmonious system. From its begin-ning in Adam Smith's "Invisible Hand" to theimpotent elegance of modern general equilibriumanalysis this one conception has dominatedbourgeois theory. The fundamental contradictionof all human existence is said to arise from theinsatiability of human wants in the face of thelimited availability of physical resources.1 Theinsatiable greed of capitalism is thus transformedinto an attribute of Human Nature; its insaneplunder of our planet is therefore only "natural,"the inevitable outcome of a battle within Natureitself. Human Nature meets Physical Nature. Inthis way greed, competition and selfishness areeternal: there is nothing we can do about them, no

    way we can eliminate them. In fact, on this basiscapitalism is presented as that social set of ruleswhich automatically permits the freest expressionto the above "intrinsic" human drives. Moreover,since it represents the optimal institutional solu-tion to an eternal "natural" conflict, capitalismremains eternally optimal. It has no limits otherthan some unimaginable mutation in HumanNature or some unimaginable destruction inPhysical Nature. Leave it alone, and capitalismwill reproduce itself smoothly, efficiently andprobably forever. So the story goes.

    Since the system is viewed as self-regulating,the process of regulation tends to be ignored.Thus, the dominant tendency within this proble-matic is to concentrate on either static or balancedgrowth equilibria. In this way the impression isgiven that the adjustment process itself is negligi-ble. Indeed, this strategy is quite necessary, sincethe notion of a prolonged adjustment process is athreat to the concept of equilibrium and hence tothe cherished optimality of the system.

    Even so, crises occur anyway. This tends tomake economists resentful, at times quite surly.Nonetheless, their ideological function requirethem to (periodically, at least) deal with thequestion of crises.

    Economists who study the history of empiri-cal phenomena are inevitably impressed not onlywith the frequency of crises but also with theirapparent regularity. In the U.S., for exampleWesley Clair Mitchell counts fifteen "crises"within the 110 years from 1810 to 1920, while PaulSamuelson lists seven "recessions" in the thirtyyears from 1945 to 1975.2 In between was theGreat Depression which lasted almost ten years!

    There are basically only two ways to absorbthis evidence into the main body of theory withoutpermanently damaging it. First and foremost, itcan be argued that in principle crises need neveroccur; that they do in fact occur may then beattributed to factors which are external to thenormal functioning of capitalist reproduction.Through no fault of its own, the system isperiodically disrupted by crises. In this traditionwe find crisis blamed either on Nature (sunspots,crop failures in general, etc.) and/or on HumanNature (psychological cycles of optimism anddespair, wars, revolutions, and political blur/- ,ders).3

    But the regularity of crises proves hard to pinon sunspots or consumer bio-rhythms, while theone-shot explanations like wars and politicalblunders are just not adequate to explain appar-ently cyclical phenomena. Consequently we get


  • the concept of the business cycle; it represents theother basic way to absorb the phenomena of crisesinto orthodox theory. Within this concept, thesystem is still viewed as being self-regulating: onlynow the adjustment process is seen as beingcyclical rather than smooth. Various factorsinternal to the system's operation give rise to self-generating cycles, so that self-reproduction has aninternal rhythm.

    It is important to note that in orthodox theorya cycle is not a crisis. In order to be consistent withthe overall theoretical structure, cycles must beviewed as being essentially "small fluctuations,"second order variations which at first approxima-tion one may justifiably neglect. In this way thecyclical nature of the adjustment process does not-represent a limit to the ability of the system toreproduce itself.

    The branch of orthodox economics known asbusiness cycle theory is a combination of thesetwo basic approaches. Regular, non-violent fluc-tuations are internal to the system: contractionsand expansions are part of the normal businesscycle. Violent or prolonged expansions andcontractions, however, arise from external factorsoriginating in Nature and Human Nature, factorswhich either turn a cycle into a crisis, orprecipitate one entirely on their own. Crises, there-fore, remain outside the normal process ofcapitalist reproduction.

    In spite of its yeoman service, business cycletheory has always occupied a minor role in laissez-faire economics. Its subject matter was toodangerous, its history too tainted by anti-capitalistsentiments, for it to be comfortably integrated intothe main body of theory. With the advent ofKeynesian economics, however, this changed. Weshall see why shortly.

    B The (Right) Keynesian Tradition

    We have so far been speaking of the "laissez-faire" tradition within bourgeois theory, since thishas almost always been the dominant one. But themassive worldwide collapse of capitalism duringthe Great Depression dealt this tradition a stagger-ing blow. The collapse itself was "easily" explainedby the faithful in a variety of ways similar to thosedescribed above: what was inexplicable was thefact that the system did not seem to exhibit anytendency to snap back to "normal" full employ-ment equilibrium. Even by official (conservative).estimates unemployment in the U.S. hoveredaround ten million people in 1939 a full tenyears after the "Great Crash."

    As the Depression dragged on, as socialunrest deepened, laissez-faire theory fell increas-ingly into disrepute and Keynesian theory rapidlytook its place.

    Keynes attacked the orthodox notion that"supply determined its own demand," for it wasthis notion which led to the conclusion thatcapitalism tended automatically to, more or less,fully utilize the available labor force and means ofproduction. Instead, in his analysis the level ofinvestment spending planned by capitalists is thecrucial factor in determining the level of outputand employment. But investment plans depend toa significant extent on the anticipation of profits,on the "expectations" and "animal spirits" ofcapitalists. Two major conclusions follow fromthis. First, since "expectations" are notoriouslyvolatile, capitalist reproduction is likely to bequite erratic. Second and even more important,there exists no automatic mechanism withincapitalism which would make capitalists plan justthe right amount of investment so as to assure fullemployment. It should be noted, however, thatthe system is presumed to be automatically self-equilibrating: it is just that the equilibrium doesnot preclude persistent unemployment or infla-tion.

    The so-called Keynesian Revolution was anambivalent one, however. Much of the "deep"structure of Keynes' analysis was the same as thatof the orthodoxy he attacked:4 the division ofsociety into producers and consumers (not classes)the same basic view of human nature, the crucialimportance of psychological "propensities" andpreferences, the role of supply and demand, andabove all the general reliance on equilibriumanalysis. It is no wonder then that a portion of theorthodoxy was able to absorb Keynes into a newversion of bourgeois theory. Conceding that therewas indeed no automatic mechanism to makecapitalist reproduction smooth, efficient and crisisfree, the neoclassical Keynesians (Bastard Keynes-ians, as Joan Robinson calls them) turned to theState as the mechanism which would bring to lifethe society pictured in the laissez-faire parables. Ifthe State did its job well, it would manipulateaggregate demand so as to maintain near fullemployment with little or no inflation; with thismodification, "the rest of the doctrines of the(orthodoxy) could be revived."5

    Since economic fluctuations are an admis-sable part of Keynesian theory, business cycletheory becomes a much less dangerous branch ofeconomics. Indeed, since the State in principle caneliminate fluctuations, it becomes imperative to


  • study cycles and crises in detail in order to knowhow to counter them. Consequently, a greatwealth of information about crises has sprung upsince the so-called Keynesian Revolution.

    Not surprisingly, Keynesians tend to see theerratic and violent history of capitalist accumula-tion as a series of errors in "policy."' Their viewson the current crisis are no exception.

    Keynes also generated another branch offollowers, the so-called left Keynesians, amongwhom the leading figure is Joan Robinson. Herviews, along with those of Michael Kalecki andJoseph Steindl, will be discussed in the nextsection.

    Ill Capitalism as Incapable of Self-Expansion

    From the very beginning, the laissez-fairevision of a harmonious, crisis-free capitalism hasbeen bedeviled by an equally old and equallypersistent notion of capitalism as being inherentlyincapable of accumulation. The internal forces ofthe system, it is argued, can at best reproduce it atsome stationary level: but as stagnant capitalismsoon degenerates. Competition sets each againstthe other, yet because there is no growth no onecan gain except at the expense of someone else.Capital is set against capital, worker againstworker, and class against class. Either theantagonisms become too intense and the systemexplodes, or else it decays into a society (likeChina of old) in which a tiny ruling elite rests on abase of mass poverty and human misery. In eithercase, a non-accumulating capitalism will not lastlong.

    Interestingly enough, this countervailing ar-gument begins from the same initial conception asthe theory it attacks. Orthodox theory has alwaysinsisted that the ultimate goal of all capitalistproduction is to provide for consumption: thatwhich is not consumed now is plowed back intoproduction in order to provide for future con-sumption. Either way, it is consumption whichrules the roost. In the dark glass of underconsump-tion theory, this same notion becomes a weapon inthe attack on capitalism. Throughout the long andcomplex history of this branch of crisis theory, thefollowing argument appears again and again: yes,the ultimate regulator of all production is indeedconsumption, currently or in the future; however,capitalist production responds not to need but topurchasing power, not to demand but to "effec-tive" demand (i.e. demand backed by money).And such is its contradictory nature, that if left toitself it is incapable of generating sufficient

    effective demand to support accumulation. Theintrinsic mechanisms of the system, in otherwords, tend to point it towards a stationary state:it requires some external source of effectivedemand external, that is, to its basic mecha-nisms in order to continue growing.

    A. The Concept of the Demand Gap

    Over the past 150 years, there have beenmany attempts to specify the exact nature of theunderconsumption problem. In spite of the varietyof formulations, however, it is quite striking howconstant is the notion that the demand forconsumer goods is the ultimate regulator ofoverall production.

    Suppose we divide all social production intotwo major branches or "Departments." Depart-ment I produces producer goods (raw materials,fuel, plant and equipment, etc.), while Depart-ment II produces consumer goods and services(food, clothing, entertainment, etc.).

    The basic tenet of underconsumption theorythen, is that the demand for consumer goods andservices determines not only the production levelof Department II (consumer goods), but also thatof Department I (producer goods). Output in theproducer goods industry is ultimately regulated bythe input requirements of the consumer goodsindustry: the demand for producer goods is there-fore "derived" from the demand for consumergoods.

    Notice that this does not merely say that theoutput of Department II influences the output ofDepartment I, and vice versa. It says somethingmuch stronger, namely that the causation isprimarily one way, that Department II is theleader and Department I is the follower.

    Parallel to this notion is a conception ofcirculation as a process whereby society's productis shared out between workers and capitalists.Thus, out of the total social product, part of it isconceived of as being the replacement of the inputsused up in producing it, and the remaining part,the net product, is thought of as being availablefor "distribution" among workers and capitalists.

    A similar breakdown is made on the incomeside. Out of the sales of all firms, an amount ofmoney is said to be set aside to replace moneyexpended for producer goods used up duringproduction. The rest is the net operating income ofthe firms which is divided into wages and profits.This net income, what orthodox economists callnet national income, is the source of effectivedemand for the net product.


  • Net production has two,sides, therefore. Onthe one hand we have goods and services and onthe other we have net money income, whichequals wages plus profits: supply on one side, andeffective demand on the other.

    We can now state the basic problem of under-consumption theory. Workers generally spend alltheir wages. They therefore "buy back" a portionof the net product, at its normal price. But sinceworkers never receive the whole of net income,they can never buy back the whole of net product.Workers' consumption always leaves a "demandgap;" moreover, the lower the share of theirwages, the greater the "demand gap."

    At this stage in the analysis the surplusproduct still remains to be sold, and capitalistincome profit still remains to be expended. Ifthese two could match up, all of the productwould be sold and the "demand gap" completelyfilled. But under what conditions will this happen?

    The early underconsumptionists tended tovisualize the net product as being composed solelyof consumer goods. Given their fundamentalpremise that the output of Department I isregulated by the input requirements of Depart-ment II, they easily fell into,-the-idea that in anyperiod of time the output of Department I is justsufficient to replace the inputs used up by thesystem as a whole. This means that although thetotal social product is made up of both producergoods (Department I) and consumer goods (De-partment II), the net product (the total minus thereplacement requirements) consists solely of con-sumer goods. *

    From this point of view, after workers spendtheir wages to "buy back their share" of the netproduct, we are left on the one hand with a surplusproduct in the form of consumer goods, and onthe other hand with the unexpended profits whichform capitalist "income." It therefore follows thatthe "demand gap" will be filled only if capitalistsspend all their profits on personal consumption.But then there can be no investment, hence nogrowth, no internally generated accumulation.

    This does not mean that capitalists will nottry to accumulate. What it in fact implies is thatthe attempts of the class as a whole to accumulatewill be self-defeating. Ater all, in the cut-throatcompetition of one capitalist against another, the

    * The net product is that part of the total product overand above that necessary to maintain the productivesystem. If we subtract workers' consumption from it, weget that part of the total product over and above the main-tenance requirements of the productive system and of theworkers who operate it: it is the surplus product.

    size of a capitalist's assets are an important indexof power. And one important way to increase insize and power is to save, invest and therebygrow. So capitalists will keep trying to accumu-late. Imagine, therefore, that we begin from theinitial situation pictured above, in which Depart-ment I produces just enough producer goods tomaintain the productive capacity of the system,and Department II produces an amount ofconsumer goods which are completely "boughtback" by workers and capitalists consuming alltheir income.' Now suppose that the next timearound capitalists spend only part of their profitson consumer goods; the rest they invest by buyingproducer goods, hiring workers, and setting upfirms in Department I and/or Department II.

    A curious thing happens at this point. Let'ssay that total profits amount to $200,000, whichthe capitalist class at first spends entirely onpersonal consumption. Now suppose they cutback their consumption to $150,000, and theremaining $50,000 they invest by using $30,000 tobuy producer goods (from the inventories ofDepartment I) and $20,000 to hire workers (out ofthe reserve army of the unemployed). The netdrop in consumer demand is only $30,000, sincethe drop in capitalist consumption demand ispartially offset by the extra consumption of thenewly hired workers. Nevertheless, demand forconsumer goods does drop, so that sales inDepartment II will fall which in turn means that itsown demand for producer goods will fall, thusdecreasing sales in Department I. Yet, the very actwhich led to all of this has simultaneouslyexpanded productive capacity in general. Theirattempt to expand capacity has therefore maderedundant not only the extra capacity they haveadded but also a part of the capacity which existedbefore. Inevitably this must cause them toretrench. Internally generated accumulation ne-gates itself.

    Since expansion occurs gradually and taketime to complete, one can imagine that it takes awhile for the lack of "effective demand" to makeits effects felt, and another while for the contrac-tion which ensues to work itself out. The conse-quence of the attempted accumulation wouldtherefore be a boom followed by a bust, with zeronet accumulation over the cycle. This, accordingto the logic of underconsumption theory, wouldbe the expected behavior of a capitalist economyleft to itself.

    Cycles of boom and bust are no strangers tothe history of capitalism. At the same time,however, the study of history makes it abundantly


  • clear that these cycles are accompanied bytremendous secular growth in actual capitalisteconomies a fact which stands in sharp contrastto the intrinsically stagnant capitalism implied byunderconsumptionist logic. Invariably, therefore,underconsumption theories have had to resort to"exogenous" (i.e. external) factors to explain thisgreat contrast between history and theory. In thenext two sections, which deal with the history ofpre-Marx and post-Marx underconsumption theo-ries, respectively, we will see what an importantposition these external elements occupy.

    B. Conservative and Radical UnderconsumptionTheories

    In the preceding section I have attempted topresent both the essential logic behind undercon-sumption arguments and the implications whichfollow from this logic. In doing so I have usedmodern conceptual tools such as Marx's twoDepartments and Kalecki's aggregate supply anddemand analysis. But these concepts are relativelynew, and quite naturally the argument does notappear in precisely this form in the actual historyof underconsumption theory. In fact, what is quitestriking about this history is that while the notionof a "demand gap" appears throughout, thecorresponding implication about the impossibilityof self-sustained capitalist accumulation is seldomg rasp ed . Par t i cu la r ly amon g th e n o n-Marxian theories, this implication is assiduouslyavoided. It is a difficult position indeed to live andwrite in the 19th century, during a period of almostexplosive capitalistic growth, and have yourtheory tell you that growth is not intrinsic tocapitalist production.

    Convinced of the soundness of their basicposition but either unaware of or unwilling toaccept its full implication, the early undercon-sumptionists almost universally adopted the posi-tion that too much accumulation would cause acrisis. They would begin by assuming that theeconomy was growing at some "sustainable" rate.Following the logic I have outlined in thepreceding section, they would then assume thatcapitalists cut back this consumption and investedthe amount so saved in additional producer goodsand workers. Thus, while investment had ex-panded productive capacity, the net cut back inconsumer goods demand and its subsequent effecton producer goods demand resulted in underutili-zation of even the capacity which had existedbefore. "Too much saving" had led to a slump.*

    But what their logic actually implied was that

    any saving would lead to a slump, a fact that wassoon pointed out by their opponents. In his excel-lent study entitled Underconsumption Theories,Michael Bleaney summarized the dilemma of theearly underconsumptionists:

    The general position of these writers was thatthere is a limit above which the rate ofaccumulation becomes dangerously high,threatening to precipitate a slump. But thelogic of the argument as they develop it is thatthis limit is in fact a zero rate of accumula-tion, as is effectively pointed out by Chal-mers. Thus they are caught in a trap, in whicheither they must draw back from the brinkand discard part of their results, or they mustopenly state the absurdity of their conclu-sions.7

    The first major economist to land himself inthis dilemma was Thomas Malthus (1820s). Trueto the underconsumptionist tradition, Malthusargued that it is the demand for consumer goodswhich regulates production, so that only a certainrate of growth was "sustainable." Of course, giventhe logic of his argument and the conclusionimplicit in it, Malthus was never able to say justwhat this "sustainable" rate of growth was.Nevertheless, he did emphasize that (too much)saving would mean that capitalist consumptionwould not fill the demand gap left by workers, sothat crises of overproduction (underconsumption)were distinctly possible in capitalism. In Malthus'hands this tendency towards underconsumptionbecame a reactionary apologetic for feudal land-owners, whose high living and conspicuousconsumption was presented as a welcome counter-balance to the tendency of capitalists to (over)save. (Malthus is also famed for his attack on theworking class through his so-called laws ofpopulation. Then, as now, these brutish "naturallaws" were never meant to represent the behaviorof the "civilized" ruling classes.)

    Simonde de Sismondi was a contemporary ofMalthus who also saw a tendency toward under-consumption within capitalism. Once again, wefind here the argument that the level of consump-tion regulates overall production, so that produc-tion can grow only as fast as consumption grows.But capitalism restricts the consumption of themasses by keeping them in poverty; the workersare too poor to buy back their own product (here

    * The underconsumptionists did not envision anyKeynesian discrepancy between planned saving andplanned investment. Capitalists plan both, and what theysave is invested, not hoarded. Hoarding does not play amajor role in underconsumption theories, as Bleaney (op.cit., pp. 50-51) points out.


  • again the ubiquitous demand gap). Moreover, ascapitalism develops, income distribution becomesmore and more unequal, so that consumption ofthe masses grows more slowly than overall wealth(the gap widens). In Sismondi, therefore, not onlydoes a tendency to underconsumption exist, but ifalso gets worse as capitalism matures. Over timecrises get worse, and competition among nationsfor external markets gets more fierce.

    Unlike the reactionary Parson Malthus, Sis-mondi was a radical who was deeply impressed bythe suffering of the peasants and workers undercapitalism. In his time he stood at the head of whatMarx called petty-bourgeois socialism, whichstruggled against the cruelty and destructionengendered by capitalism and sought to reform itso as to ameliorate these conditions. Sismondihimself championed radical changes in incomedistribution in favor of peasants and workers, andlooked to the state to carry these and othereconomic reforms out.8

    Both Malthusian and Sismondian schools ofunderconsumption mention external markets assources of consumption demand. In Malthus thisis only a passing reference; in Sismondi, however,foreign markets are an important outlet fordomestic overproduction, and he sees increasinginternational rivalry as arising out of the worsen-ing underconsumption problem. Of course, inorder for international trade to be a solution tothis problem, a given nation must export more toothers than it imports from them. This isobviously impossible for the world as a whole. Ifall trade is confined to capitalist spheres only,then foreign trade is internal to the world capital-ist system and offers no escape from the under-consumption problem. Sismondi consequentlydoes not present foreign trade as a general solutionto the problem.

    Between the time of Sismondi (1850s) and thetime of J.A. Hobson (1900s) came the great water-shed in capitalist history which marks thebeginning of the Age of Imperialism. In the yearsbetween the 1870s and 1914, for instance,European foreign investments rose over 700%,much of it going to the so-called Third World. It istherefore not at all surprising that by the 1900sforeign trade, through imperialism, began toappear to be a solution to the problem of under-consumption. After all, if one conceives of theworld in terms of the imperialist capitalist nationsand the underdeveloped Third World, it becomespossible to also imaging this Third World absorb-ing the excess savings of the developed capitalistcountries either directly in the form of foreign

    investments, or indirectly in the form of commodi-ty exports. Both in Hobson and in Rosa Luxem-burg (whom I will discuss in the next section), theconnection between underconsumption and im-perialism becomes very important.

    Hobson begins in the now familiar way ofunderconsumptionists. He explicitly identifies theultimate object of all production, even undercapitalism, as being the production of consumergoods. Moreover, he is the first one to explicitlytreat Department I (the producer goods industry)as being strictly subordinate to Department II(consumer goods), so that the whole productionprocess may be treated as a vertically-integratedsystem beginning from raw materials and proceed-ing in successive stages to the final product whichconsists of consumer goods alone. Lastly, he toobegins by postulating a "sustainable" rate ofgrowth (which of course he cannot define) andthen goes on to show that (too much) saving leadsto a slump. Crises arise from (over) saving.'

    Hobson also introduces the concept of the"surplus," which plays an important role in hissubsequent analysis. Generally speaking, the"surplus" is defined by Hobson to be the excess ofthe total money value of the output over thestrictly necessary costs of producing that output.10

    This concept involves distinguishing betweennecessary and unnecessary cpsts of production, aswell as between costs of producing and otherexpenses (such as selling costs, sales taxes, etc.). Itis a broader concept than what I defined earlier asprofits (sales minus all costs), but we need notpursue the difference here.

    In any case, Hobson's notion of surplusincludes unnecessary "costs" such as monopolyprofits and rent of land (since these do not stemfrom production of any sort). As capitalismdevelops, these "unearned incomes" swell, andsince their recipients tend to consume little,oversaving tends to occur. There is, therefore, aworsening problem of underconsumption.11

    According to Hobson, foreign trade providesan outlet for excess savings and a market forexcess production, even under competitive capital-ism. However, as industry becomes more concen-trated and monopoly becomes widespread, theunderconsumption problem moves to a qualita-tively higher level. On the one hand monopolyprofits swell the surplus, leading to greatersavings; on the other hand since monopoliesachieve these excessive profits by raising prices,they tend to shrink the market. The same factorswhich expand savings thus reduce the outlets forthem. Imperialism arises as the solution: imperial-


  • ism is the highest stage of underconsumption.However, this need not be so, says Hobson.

    The root cause of crises and imperialism lies in theinequality of income and the excessive incomes ofmonopolists and rentiers, and the solution lies inappropriate reforms:

    Let any turn in the tide of politico-economicforces divert from these owners their excess ofincome and make it flow, either to theworkers in higher wages, or to the communityin taxes, so that it will be spent instead ofbeing saved, serving in either of these ways toswell the tide of consumption there will beno need to fight for foreign markets or foreignareas of investment.12

    A surprising number of theses advanced byHobson in the 1900s reappear in subsequentMarxist analysis. Writing in 1916, Lenin empha-sizes the connection between monopoly andimperialism, though he rejects Hobson's under-consumption analysis. On the other hand, in the1920s, the German revolutionary Rosa Luxemburgargues that the roots of imperialism in fact lie inthe underconsumption problem, though she ofcourse rejects the conclusions which Hobsondraws from this. More recently, in the UnitedStates, influential works by Marxists Paul Sweezyand Paul Baran have revived Hobsonian notionssuch as the view of overall production as avertically integrated sector, the concept of the"surplus," the notion that monopoly tends tomake the surplus rise, and above all the argumentthat the absorption of the surplus represents anintrinsic problem of capitalist production whichbecomes more acute with the prevalence ofmonopoly. We turn to these theories next.

    C. Marxian Underconsumption and Dispropor-tionality Theories

    In the early underconsumption theories, theproblem is invarial D -f?d in terms of too great arate of accumulation. We have seen that accordingto their own logic, however, any accumulationtended to negate itself. Inevitably, underconsump-tionists were driven to the conclusion thatcapitalism tended' towards stagnation, that aself-expanding capitalism was impossible.

    Marx completely destroyed this argument. Inorder to see why, we need to discuss some of theconceptual advances made by him.

    We are already familiar with the first greatadvance, which was to conceptualize overallproduction in terms of two major branches orDepartments, producer goods (I) and consumergoods (II). This means that the total product over

    any period of time is composed of both types ofgoods.

    The second breakthrough by Marx was toclarify the nature of effective demand. The under-consumptionists, it will be recalled, identifiedbasically three types of effective demand: replace-ment demand which buys back producer goods toreplace those used up, workers consumptiondemand which buys back their "share" of theproduct, and capitalists' consumption and netinvestment demand which must fill the "demandgap" in net output.

    Marx's first point of departure involves aquestion of time. Suppose the production processin each Department takes a given length of time,say one year. Well then, the producer goods usedup in the overall process cannot be bought out ofthis year's production, because the first 'inishedproducer good which derive from productionstarted in this year won't roll off the assembly lineuntil the end of the year. Similarly, workersemployed during this year cannot "buy back" theconsumer goods resulting from their currentactivities because these goods won't be ready tillthe end of the year; nor can capitalists consumewhat is not yet available.

    Let's go back to the beginning of the year. Tokeep the example as simple as possible, assumethat all goods to be used over the year are boughtat the beginning of the year (this is an expositionaldevice only). Capitalists decide the level ofproduction they would like for the current year.They therefore buy a certain amount of producergoods, and hire a certain number of workers; theworkers in turn use their wages to buy consumergoods. At the same time, capitalists also must buya certain amount of consumer goods for their ownpersonal consumption over the year. Notice thatthe effective demand originates entirely with thecapitalist class: workers' wages are part of theyear's gross investment expenditures by capitalists.It is quite illegitimate to treat consumption andinvestment as being functionally independent,since the bulk of consumption comes from wages,which are themselves a necessary aspect of investment expenditures. '

    At the beginning of the year, therefore, it isthe capitalist class through its consumption andinvestment expenditures which determines effec-tive demand. But who sells the commodities?Why, the capitalist class, of course! The beginningof this year is also the end of the last year; it istherefore also the time when the finished productof last year's production process becomes avail-able. Last year's production provides the capitalist


  • class with the commodity-supply available for saleduring this year; this year's expenditures by thecapitalist class on gross investment and personalconsumption determine the effective demand forthat commodity supply. If this sounds bizarre, itshould be remembered that capitalist reproductionis bizarre. Production and consumption decisionsare undertaken by hundreds of thousands ofindividual capitalists with no thought whatsoeverfor the reproduction of the system as a whole.Though it is the capitalist class which determinesboth ends of the supply-demand relationship,capitalists do not do so as a class but rather asindividuals. The difficult part is to explain whythey ever manage to "come out right." We willreturn to this point shortly.

    It is not difficult to go on from here to showthat steady growth is easily possible, with effectivedemand in each year being just sufficient to buythe available supply at "normal" prices.13 If invest-ment grows by 10%, then output grows by 10%.If therefore capitalist consumption also grows by10%, each year's output will find waiting for it theeffective demand to buy it. After Marx, thepossibility of "balanced growth" has become acommonplace.

    Balanced growth implies that productivecapacity and effective demand can grow atroughly the same rate. Taken by itself, however, itdoes not necessarily imply that capitalism achievesanything remotely like that. Nor does it tell usanything about which way the causation mightrun if such growth was indeed possible on theaverage. Nonetheless, the fact that expandingreproduction is possible poses a distinct threat tounderconsumption theories. It is in the light of thischallenge that we encounter Marxist versions ofunderconsumption theory.

    A little background on Marx's writing is inorder here. During the period 1858-1865, Marxwrote and rewrote the bulk of the manuscriptsfrom which his great three volume work, Capital,is taken. Volume I was published by 1867, butVolume II in which the analysis of the capitalistreproduction process appears was never putinto final form, even though it was revised in theearly 1870s and again in the late 1870s. Marx didnot live to complete this task, and the latter twovolumes were compiled and published by Engels.^During Marx's lifetime, therefore, the publishedparts of Marx's work did not deal with reproduc-tion and growth.14

    In Volume I Marx demonstrates that a surplusproduct can arise only if workers as a whole workmore hours in a given day than it takes for them to

    produce the goods they themselves consume andgoods needed to replace those used up in theproduction process. It is this surplus labor time ofworkers over and above that necessary for them tomaintain themselves and the productive system,which provides the surplus product appropriatedby the capitalist class.

    In Czarist Russia, this struck a responsivenote. Capitalism had begun its destruction ofsocial forms, in particular the ancient peasantcommune, the mir. In the 1850s, it was beingargued by some populists that the mir could serveas the basis for a direct transition to socialism,without having to go through the horrors ofcapitalist industrialization. By 1880, Volume I ofCapital had provided Marxist populists not onlywith a devastating critique of capitalism in generalbut also by means of a little extrapolation with an important theoretical weapon againstcapitalism in Russia.15

    The Marxist populists saw Marx's emphasison surplus labor-time as proof of the impossibilityof capitalism in Russia. In classic underconsump-tionist fashion, they reasoned that since workersproduced more than they consumed, the homemarket would never be sufficient to permitgrowth. The developed Western capitalist nationshad escaped this dilemma by finding foreignmarkets; but Russia, they argued, was toounderdeveloped to compete effectively on theworld market. Capitalism therefore was not viablein Russia. Organizing the peasants was the key tosocialism.

    Volume II of Capital was published in 1885,two years after Marx's death. Even so, fifteenyears later the Marxist populists were still insistingthat "it is impossible for a capitalist country toexist without foreign markets."16 But by now acounter-argument had developed within RussianMarxism; and it counted some heavy names on itsside: Bulgakov, Tugan-Baranowsky, Struve, Le-nin.

    This latter group of Marxists made two majorcriticisms of the populist underconsumption argu-ment. First, they noted that it was a fact thatcapitalists and commodity relations were rapidlygrowing everywhere in Russia. Lenin's first bookThe Development of Capitalism in Russia (1899)was aimed at making just that point. Secondly,Lenin and the others attacked the logical basis ofthe populist argument. The basic error, they said,lay in imagining that even under capitalism,consumption was the goal of production. Capital-ism produced for profit, not consumption, andMarx's analysis of expanded reproduction estab-


  • lished beyond a doubt that this profit-motivatedproduction was entirely capable of generating itsown internal markets. Underconsumption was notan intrinsic problem. Capitalism was alreadythere, it was viable and spreading, and organizingthe urban proletariat was an urgent task.

    That round of the debate was decisively wonby Struve, Bulgakov, Tugan-Baranowsky andLenin. But their victory only set the stage foranother, even more important series of questions:if capitalism was indeed capable of self-sustainedgrowth, what is there to prevent it from growingforever? What are its limits, in other words?Moreover, how are we to understand the devastat-ing crises it is periodically subject to?

    Tugan-Baranowsky's response was to arguethe extreme position that capitalism was totallyindependent of consumption, provided Depart-ments I and II grew in the correct proportions toone another. But, he argued, given the anarchy ofcapitalist production, this correct proportionalitywas a matter of chance. The trial-by-error natureof capitalist production would therefore period-ically give rise to such great imbalances thatreproduction would be interrupted and a crisisbreak out. Lenin rejected Tugan-Baranowsky'sassertion that consumption was irrelevant, but atthis time other than emphasizing the anarchy ofcapitalist production as a source of crises, he didnot provide a clear cut crisis theory. He was not toreturn to this subject. In Germany, some ten yearslater the disproportionality theory of crisescropped up again, this time in Rudolph Hilfer-ding's massive work on monopoly capitalism.Both Tugan-Baranowsky and Hilferding werelater to argue that since it was the anarchy ofcapitalism which led to crises, planning wouldeliminate crises. "Organized capitalism," in Hilfer-ding's words, was the solution, and the parliamen-tary path to State control was the means.17

    Rosa Luxemburg refused to accept thisresolution of the debate. As a revolutionaryactivist, she was completely opposed to thereformism which the disproportionality theoryseemed to engender. Once one admits "thatcapitalist development does not move in thedirection of its own ruin," she declared, "thensocialism ceases to be objectively necessary." Toabandon the theory of capitalist collapse was toabandon scientific socialism. And so she set out torevive the Marxist underconsumption debate.18

    Since it was Marx's examples of expandingreproduction (balanced growth) which proved tobe the decisive factor in the earlier debate amongRussian Marxists, Luxemburg attacked these

    examples directly. Marx plainly demonstrated theabstract possibility of expanded reproduction, sheconceded , but he did not seem to realize that itwas nonetheless impossible in reality because,from a social point of view, the capitalist behaviorit requires makes no sense. Imagine that at theend of a production cycle the whole social productis deposited in a warehouse. At this pointcapitalists come forward and withdraw a portionof the total product to replace their producergoods used up in the last cycle, and workers comeand withdraw their means of consumption. Thisleaves the surplus product, from which capitalistswithdraw a portion for their personal consump-tion. Now Luxemburg asks, where do the buyersfor'the rest of the product come from? (This is ofcourse the traditional underconsumption problemof filling the "demand gap"). If Marx is right, shesays, then it is the capitalist class which buys backthe rest of the product in order to invest it and thusexpand productive capacity. But that makes nosense at all, for "who are the new consumers forwhose sake production is ever more to beenlarged?" Even if capitalists did what Marx saysthey will, in the next period productive capacitywill be even greater, the gap to be filled evenlarger, and the problem even more intractable.Marx's "diagram of accumulation does not solvethe question of who is to benefit in the end byenlarged reproduction. . ." Expanded reproduc-tion is algebraically possible but socially impos-sible.20

    It follows that actual capitalist accumulationcan be explained only through some forcesexternal to "pure" capitalist relations. Luxemburgnotes that the Malthusian solution of a third classof unproductive consumers makes no sense, sincetheir revenue could only come from profits orwages. Similarly, foreign trade among capitalistnations also provides no solution for capitalism asa whole, since it is internal to the world system.She therefore argues that capitalist accumulationrequires a strata of buyers outside of capitalistsociety who continually buy more from it thanthey sell to it. Thus trade between capitalist andnon-capitalist spheres is a prime necessity for thehistorical existence of capitalism, and imperialismnecessarily arises as capitalist nations struggleover control of these all important sources ofeffective demand. Moreover, as capitalism ex-pands to cover the globe the non-capitalist milieushrinks correspondingly, and with it shrinks theprime source of accumulation. The tendency tocrises is heightened, and competition amongcapitalist nations for the remaining non-capitalist


  • areas intensifies. World crises, wars and revolu-tions are the inevitable outcomes of this process.

    Even if Luxemburg were right about theimpossibility of accumulation, her solution wouldnot work since it requires the "Third World" tocontinually buy more than it sells. Where wouldthe excess revenue come from?

    But in fact she is also wrong about thepossibility of accumulation. To see this we need toreturn briefly to the analysis presented at thebeginning of this section. Recall that at the end ofthe production cycle, it is the capitalists who are inpossession of the whole social product. At thesame time, it is also their gross investment andpersonal consumption expenditures which are theoriginal source of effective demand for this veryproduct (since workers' wages are a part of overallinvestment). Now, aside from their own personalconsumption, their remaining expenditure (grossinvestment) is in no way motivated by consump-tion as such. It is motivated entirely by the antici-pation of profit. What Marx's examples show isthat if capitalists did undertake the appropriateamount of investment, then they would indeed beable to sell their product and make the anticipatedprofits. If this success spurs them to reinvest onceagain in anticipation of yet more profits, theywould be rewarded once again, and so on. All thewhile consumption would expand due to thegrowing employment of workers and the growingwealth of capitalists. But this expansion ofconsumption would be a consequence, not acause.*

    Yet if this refutes Luxemburg's criticisms ofexpanded reproduction, it still does not answer thetwo crucial questions she began with. First, whatforces, if any, make expanded reproductionpossible in reality? And second, is it not true thatif expanded reproduction is actually possible,"capitalist development does not move in thedirection of its own ruin?"

    That which theory debates, reality decides. In1929 a devastating worldwide capitalist crisiserupted, to be followed by over ten years of deepdepression and unemployment. Given this back-ground, the problems of capitalist reproductiononce more rapidly rose to prominence.

    The first major attempt to revive undercon-sumption theory as an explanation of crises was

    * Readers familiar with Volume I of Capital mightrecall that Marx distinguishes two types of circuits involv-ing purchase and sale: C-M-C and M-C-M'. In the formerthe object is consumption, but in the latter the object is theexpansion ot capital. It is the latter which is the dominant(regulating) circuit of capitalist production. Luxemburgforgets this.

    made by Paul Sweezy, in his influential book TheTheory of Capitalist Development (1942). Sweezyexplicitly set out to formulate an underconsump-tion theory "free of the objections which havebeen levelled at earlier versions."21

    In this early attempt Sweezy is still very muchin the grip of the traditional underconsumptionnotion that the demand for consumer goodsregulates overall production. From this point ofview Department I appears as part of the verticallyintegrated productive apparatus of Department IIso that changes in the output of Department I(producer goods) are in effect changes in thecapacity to produce consumer goods. In addition,Sweezy argues that "empirical evidence" suggeststhat a 1% change in Department I output willincrease capacity output of consumer goods by1 %. This is a virtual replay of Hobson, whom weanalyzed earlier.

    Now consider effective demand, which as wehave seen is composed of capitalist consumptionand total investment expenditures (the latter inturn being composed of expenditures on producergoods and on hiring workers). As capitalismdevelops, Sweezy notes, mechanization proceedsapace and it takes more and more machines andmaterials to back up one worker; this means thatcapitalist investment expenditures on producergoods rise faster than those on wages. Given hisanalysis of production, the investment expendi-tures on producer goods imply proportionalincreases in consumer goods capacity, whereas themore slowly rising expenditures on wages ofcourse translate into workers' consumption. Itappears, therefore, that the capacity to produceconsumer goods expands faster than the consump-tion demand of workers. A "demand gap" thusopens up. Of course, capitalist consumptiondemand might fill the gap. But as capitalismdevelops capitalists tend to invest proportionatelymore, and consume proportionately less, of theirprofits, so that their consumption lags behind theproductive capacity of Department II. Sweezyconcludes:

    . . .it follows that there is an inherent tenden-cy for the growth of consumption to fallbehind the growth in the output of consump-tion goods . . . this tendency may expressitself either in crises or in stagnation, orboth."The fundamental error in Sweezy's analysis is

    the traditional underconsumptionist one of reduc-ing Department I to the role of an "input" intoDepartment II. Once this assumption is made, itnecessarily follows that an increase in production


  • of producer goods must expand the capacity ofconsumer goods. But this is false: producer goodsmay be used to make producer goods also, and aswe noted in the critique of Luxemburg, expandedreproduction requires that they be so used.Contrary to Sweezy's reasoning, it is perfectlypossible to have a rising ratio of machines andmaterials per worker and a proportional growth inthe outputs of both Departments, while stillhaving expanded reproduction.

    Sweezy's second attempt, made along withPaul Baran, came over twenty years later inMonopoly Capital. In the first attempt, as we haveseen, Sweezy argued that capitalism had anintrinsic tendency to expand the productioncapacity of Department II faster than consumptiondemand. Monopoly Capital, written in the light ofMarx, Keynes and Kalecki, no longer restrictsitself to Department II or consumer demand alone.Instead, it is argued here that modern capitalismhas a tendency to expand total productivecapacity faster than internally generated effectivedemand so that in the absence of externalfactors, "monopoly capitalism would sink deeperand deeper into a bog of chronic depression."23

    It follows from this diagnosis that "the fairlylong periods during which the (actual) accumula-tion process has proceeded in a vigorous fashionwith . . . the demand for labor power expandingrapidly and productive capacity being utilized ator close to full capacity" must be explainedthrough external factors,24 Thus Baran and Sweezypoint to major innovations (steam engine, rail-roads, automobile), imperialist expansion andwars, and the stimulation of demand in generalthrough advertising, government policy, etc., asbeing crucial factors in overcoming the inherentlystagnant nature of monopoly capitalism.

    The association of monopoly with slowgrowth and excess capacity is not new. Manytheories (as we shall see) attempt to explain thiscorrelation. Baran and Sweezy's specific contribu-tion is their argument that these phenomena arisefrom the persistent tendency of monopoly capital-ism to ouer-expand productive capacity and thusdrive itself towards crises and/or stagnation. Wemust therefore seek out the logical basis of thisargument.

    Recall that in Marx's analysis it is totalinvestment and capitalist consumption expendi-tures which determine effective demand (totalinvestment includes expenditures on wages, whichin turn determines workers' consumption). More-over, insofar as the personal consumption of thecapitalist class responds more or less passively to

    past and present profits, it is total investmentwhich is in fact the crucial variable.

    Now suppose that at the beginning of a givenyear, total investment expenditures for next year'sproduction are large enough to expand productivecapacity but not large enough to buy all of theexisting social product. Then capitalists will on theone hand have initiated an expansion of theirfuture productive capacity, while on the otherthey will find demand insufficient for even theirpresent capacity.Given the anarchic nature of capitalist pro-duction, such an outcome is to be expected fairlyoften. The question is, is this merely one aspect ofthe regular fluctuations in capitalist reproduction,or is it something more? Marx, for instance,argued that capitalists are driven to accumulate asrapidly as objectively possible, so that a discrep-ancy such as the above tends to be self-correcting.But if one could somehow argue that in eachperiod investment tends to remain in the rangedescribed above large enough to expandcapacity but not large enough to purchase thepreceding period's supply then of courseproductive capacity will outrun effective demandand the system will be faced with a demand gap or"realization problem." This is precisely the argu-ment implicit in Baran and Sweezy's assertion thatthe (potential) surplus expands faster than thesystem's ability to.absorb it. Yet, though they tendto lay much of the blame for this problem onmonopoly, they do not discuss why monopolistswould persist in over-expanding productive capac-ity in the face of insufficient demand. The crucialelement of their whole thesis therefore remainsunexplained. In his recent survey of Marxist crisistheories, Erik Olin Wright notes this all importantdeficiency:

    The most serious weakness in (this) undercon-sumptionist position is that it lacks anytheory of the determinants of the actual rateof accumulation. . . Much underconsump-tionist writing has, at least implicitly, optedfor Keynes' solution to this problem by focus-ing on the subjective anticipation of profit onthe part of capitalists as the key determinantof the rate of accumulation. From a Marxistpoint of view this is an inadequate solution.I have not yet seen an elaborated theory ofinvestment and rate of accumulation by aMarxist underconsumptionist theorist, andthus for the time being the theory remainsincomplete.25

    In their book, Baran and Sweezy cite contri-butions made by Joan Robinson, Michael Kalecki,and Joseph Steindl. Since these authors are also an

    is i


  • integral part of the left-Keynesian theoreticaltradition, it behooves us to investigate the implica-tions of their respective analysis for the questionof crises.

    Investment plays a crucial role in bothKeynesian and Marxian analysis. But in Keynesiantheory the emphasis is very much on the short-rundeterminants of investment decisions. Insofar asthe above authors treat investment decisions,therefore, they tend to focus primarily on theshort-run and only secondarily on long-runstructural changes. Joan Robinson's early workonly treats structural change in passing, whereasher later works rely mainly on Kalecki.26 Kaleckiin turn, when he briefly deals with the long-run,simply assumes that in the absence of externalfactors capitalism tends toward stagnation. It isinnovation, therefore, which is the major factor inpushing investment above the level necessary tojust reproduce the system, and he argues that it isthe decline in the intensity of innovations inmonopoly capitalism which accounts for itsrecent slow growth.27 This is all very ad hoc,though, and in his last major work (1968) Kaleckiemphasizes that a satisfactory explanation oflong-run determinants of investment was stilllacking.28

    Lastly, Steindl begins by notion the incom-pleteness of Kalecki's long-run analysis, and setsout to remedy this defect. In the final analysis,however, he too is forced to postulate a decline inthe intensity of innovation as the primary factor inthe slow growth of modern capitalism, though heemphasizes that monopoly tends to exacerbate theeffects of this decline. Like Kalecki before him, hetoo ends by declaring that a satisfactory explana-tion has yet to be found.2' It is not surprising,therefore, that Baran and Sweezy prefer to set outtheir own versions of the problem.

    IV Capitalism as Self-Limiting Accumulation

    Radical and Marxian underconsumption the-ories tend to focus on effective demand as thelimiting factor in capitalist accumulation. InMarx's own analysis, however, effective demandis not an intrinsic problem. On the contrary, in hisview capitalists are driven to accumulate asrapidly as possible, so that self-expanding repro-duction, not stagnation, is the normal tendency ofthe system. This does not imply that theaccumulation process is smooth, or that partialcrises may not occur along the way due to cropfailures, etc. But it definitely does imply that thelimits to the accumulation process do not arise

    from an insufficiency of demand.Does this mean, as Rosa Luxemburg so

    eloquently argues, that once one rejects undercon-sumption theory one is forced to accept the viewthat accumulation (and hence capitalism itself) iscapable of indefinite extension? Not at all.According to Marx, the limits to accumulation areentirely internal to the process. "The real barrierof capitalist production is capital itself."30

    Capitalist accumulation is motivated byprofitability. But, according to Marx, accumula-tion progressively lessens profitability, so that ittends to undermine itself. This is the famous lawof the tendency of the rate of profit to fall, whichwe shall turn to shortly. At the same time,accumulation implies extension of capitalist rela-tions, increase of the proletariat and of itsstrength.

    Declining profitability means declining ratesof accumulation and increasingly fierce competi-tion among (national and international) capitalistsfor markets, materials and cheap labor-power. Asweaker capitals are eliminated, economic concen-tration and centralization (i.e., "monopoly") in-creases. Moreover, it becomes increasingly neces-sary for capitalists to attack wages either directly,through mechanization, or through import ofcheap labor-power and/or export of capital topoorer countries.

    At the same time, the size of the working classand the extent of its collective experience instruggling against capital is continually on the rise.Thus capital's increasing attack on labor is metwith an increasing resistance and counter-attack(over the long-run). The class struggle intensifies.

    It is important to realize that the tendency forprofitability to decline (as Marx derives it) is notcaused by high wages although rising real wagesmay well exacerbate it. This means that theperiodic crises which result from declining profita-bility cannot be attributed to labor's demands orresistance, though of course different historicalstages and political situations are very importantin explaining how the system as a whole reacts toeach crisis. As long as capitalist relations prevail,however, its general tendencies will continue tooperate. Consequently Marx emphasizes that thetask of the proletariat is not only to resist capitalbut to overthrow it.

    It should be obvious from this brief sketchthat rising "monopoly," declining rates of accumu-lation and deepening class struggles can beexplained as consequences of the basic laws ofcapitalist development, rather than as factorsgiving rise to new laws as is attempted by Baran


  • and Sweezy, for instance.* Since the law of declin-ing profitability is central to this explanation, wemust examine it further.

    A. Marx's Theory of the Falling Rate of Profit

    The question of profitability has two impor-tant aspects. First, what is the basis of profitabilityand what determines its extent? Second, how doescapitalism develop this basis and what effect doesthis in turn have on its extent?

    In answer to the first question, Marx beginswith the labor process. In all societies, he notes,the objects necessary to satisfy human needs andwants imply a certain allocation of society's labor-time, of its productive activities, in specificproportions and quantities. Otherwise the repro-duction of the society is impossible.

    While the allocation of social labor isfundamental to all societies, the extraction ofsurplus labor is the basis of all class societies. Thissurplus labor forms the material and social basis ofthe class relation. The extraction of surplus labormust be enforced, for it provides the ruling classnot only with its means of consumption, but alsowith its means of domination.

    In most societies, the allocation of sociallabor-time and the extraction of surplus labor aresocially regulated, by tradition, by law, by force.But in capitalist society, productive activity isprivately undertaken by individual capitalists onthe basis of potential profit. Reproduction is notan explicit consideration, and yet it must and doestake place. On the surface, it is money prices andprofits which provide the day to day "feedback"which determines capitalists' decisions. But, Marxargues, in reality it is the total labor times (laborvalues) involved in the production of commoditieswhich regulate the money phenomena. Thisregulation of prices and profits by labor valuesand surplus value is in fact the manner in whichthe social requirements of reproduction manifestthemselves in capitalist society. We will hence-forth deal directly with labor values and surplusvalue, since these are the real regulating elements.

    During the labor process, workers use instru-ments of labor (plant and equipment) to transformmaterials into finished products. The total labor-time required for the finished product is therefore

    * Incidentally, it is worth noting that when, as aconsequence of declining profitability, capitalists curtailtheir investment expenditures, part of the product availablewill not be sold and it will appear that the crisis is caused bylack of effective demand, by "underconsumption." But infact this "underconsumption" is only a reaction to the crisisin profitability. It is a symptom, not a cause.

    composed of two parts: first, the labor-timeimplicit in the means of production (materials,plant and equipment) used up; and second, thecurrent labor time expended by workers in thelabor process itself. Marx calls the first element"constant capital" (C) since it reappears in the finalproduct, while he calls the second "value added byliving labor" (L). The total labor value of any finalproduct is therefore C + L.

    Out of the final product, part is just theequivalent of means of production used up. Itslabor value will therefore be C, since this is thelabor value of the actual means of production usedup. This leaves us with the net product on onehand, and value added by living labor (L) on theother. The net product is the material equivalentof living labor time L.

    If there is to be a surplus product, then onlypart of the net product must go to replace theconsumer goods used up by workers. The valueadded by living labor (L) is therefore composed oftwo parts, one of which corresponds to the labor-value of the workers' consumption requirements(V) and the other to the labor value of the surplusproduct (S). In other words, it is the differencebetween the time workers actually put in (L), andthe time necessary to reproduce themselves (V) their surplus labor time (S) which gives rise tothe surplus product and hence to real profits: S =L-V.

    The division of living labor time intonecessary (V) and surplus labor time (S) is therforethe hidden basis of capitalist society. Marx callsthe ratio S/V "the rate of surplus value" or "therate of exploitation." Other things being equal, thegreater the rate of exploitation the greater theamount of surplus value and hence the greater theprofit.

    The time workers actually put in (L) isdetermined by the length of the working day. Thetime necessary to reproduce themselves (V), on theother hand, is determined by both the amount ofgoods they consume (their "real wage") and thelabor-time it takes to produce these goods. Themass of surplus value (S) and the rate of exploita-tion (S/V) can therefore be increased in two ways:directly, by lengthening the working day L so thatsurplus labor time is directly increased; andindirectly, by lowering the necessary labor-time Vso that more of a given working day is spent insurplus labor-time. This latter method of increas-ing S and S/V requires that either workers' realwages be reduced or that the productivity of theirlabor be raised so that it takes them less time toproduce their means of consumption, or both.


  • Capitalists constantly try all methods ofincreasing the rate of exploitation. But over timethe growing strength of the working class hassharply restricted attempts to lengthen the work-ing day and/or lower the real wage. Thusincreasing the productivity of labor has come to bethe principal means of raising the rate of exploita-tion. But the paradoxical thing about capitalism,according to Marx, is that the very means bywhich it raises the rate of exploitation tends tolower the rate of profit. The rising productivity oflabor manifests itself in a falling profitability ofcapital.31

    The rate of surplus value S/V expresses thedivision of the working day into necessary andsurplus labor-time. It measures the degree ofexploitation of productive workers. But to capital-ists the crucial thing is the degree of profitability ofcapital. From their point of view, they investmoney in means of production (C) and in workers(V), with the intention of making profit (S). Theamount of profit (S) relative to their investment (C+ V) is the capitalist measure of success. In otherwords, it is the rate of profit S/(C+V) whichregulates the accumulation of capital.

    This is where the paradox comes in. In theircontinuing battles against one another,* individ-ual capitals are constantly forced to lower unitcosts so as to gain an edge over their competitors(the current battle over pocket calculators is anexcellent example of this process). As far assuccess in battle of sales is concerned, anythingwhich lowers unit costs will do.

    But capitalists are also perpetually engaged inanother battle the battle of production, in thelabor process. And it is here that mechanizationarises as the principle means of raising theproductivity of labor and hence lowering unitcosts. Capitalists hire workers for a specifiedperiod, and their aim is to squeeze the maximumpossible productivity out of them during thelabor-process, at the lowest possible cost. Thisimplies not only struggles over the real wage andthe length and intensity of the working day, butalso over the nature of the labor process itself.From the very beginning capitalists have sought to"perfect" the labor process by subdividing it intoincreasingly specialized and routinized tasks. With

    * These battles are what Marx calls the "competitionof capitals." But this use of the term competition is not thesame as in "perfect competition," the opposite of which is"monopoly." In Marx the progressive concentration andcentralization of capitals implies fiercer "competition ofcapitals" over progressively greater parts of the world. Theso-called "monopoly" stage of capitalism does not super-cede competition but rather intensifies it.

    capitalist control of the labor process humanproductive activity is made increasingly mechani-cal, automatic. It is no surprise then that thesemechanized human functions are progressivelyreplaced by actual machines. As machines replacesome human functions, the others are even moresubject to the tyranny of the mechanical, untilsome of these functions too are replaced bymachines, and so on.*

    The tendency towards mechanization is there-fore the dominant capitalist method of raising thesocial productivity of labor. It arises out ofcapitalist control of the labor process, of humanproductive activity. As such, neither growingworker resistance nor rising real wages are theintrinsic causes of mechanization, though theymay well speed up this tendency.

    Increasing mechanization gives rise to whatMarx calls a rising technical composition ofcapital. Ever greater masses of means of produc-tion and materials are set into operation by a givennumber of workers. According to Marx, this inturn implies that out of the total labor value (C +L) of the final product, progressively more comesfrom the means of production used up andprogressively less from living labor. In otherwords, the rising technical composition is reflectedin value terms as a rising ratio of "dead to livinglabor," of C to L.

    The rate of profit, as we have seen, isS/(C+V). But S = L-V, since surplus labor-time(S) equals the time workers actually put in (L)minus the time necessary to reproduce themselves(V). Therefore, even if "workers lived on air"(V=0), the most that S could be is smax/C=L/C.Consequently, L/C is the ceiling to the rate of pro-fit, while the floor is of course zero. Now, if arising technical composition does indeed reflectitself as a rising ratio C/L hence a falling ratioL/C then the actual rate of profit will be pro-gressively squeezed between a descending ceilingand an unyielding floor, so that it must itselfexhibit a downward tendency. This is what Marxmeans by the tendency of the rate of profit to fall.

    The falling tendency described above is inde-pendent of how L is divided between V and S, andhence independent of the rate of exploitation S/V.In fact, if the real wage of workers were constant,the rising productivity of labor due to mechaniza-tion would continually raise S/V; the greater theproductivity of labor, the less time it takesworkers to produce a given bundle of consumer

    * For a brilliant analysis of the modern labor process,see Harry Braverman's Labor and Monopoly Capital,Monthly Review Press, New York, 1974.


  • goods so that a greater portion of a given workingday becomes surplus labor-time. Even when realwages do rise, as long as they rise less rapidly thanproductivity, the rate of exploitation will still rise.It is perfectly possible, therefore, to have both arising real wage and a rising rate of exploitation."This is in fact the general situation pictured byMarx, on the grounds that workers can nevercapture all the productivity gains of mechaniza-tion without bringing accumulation to a halt andthus killing the golden goose.* For Marx the classstruggle over real wages operates within certainobjective limits, the limits given by the accumula-tion of capital. These limits are intrinsic tocapitalism itself, and can be transcended only byoverthrowing it.

    Almost all Marxist commentators accept as afact that mechanization is an overwhelming realityof capitalist production. However, one importantschool of thought attributes mechanization not tothe capitalist control of the labor process, as doesMarx, but rather to capital's reaction to growingworker resistance and/or rising real wages (in thelong-run). Typically, they begin by postulating arise in real wages under given conditions ofproduction, which leads to a fall in the rate ofprofit, which in turn induces capitalists to substi-tute machines for workers. From this point ofview, of course, mechanization and its attendantrise in the productivity of labor are the principalmeans of increasing profitability, while risingwages tend to diminish it. Depending on whichfactor prevails they say, the rate of profit can goeither way. * * Paul Sweezy and Maurice Dobb, forinstance, both hold this point of view.33

    This analysis is correct as far as it goes.Rising real wages will indeed induce mechaniza-tion, and this may or may not offset the effect ofthe higher wages on profitability. But in Marx therising real wages are themselves made possible bya prior cause, namely the mechanization arisingfrom the battle of production. Thus the effect thatSweezy and Dobb analyze is a secondary one,superimposed on (and indeed only possible

    * " This is precisely the point Marx makes in VolumeI of Capital, in the first part of the chapter entitled "TheGeneral Law of Capitalist Accumulation" (Ch. XXV,Section I), when he notes that real wages can rise only ifthey do "not interfere with the progress of accumulation"(p. 619).

    ** "For a more detailed discussion of this position,as well as some of the mathematics (such as the so-called"choice of technique" theorems) used to support it, see"Political Economy and Capitalism: Notes on Dobb'sTheory of Crises," by this author, forthcoming in theCambridge Journal of Economics.

    because of) the primary one. Given that theyignore the primary cause, it is not surprising thatthey can find no particular reason for the rate ofprofit to fall.

    Another major objection to the law arguesthat mechanization (whatever its cause) does notnecessarily imply a falling tendency to the rate ofprofit. Consider a given number of workers, sothat L is given. Mechanization means that themass of means of production employed by theseworkers increases. But this is also accompanied bya rise in the productivity of labor and hence a fallin the labor value of commodities, since it nowtakes less time to produce a given commodity.Therefore the labor value of the means ofproduction (C) will not rise as fast as their massand may even fall. Marx argues that nonetheless Cwill rise, so that C/L will rise and the fallingtendency will operate. But, say the critics, supposethe labor value of the means of production fall asfast or even faster than its mass rises? Then C/Lwill stay constant or even fall, and no downwardpressure will be exerted on the rate of profit.

    It must be said at the outset that this objectionis a valid one, since it points to a gap in the fallingrate of profit argument. As constructed in thecurrent literature, there is a strong presumptionthat a rising ratio of machines to workers alsoimplies a rising ratio of "dead" to living labor(i.e. of C to L). But attempts to specify the exactconnection between the two (such as Yaffe's)34

    have not been satisfactory, so that the possibilityof the scenario pictured by the critics alwaysremains open. This issue is still very much thesubject of debate, and is treated at greater lengthin the forthcoming article referred to in thepreceding starred footnote.

    Another currently popular objection has todo with the notion that capitalists would neverchoose to employ a technique of production whichlowers their rate of profit. A falling rate of profit istherefore automatically excluded. This argumentis often stated mathematically, as in the so-called"Okishio Theorem,"35 but its basic presupposi-tions underlie a widely shared analytical frame-work ranging from left Keynesians such as JoanRobinson to Marxists such as Bob Rowthorn. Interms of the above discussion, the crucial errorhere lies in the presupposition that technicalprogress is merely a question of capitalist "choice"and not one of necessity. Marx noted long ago thatunder capitalism it is the necessity of competitionthat forces capitalists to choose the technique witha lower unit cost, even when it implies a lower rateof profit. Whoever makes this move first will


  • undersell the others. The only "choice" theremaining capitalists will then face is betweenmaking some profits at a lower rate than beforeand making no profits at all because their productcosts too much.36

    Lastly, some Marxists reject the notion of arising C/L on empirical grounds. Since C is thelabor value of the means of production, and L isthe value added by living labor, their moneyequivalents are K, the money value of the meansof production, and Y, the money value added or"net national product." On this basis the "capital-output ratio" K/Y is examined, and since officialstatistics indicate that it tends to be constant overlong periods, this is said to militate against thenotion of a rising C/L.37

    It is interesting that these same Marxistsstrenuously oppose accepting at face value theofficial statistics on unemployment, the extent ofpoverty, the incidence of malnutrition, etc. onthe grounds that bourgeois conceptions of thesecategories so dominate their construction as tomake them practically useless. Unemploymentstatistics, for instance, do not count those whohave given up looking for work, those who neversucceeded in finding jobs in the first place (such asblack teenagers), and those who do not enter thework force because of the hopelessness of it (suchas housewives). It is not uncommon, therefore, forradicals and Marxists to estimate "real unemploy-ment" to be two to three times the official figure.And yet when it comes to absolutely fundamentalcategories such as "capital" and "value added,"official statistics are suddenly accepted withoutquestion. We will return to this important point inthe discussion of "profit squeeze" theories ofcrises. For the moment it is sufficient to note thatthe one Marxist statistician who has bothered toexamine how these statistics are gathered, and tocorrect them for the conceptual differences be-tween Marxist categories and orthodox ones, hasfound precisely that the "capital-output" ratioseems to rise steadily.38

    B. A History of Falling Rate of Profit Theory

    The tendency of the rate of profit to fall ascapitalism develops was widely accepted byClassical economists as an incontrovertible fact.The problem lay in explaining this phenomenon.

    Adam Smith (1770s) for instance, noted thatwhen more capitals crowd into a particularindustry, they expand supply, drive down prices,and hence lower profits. In the same way, heargued, as accumulation progresses capital as a

    whole will become more plentiful and this willdepress the rate of profit.

    Critics quickly pointed out that capitalscrowd into a particular industry only when theindustry has a rate of profit above the average;moreover, by so doing they merely drive its rate ofprofit back down towards the average. Theaverage rate consequently remains unexplained,and there is no reason given in Smith whyaccumulation should alter it in any way.

    Some forty years later, David Ricardo (1810s)offered an alternate explanation. As the societydevelops, he argued, more land has to be broughtinto cultivation to feed the growing population.This means bringing progressively less fertile landunder cultivation, so that food becomes increas-ingly expensive to produce. In Marxian terms, thelabor value of food rises. For a given working day,therefore, necessary labor-time rises and surpluslabor-time falls correspondingly. Thus the rate ofsurplus value falls as society develops, and with itfalls the rate of profit not because worker's realwages rise, but because the productivity ofagricultural labor falls.

    The crucial conclusion in Ricardo is that theproductivity of agriculture tends to decline. In hiscritique of the Ricardian theory of rent, Marxdemonstrates that this conclusion is neitherlogically nor empirically true. Indeed, capitalisthistory is characterized throughout by a risingproductivity of labor, both in industry and inagriculture. As we have seen in the precedingsection, Marx's own explanation of the falling rateof profit is based on a rising productivity of sociallabor and a rising rate of surplus value.

    The rate of profit falls, not because labourbecomes less productive, but because it be-comes more productive. Not because theworker is less exploited, but because he ismore exploited. . ,39

    Marx considered his own explanation of the"tendency of the rate of profit to fall as [capitalist]society progresses" to be "one of the greatesttriumphs over the great stumbling block of allprevious economics." It is the lynchpin of hisanalysis of laws of motion of the capitalist system.And yet, curiously enough this law plays arelatively minor role in much of the history ofMarxist thought. It is completely absent fromunderconsumption theories, for instance, and aswe shall see in the next section, it is similarlyabsent from "profit-squeeze" theories.

    Part of the reason for this neglect stems fromthe previously examined objections to the logic ofMarx's derivation of the falling tendency. But


  • another, perhaps even more important basis forrejecting this law is a political one.40 To conceiveof capitalism as being subject to "laws of motion,"it is said, is to treat a human social arrangement asif it were a machine or some physical process. Thisdownplays and degrades the role of human beingsin determining the course of events. People, notlaws of motion, make history. Furthermore, it isargued, belief in the proposition that the rate ofprofit tends to decline will lead to a fatalistic andpassive attitude towards the task of overthrowingcapitalism. Finally, it is sometimes added that inany case the analysis of the causes of crises is tooabstract an issue to be of use in the practicalpolitics of class struggle.

    There is no question that Marx did conceiveof capitalist history in terms of laws of motion,and of human history in general in terms ofobjective forces acting on and thus limiting humanaction. And yet this is the same Marx whoelevated class struggle to the highest level, whoactively championed the immediate overthrow ofcapitalism (not in some fatalistic future), and whoparticipated in the most practical politics on thebasis of his theoretical analysis. Is there acontradiction between these two aspects of Marx?

    Not at all. On the contrary, as HenrykGrossmann (Germany), Paul Mattick (U.S.) andDavid Yaffe (Britain) argue, it is precisely fromMarx's theoretical framework that revolutionarypolitics flow.

    Grossman was the first major Marxist to shiftthe discussion of crises away from underconsump-tion and disproportionality theories. Heavilycritical of these theories on both logical andpolitical grounds, Grossmann emphasized insteadthe centrality of the law of the falling rate of profitto a theory of crises. Of particular importance inMarx, he noted, is the fact that as the rate of profitfalls, the growth in the total amount of profit mustslow down and eventually halt. At the pointwhere new investment no longer generates addi-tional profits, investment will be curtailed and acrisis will break out.41 As the crisis spreads,weaker and less efficient capitalists will be wipedout, and stronger ones will be able to buy up theirassets at abnormally low prices. With risingunemployment the position of workers is weak-ened. Real wages tend to fall while the labor'process tends to be intensified, so that the rate ofexploitation rises. All of these factors raise the rateof profit. Thus each crisis itself sets the stage forthe recovery and the next cycle of boom and bust.

    None of this says when a particular crisis willbreak out, since many factors can retard or

    accelerate the effects of the falling rate of profit. Inthis sense class struggle is crucial not only in thequestion of the timing of crises, but also in thearena of fighting their effects. Even more impor-tant to Grossmann, however, is that crises are"objectively revolutionary situations." To showthe necessity of crises within capitalism is thereforeto show the necessity of preparing in advance for,and seizing the moment of these objectivelyrevolutionary periods. Lastly, based on his read-ing of Marx, he makes an important connectionbetween theory and practice:

    . . .no economic system, no matter howweakened, collapses by itself in automaticfashion. It must be "overthrown." Thetheoretical analysis of the objective trendsleading to a paralysis of the system serves todiscover the "weak links." Change will comeabout only through the active operation ofthe subjective factors.42

    Paul Mattick elaborated on Grossman's workin a variety of ways. Of particular importance isMattick's point that the reason Marx speaks ofcapitalist society in terms of laws of motion isprecisely because capitalism is regulated not byconscious human decision by rather by "thing-likerelations" the relations of the market, of pricesand profits. Like Grossmann before him, Mattickemphasizes that crises provide revolutionary andreactionary opportunities, but only class strugglecan determine which path will be chosen. Whethercapitalism turns to fascism, or is turned intosocialism, is not determined before hand.43

    In the last few years, David Yaffe has set outboth to present Marx's economic analysis and toapply it to the current crises. The full extent of hisanalysis is beyond the scope of this discussion. Asfar as crisis theory is concerned, in addition topoints similar to those made by Grossmann andMattick, Yaffe adds the following. First, thatbecause a crisis shows up in terms of prices andprofits, there is a tendency to think of prices andprofits as causes of crises. For instance, since bydefinition profit is the difference between sales andcosts, anything which causes a fall in profitabilitywill necessarily imply another way of defining afall in profits. But one portion of costs is merelythe price of some goods, like materials, etc. (andhence the sales of some other industries). There-fore any decline in profitability tends to becompared to the remaining portion of costs, towages, and from here it is only a short step to theargument that "high" wages are the cause of thedecline. In this way an effect is made into a cause.

    Similar points can be made for stagnation,rising unemployment, inflation, rising state expen-


  • ditures and sharpening class struggles all over theworld. Each of these, Yaffe argues, is a phenome-non of the development of the crisis, not a cause.As the rate of profit falls, accumulation will slowdown and unemployment will rise. Capitalists willincrease prices to try and maintain profitability,thus giving rise to an inflationary spiral. At thesame time the State is forced to step in, on onehand to maintain employment at politicallyacceptable levels, and on the other to subsidizeand even take over ailing industries. Stateexpenditure therefore increases rapidly. But thedeficit financing of the State only acceleratesinflation, while its support of employment levelsprevents wages from falling enough to help restoreprofitability. In this way the contradiction isdeepened, and it becomes harder and harder tofind policies which "work." This, says Yaffe, is thestage we are now in, all over the capitalist world.44

    C. Class Struggle and the Profit Squeeze

    Every crisis underscores the importance ofprofits for capitalist production, and raises anewthe question of what regulates profitability.

    Every decline in profitability, in turn, tends,sooner or later, to be traced to high wages. Now,it is certainly true that a reduction in wages, otherthings being equal, will raise profits. But it doesnot follow that a given decline in profits is neces-sarily due to excessive wages. The question is,how do we tell which is cause and which is effect?

    In Marx's analysis, a rising real wage isexpected to accompany a rising rate of exploita-tion, so that by itself the wage rise will notcontribute to a fall in profitability. In Marxistterms, therefore, only when the rise in real wagesis large enough to actually lower the rate ofexploitation can we say that the fall in profitabilityis due (in part, at least) to "high wages."*

    Marx of course rejects this explanation, onthe grounds that the accumulation of capital itselfprovides objective limits within which wagestruggles are confined, so that, in general, the rateof exploitation rises. In fact, he argues that therate of profit falls precisely because workersbecome more exploited, not less.

    At the most abstract level, the moneyequivalent of the rate of surplus value S/V is theratio of "profits" to "wages" TT/VV. A fall in theprofit-wage ratio could then be taken as evidence

    * "Nothing is more absurd . . . than to explain the fallin the rate of profit by a rise in the rate of wages, althoughthis may be the case by way of an exception." (Marx,Capital/V.I, Ch. XVI, p. 240)

    of an excessive rise in real wages. But thisreasoning is false.

    First of all, it is perfectly possible to haveworkers exploited more and hence to have themproduce a greater surplus product, while at thesame time to have capitalists unable to sell thisgreater product and hence be unable to translate itinto money profits. For instance, in a crisisbrought about by a falling rate of profit (a laMarx), as some capitals go out of business otherswill be deprived of buyers for part of theirproducts. Prices will fall, and with them will fallprofits and the ratio of profits to wages. Tocompensate for this, the surviving capitalists willdrive their workers even harder, exploit them evenmore, in an effort to lower costs and stay inbusiness. In the throes of a crisis, therefore, afalling profit-wage ratio will be accompanied by arising rate of exploitation. Moreover, under thesecircumstances both are symptoms, not causes, ofthe crisis.

    But the above pattern would not hold prior tothe outbreak of a crisis. Would it not belegitimate, then, to view the "profit-wage" ratio asan index of the rate of exploitation during non-crisis periods? If so, a falling profit-wage ratioprior to a crisis would be powerful evidence thatworkers had indeed succeeded in raising their realwages fast enough to lower the rate of exploitationand hence precipitate the crisis. It is precisely thistheoretical identification of TT/W as an index ofS/V which defines the "profit-squeeze" branch ofMarxist crisis theory, as set out by Glyn, Sutcliffeand Rowthorn in England and Boddy and Crottyin the United States.45

    Ostensibly, their argument rests on theempirical observation that crises are preceded by afall in the profit-wage ratio. But this sameobservation is frequently made by bourgeoiseconomists too, as in the recent case of WilliamNordhaus of the Brookings Institute.46 UnlikeNordhaus, however, the Marxists go one stepfurther by identifying the observed profit-wageratio with the rate of exploitation. It follows fromthis that the decline in profitability is really anexpression of a fall in the rate of surplus value,which in turn can only be due to a sufficientlylarge rise in real wages. Ironically, whereas thebourgeois economist Nordhaus blames the declineon the "cost of capital," the Marxists attribute it to"labor problems!"

    In a sense, profit-squeeze arguments are asold as capitalism. Nobody knows better thancapitalists how important profits are to thesystem, and for obvious reasons none have been


  • quicker to blame high wages for precipitatingcrises. In this sense, a capitalist version of theprofit-squeeze argument crops up with everycrisis.

    At a slightly more abstract level, bourgeoiseconomists have long argued that falling profita-bility is due to the fact that workers have been ableto increase their "share" of net national product (atthe expense of the capitalist "share," of course).Remarking on two of his contemporaries, theFrenchman Frederic Bastiat (1840s) and the Ameri-can Henry Carey (1860s), Marx notes that though"they accept the fact of the tendency of the rate ofprofit to fall. . .they [mistakenly] explain it simplyand entirely as due to the growth in the value oflabour's share. . .""

    In many respects, the current Marxist profit-squeeze theory is similar to that of Bastiat andCarey. Erik Olin-Wright, in his survey of Marxistcrisis theories, summarizes the modern version inthe following way:

    The essential argument is very simple: therelative share of national income going toworkers and capitalists is almost entirely aconsequence of their relative strengths in theclass struggle. To the extent that the workingclass develops a strong enough labor move-ment to win wage increases in excess of pro-ductivity increases, there will be a tendencyfor the rate of exploitation to decline and thusfor the rate of profits to fall (to be "squeezed"by rising wage bills). Such a decline in profitsresults in a corresponding decline in invest-ments and thus even slower increases inproductivity. The end result is economiccrisis.48

    The modern Marxist version, therefore, fol-lows the economic logic of Bastiat and Carey intaking the tendency of the rate of profit to fall tobe a consequence of a falling rate of exploitation.But there is a crucial political difference betweenthe two versions in that whereas the bourgeoiseconomists decry this situation, the Marxistscelebrate it. The Marxist profit squeeze theorymakes class struggle over working conditions thecrucial factor that (in the last instance) determinesthe course of capitalist reproduction. To theseMarxists, the fact that the development of thesystem has reached a stage where labor is strongenough to precipitate crises is a very hopeful sign.If the working class is able to bring the system toits knees through its wage demands, then it mayalready be strong enough to resist the attacks onthese real wages which are part and parcel of the"recovery" process. They may perhaps even bestrong enough to "solve" the crisis by taking over

    state power.The great virtue of this theory is its simplici-

    ty. Even in capitalism, we have "politics incommand." The practical politics of class struggle,not some abstract laws of motion, are what weneed to analyze in order to understand capitalisthistory. Capitalist accumulation is indeed inter-nally limited, but it is labor, not "capital itself" (asMarx says), which is the ultimate barrier toaccumulation.

    Simplicity is really a virtue only if the simpleexplanation is the correct one. The penalty forbeing wrong, after all, is failure. And so we goback to the central theoretical point, and ask: canwe in fact impute a falling rate of exploitationfrom an observed fall in the profit-wage ratio? Inother words is TT/W indeed an index of S/V? Toanswer these questions, we need to trace out themoney forms of S and V.

    Consider the end of a reproduction cycle oftotal capital. Beginning with the sales receipts, wecan then trace out the disbursement of this moneysum.

    Suppose total sales (M') amount to $100,000.From these, capitalists set aside $40,000 to replacethe costs of materials and machines used up (C*) inproducing the commodities which were sold, and$20,000 to replace the wages advanced (V*) to theproduction workers who were employed in theproduction process.* The remaining amount,$40,000, is what capitalists themselves call grossprofits on sales (S*). It is the receipts from the saleof commodities minus the material and labor costsof producing these commodities. From the point ofview of the system as a whole, these gross profitsrepresent the money equivalent of the surplusproduct.

    From the Marxian point of view, "grossprofits" (S*) represent the money equivalent ofsurplus labor-time of production workers, whilethe wages (V*) of these workers represent themoney equivalent of their necessary labor time.The proper index of the exploitation of productionworkers i.e. of the rate of surplus value istherefore:

    S * / V p * = 4 0 , 0 0 0 / 2 0 , 0 0 0 = 2 0 0 %

    But to capitalists matters look very different.From gross profits they still have to deduct the

    * I am using the term production workers because it isnot possible to adequately develop the concept ofproductive labor within the confines of this paper.Similarly, I use the term commodity to cover both goodsand services which are sold for money. The distinctionbetween productive and unproductive labor does notreduce to the simplistic distinction between goods andservices.


  • money they have expended in trying to sell thecommodities. These selling expenses, as capitalistscall them, consist of the material (C*j) and laborcosts (V*j) of transforming the produced com-modities into money sales. In addition, they mustalso deduct indirect taxes T (sales, license andproperty taxes, etc.), because from their point ofview these too are an "expense" of business. Whatis left after all these deductions is called netcorporate income* (TT). If selling expenses C* +V*, = $15,000 + $10,000, and indirect taxes T =$5,000, then net corporate income TT = $10,000.

    From the point of view of the capitalist classboth selling expenses and indirect taxes aregenuine expenses of business. Indeed, even fromthe point of view of the system as a whole theymay be' regarded as strictly necessary expendi-tures, since both commercial capital (wholesalers/retailers) and the state perform indispensablefunctions. But the fact that they are indispensableexpenditures does not in any way alter the factthat they are also derivative forms of surplusvalue. It is necessary to produce the surplusproduct before it can be sold; selling it onlychanges the title of ownership of this product, notits magnitude. The extent to which part of thesurplus product is absorbed by title-changingactivities (buying and selling) and by stateactivities** is merely an index of the distributionand legitimation expenses of the system.

    Unfortunately, the profit-squeeze theoristsfail to grasp this crucial point.*** Invariably, theyidentify the rate of surplus value with TT/VV, theratio of net corporate income to all wages. Interms of the above illustration, TT= $10,000, andW = wages of production workers + wages andsalaries of salespeople, etc. = $20,000 + $10,000= $30,000, so that TT/W = 10,000/30,000 =33V3%. This is a far cry from the true rate ofsurplus value, S*/V* = 200%.

    By confusing TT/W with S*/V*, the true rateof exploitation in any given time period is vastlyunderstated, as the above example illustrates (themagnitudes involved correspond quite closely tothe actual current magnitudes I get on the basis ofa much more complex and detailed analysis of the

    * Net corporate income will in turn be split into corporate income tax, pure rent (as opposed to depreciationand maintenance of buildings and equipment, which isproperly speaking a part of costs of production or of sellingexpenses), interest/ dividends, and retained earnings.

    * "Other than actual state production.

    * * *This important criticism of the profit squeeze logic isalso made in a critique of Boddy-Crotty (op. cit.)appearing in The Communist, Vol. 1, No. 2.

    U.S. economy*). What is worse, TT/W has a builtin downward bias over time relative to the truerate S*/V*, because in all capitalist economies,both selling expenses and indirect taxes have risensharply. This is particularly true since World WarII. It is fallacious, therefore, to explain theobserved fall in the "profit-wage" ratio TT/W bymeans of an imputed fall in the rate of exploita-tion. On the contrary, it is quite possible that arising rate of exploitation, accompanied by afalling rate of profit a la Marx, has resulted in adeclining rate of accumulation and rising unem-ployment. In the face of this, increased inter-capitalist rivalry and increased state interventionappear as responses to a deepening crisis, notcauses of it. Empirically, these responses manifestthemselves as increased selling expenses andincreased taxes, which show up as declining TT/Weven though S*/V* is rising. This, in effect, isYaffe's explanation of the current world crisis.

    It is worth repeating that an observed declinein the "profit-wage" ratio TT/W does not by itselfprovide us with an explanation. To get behindmere observation, we need a theory of thedeterminants of profits in order to know whichfactors are accountable for the empirical trend.But we also need to know how the empirical cate-gories correspond to the theoretical ones, forotherwise we will end up imputing the wrongcause. This is precisely the error made by theprofit squeeze school: they base themselves on thetheory of surplus value, and yet they completelyfail to consider the difference between thiscomplex and powerful Marxian category and thebourgeois category of "profit" (net operatingincome). In this way they mistakenly attribute thesecular decline in profitability, and hence thecurrent world crisis itself, to a squeeze on profitsemanating from wages.


    History teaches us that capitalism is periodi-cally subject to ruptures in its economic and social

    * In actual practice, reconstructing the money equiva-lents of S*, V*, and W is vastly more complicated thanindicated above. In effect, it involves transforming nationalincome accounts, based on Keynesian categories, intoMarxian accounts based on the value categories of Marx.This is a theoretically difficult task, as well as beingempirically difficult. Nonetheless, it is quite feasible andindeed absolutely necessary. On the basis of fairly detailedcalculations (which I obviously cannot present here), I findthat the real rate of surplus value S/Vp* rises from 1900 to1972, while TT/W declines over the same period. Rising"expenses" and rising taxes account for a large part of thisdiscrepancy.


  • fabric. At these times the social tensions inherent inthe system stand out in sharp contrast. Thebourgeois platitudes of the various orthodoxiesbegin to wear thin, to take on a desperate air, and thestruggle of the classes breaks out into the open.

    We are learning this lesson of capitalisthistory once again. The post war boom which was tousher us through the golden gates of the 21st centuryis now officially dead. All over the capitalistworld, political and economic crises abound.International competition intensifies as capitalistsstruggle to survive; banks fail, giant industries fail,the international monetary system itself lurches fromone crisis to another; unemployment deepens whileprices continue to rise and everywhere the classstruggle sharpens.

    How are we to understand this latest crisis ofcapitalism? Certainly we must study and analyze itin detail, not just locally or nationally but on aworld scale. But that by itself will never beenough. We must at the same time understand thatcrises are nothing new to capitalism. Theirperiodic and devastating appearances have beenrecognized, analyzed, and theoretically grasped bymany others long before we even came to ask thequestion. To understand this is to understand thenecessity of studying the explanations of ourpredecessors, so that we may benefit from them andto build upon them. If the task is to overcome thesystem, then it is imperative to understand it. Theprice of ignorance is failure.

    The object of this paper has been to present andanalyze the basic positions which have historicallyemerged on the question of capitalist crises. I havetried to be as rigorous as possible in this task, whileat the same time not assuming any prior knowledgeof the subject. In so doing I have tried to present notonly what a particular type of theory says, but alsowhy it makes that argument, how the argumentdevelops historically, and what political positionshave been associated with it at various times.

    Rather than summarize what has been dis-cussed in the body of this paper, I would insteadlike to focus on three lessons which I believe areimplicit in the history of crisis theories.

    The first lesson has to do with the relationbetween theory and politics. Each theoreticalposition implies a certain mode of changing thesystem. In that sense, every theory has politicalimplications for the practice based upon it. But it isimportant to realize that no simplistic connection

    FOOTNOTES1. See Alchian, A.A. and Allen, W.R., Exchange andProduction Theory in Use (Belmont, Ca.: WadsworthPublishing Co., 1969), Chs. 1-4, for a forthright presentationof the neoclassical conception.


    may be made between a particular set oftheoretical concepts and the politics which canbe expected to be allied with them. Take thecase of underconsumption theory, forinstance. Its proponents have included thereactionary Parson Malthus, the petty bourgeoissocialist Simone de Sismondi, the revolutionaryactivist Rosa Luxemburg, and the wholemodern "monopoly capitalism" school basedon the work of Paul Sweezy and Paul Baran.Its opponents, on the other hand, includebourgeois theorists of all stripes, from Ricardoonwards, but also Marx and Lenin. Neitheramong the supporters of underconsumptiontheory nor among its critics can anycommon political position be discerned.Similar arguments can be made for every othercrisis theory.

    The second important lesson has to dowith theory and*the "facts." It is a very seriousmistake to assume that "facts" are somehowgiven independently of any conceptualframework. Even a brief study of the history ofnational income accounts quicklydemonstrates that the data we are confrontedwith at any given time is the numericalrepresentation of particular theoreticalcategories. These data are of course based onevents in the real world, but the manner inwhich these events are codified andenumerated also depends on a theory about theworld. The pattern which emerges on the basisof Keynesian categories (which underlycurrent national income accounts) need not beat all the same as that which emerges on thebasis of Marxist categories. In the discussionof the profit squeeze theories, for instance,we saw how important it was not to confusethe profit-wage ratio (TT/W) with the rate ofexploitation (S/V). It would be a terrible lossindeed to abandon a correct theory because itdoes not correspond to "facts" which arebased on entirely different categories.

    The third lesson has already beendiscussed at the beginning of this section. Tobriefly reiterate it, in analyzing the crisis it isnot sufficient to just study its phenomena. It isequally necessary to study the explanations ofcrises, both past and present. Otherwise we arevery likely to simply reinvent what hasalready been invented, and make the samemistakes which others long ago have made.It has often been said that those who ignorehistory are condemned to repeat it. To this itshould perhaps be added that those whoignore theory are condemned to reconstruct it.

  • 2. Wesley Ctair Mitchell, "Business Cycles," in Readingsin Business Cycle Theory, American Economic Association(London: George Allen and Unwin, 1961), p. 43; Samuel-son, Paul, Economics (New York: McGraw-Hill Book Co.,1976), pp. 250-251.

    3. Samuelson, op. cit., p. 257.4. Robert Lekachman, A History of Economic Ideas

    (McGraw-Hill Book Comapny, 1976), p. 343.5. Joan Robinson, Economic Heresies (New York: Basic

    Books, Inc., 1971), p. x.6. Lekachman, op. cit., p. 347-348. This was very much

    Keynes' own perspective, and it continues to be reflected inthat of his followers.

    7. Michael Bleaney, Underconsumption Theories: A History and Critical Analysis (New York: InternationalPublishers, 1976), p. 63.

    8. Michael Barrat-Brown, Economics of Imperialism(London: Penguin Books, 1974), p. 170.

    9. Bleaney, op. cit., pp. 153-168.10. Ibid., p. 180.11. Ibid., p. 171.12. Hobson quoted in Bleaney, op. cit., p. 166.13. For a discussion of what "normal" prices are and how

    they are determined in Marx, see my article, "Marx'sTheory of Value and the Transformation Problem/ " inThe Subtle Economy of Capitalism, ed. Jesse Schwartz,(Santa Monica, Cal.: Goodyear Publishing Co., Inc.,1977), pp. 106-137.

    14. Karl Marx, Grundrisse, (London: Penguin Books,1973) translated with a foreword by Martin Nicolaus, pp.56-58.

    15. Russell Jacoby, 'The Politics of the Crisis Theory:Towards the Critique of Automatic Marxism II," Telos, 23,Spring 1975, pp. 5-11.

    16. Ibid., p. 10. The quote is from Danielson.17. Ibid., pp. 14-16.18. Ibid., p. 22.19. Bleaney, op. cit., p. 89.20. Ibid., p. 193.21. Paul Sweezy, The Theory of Capitalist Development,

    (New York: Monthly Review Press, 1942), p. 179.22. Ibid., p. 183.23. Paul Baran and Paul Sweezy, Monopoly Capital

    (New York: Monthly Review Press, 1968), p. 108.24. Paul Sweezy, "The Economic Crisis," Monthly

    Review, Vol. 26(10), March 1975, pp. 1-8.25. Erik Olin-Wright, "Alternative Perspectives in the

    Marxist Theory of Accumulation and Crisis," in Schwartz,pp. 215-226.

    26. Bleaney, op. cit., p. 225.27. Ibid., pp. 245-248.28. See Joseph Steindl, Maturity and Stagnation in

    American Capitalism, (New York: Monthly Review Press,1976), p. xvii, footnote 7.

    29. Ibid., pp. xv-xvi.30. Karl Marx, Capital (New York: International Pub

    lishers, 1967), Vol. Ill, p. 250.31. Ibid., p. 213.32. Karl Marx, Capital (New York: International Pub

    lishers, 1967) Vol. I, p. 604.33. Sweezy, op. cit., p. 88; Maurice Dobb, Political

    Economy and Capitalism (London: Routledge & KeganPaul, Ltd., 1937), pp. 108-114.

    34. David Yaffe, "Inflation, the Crisis and the Post-WarBoom," Revolutionary Communist, No. 2,1976, pp. 5-.

    35. N. Okishio, "Technical Change and the Rate ofProfit," Kobe University, Vol. 7,1961, pp. 85-99.

    36. Marx, op. cit., Vof. Ill, p. 264. See also a realexample of this process which Marx provides and analyzesin the Grundrisse, op. cit., pp. 383-385.

    37. Geoff Hodgson, "The Theory of the Falling Rate ofProfit," New Left Review, 84, March-April 1974.

    38. See Victor Perlo, "Capital-Output Ratios in Manufacturing," Quarterly Review of Economics and Business,8(3) Autumn. 1966, pp. 29-42.

    39. Karl Marx, Theories of Surplus Value (New York:International Publishers, 1967), p. 439.

    40. The following discussion draws heavily from Jacoby,op. cit., section V.

    41. Ibid., p. 35.42. Ibid., p. 37.43. Ibid., p. 43.44. Yaffe, op. cit., pp. 5-32.45. See Andrew Glyn and Bob Sutcliffe, British Capital

    ism, Workers and the Profit Squeeze, (London: PenguinBooks, 1972); Bob Rowthorn, "Mandel's Late Captialism'"New Left Review, 98, July-August 1976, pp. 59-83; andRaford Boddy and James Crotty, "Class Conflict andMacro-Policy: The. Political Business Cycle," Review ofRadical Political Economics, Vol. 7, No. 1,1975.

    46. William Nordhaus, 'The Falling Share of Profits,"Brookings Papers, 1976, No. 1, pp. 169-208.

    47. Marx, Grundrisse, op. cit., p. 755.48. Wright, op. cit., p. 216.



View more >