The Diversity of ClassAnalyses: ACritiqueof Erik Olin Wright and Beyond

STEPHEN RESNICK AND RICHARD WOLFF (University ofMassachusetts)

ABSTRACT Class analyses areboth very oldand quite new. This essay arguesthat Marx contributeda new class deŽnition andanalysis focusedon the production,appropriation, and distributionof surplus labor. Yet, that innovative,new class analysis was lost bybeing dissolved into either pre-Marxian conceptualizationsof class in terms ofproperty and power or later social theories in which class was determinedby people’ s consciousness andself-identiŽ cations. In this context, the essay pays special attentionto the recent work ofE.O. Wright. Class analysis basedon the surpluslabor deŽ nition of class is comparedand contrasted with Wright’s differentlybased deŽnition and analysis.

Introduction The pages of InsurgentSociologist and then Critical Sociology have been an importantlocus of work, too numerous to list here, inwhichclass analysis featuredprominently and effectively. However,the complexand rich traditionof class analysis thatbore fruit in this work, as elsewhere, also produceda multiplicityof different deŽ nitions of class. The result has been adivergencein types ofclass theory informing correspondingly different analytical andpolitical projects. We believe, forreasons outlined below, thatit important henceforth to take intoaccount explicitly the irreducibly basicdifferences in how class is deŽ ned and used. Unfortunately, however,

Critical Sociology, Volume 29,issue 1 also availableonline Ó 2003Koninklijke Brill NV,Leiden www.brill.nl 8 Resnick &Wolff ² workkeeps appearingin which one or another class deŽ nition and deploymentis used as if itwerethe only orthe ultimateor the universally agreedkind of class analysis. Instead,we propose that henceforth each newclass analysis should(1) acknowledge the multiplicityof possible class deŽnitions and deployment, and (2) offer an argumentfor whichever deŽnition(s) and deployment each uses. Tomake the case forthis sea change inthe methodof class analytical work,we will Ž rstsketch the basicdifferent kinds of class analysis that have dominatedthe tradition.In thiscontext, we willpay special attention tosituating there the recentwork of a well-knowncontemporary class analyst, ErikOlin Wright(EOW). Then wewill sketch akindof class analysis sharplydifferent from those dominating kinds and now rapidly generatinga newclass analytical literature.We willexplicitly arguewhy thiskind of class analysis needs tobe incorporated into a classanalytical traditionnow mature enough to require each newpiece of class analysis tojustify its partisan position within that tradition. The next three articles inthis section apply this new kind of classanalysis toissuesof racism,class strugglesinside modern households, and Third World development. The propertied(rich) versus the propertyless(poor): can one think ofa moretraditional way to conceive of any society’s classstructure? Every day, incasual conversations, journalistic accounts, and academic presentations, peopleacross the politicalspectrum seek tounderstand social issues by analyzing them interms of how those possessing more wealth interact withthose possessing less. They call these twogroups classes. Certainly, noother conception of class played a moreprominent role in twentieth centurysocialism. For that century’ s radicals,socialists and communists, as formany oftheir opponents, altering a society’s structure(distribution) of propertyownership more or less deŽned a revolutionin its class nature. Indeed,the classiŽcation of individuals into the propertyless-poorand the propertied-rich,for all sortsof social purposes, goes back thousands of years inEuropean (and other regions’ ) histories. 1 Over the lastcentury, the classiŽcation has evolved intoa developedanalytical frameworkand traditionespecially withinMarxism. 2

1 See,for example, G.E.M. de Ste. Croix’ s discussionof Plato and Aristotle’ s property classiŽcations (1981: 70-80). 2 Marxism’s “relationsof production” often are conceived as “classrelations” understood, in turn, tobe property relations (Howard and King 1985: 6-7). Support for a property notionof class can befound throughout the Marxiantradition from itsearly beginnings (Engels1976: 181; Kautsky 1971: 43) to its Soviet period (Stalin 1954:683-684; Bukharin andPreobrazhensky 1988:28; Trotsky 1987:248) to its more recent developments(Sweezy 1964:242; Lange 1963: 16; Kuczynski 1971:10; Cutler et al.1977: 243; Hirst 1979:96; Callinicos1982: 149-150; Roemer 1988: 72-89). A relatedbut still different notionis ACritiqueof Erik Olin Wrightand Beyond 9 ²

Despiteits venerable standingwithin Marxism – andindeed within socialtheory generally –the propertyconception of class always had criticswho advocated alternative conceptions.Of these, the twobest known have advancedpower and consciousness conceptions of class. From their perspectives,class analysis focusesproperly not on property but rather onpower or on consciousness. For those embracing power theory, what mattersmost is the distributionof power across the populationof any societychosen for scrutiny. 3 Whohas whatkind of authority/ control overwhom? Instead of the richversus the poor,power theorists of class juxtaposerulers and ruled. In furthercontrast, theorists of class who stressconsciousness push property and power into the backgroundof their investigationsin favor of primary attention to the consciousnessdisplayed bysocial groups. Classes only exist forthem if,when, and to the extent thatindividuals think ofthemselves asa classuniŽ ed in speciŽ c waysand differentiatedfrom others in speciŽc ways. 4 Howanyone undertakesa classanalysis andwhat conclusions he/ she reaches dependupon which deŽ nition of class – property,power, or consciousness– the personuses as his/ her“ entry point”into social understanding.By entry pointwe mean the focalconcept(s) around which adiscourseis organized (Resnick andWolff 1987: 25). For example, in asocietywhere radical change transfersthe ownershipof the means of productionfrom private shareholders to the stateas representative of all workers,that change willcount as a classrevolution for those who deŽ ne classin property terms and who make thatdeŽ nition an entry-pointinto socialanalysis. However,those who use class as an entry pointbut deŽ ne classdifferently – say, asa radicaldemocratization of power – may well reacha differentand even oppositeconclusion about that same change in the socialdistribution of property. Should they Žndthat power was merely transferredfrom one elite (private)to another (state), then the property change falls farshort of aclassrevolution for them. Class-qua-consciousness theorists,if they Žndthat the consciousnessof social groups was little altereddespite changed distributions in property and/ orpower, would thereforeargue that no class revolution was achieved despitethose changed

Elster’s (1986:319-397) conception of class in terms ofindividuals’ access to a broadly deŽned notion of “ endowments”(property, skills, and still other “cultural traits”). This notionis presentedin andused by Eric OlinWright toformulate his notion of the “middle class”(1997a: 22-23). 3 Formulationsof class in terms ofpower are in Dahrendorf (1959:137), Mosca (1939: 50ff)and Mills (1956: 3-20; 1962: 106ff). Marxists also understand class in terms ofpower conceivedas struggle: Laclau (1977: 106), Przeworski (1987:67), and Jessop (1980: 63). 4 Thompsonpresents what wecall belowa compositenotion of class but in the end reducesthat notionto consciousness (1963: 9-10). 10 Resnick &Wolff ² distributions.Property, power, and consciousness – like the differentclass conceptsthey inform– areneither equivalent,nor interchangeable, nor linked inany Žxed way.A change inany onecan occur together with asimilaror opposite or no change inthe others.Class analyses and classstruggles have, fora longtime, re ected a rangeof irreducibly differentconcepts of class. What seems tous long overdue is an explicit recognitionand discussion, among class analysts, ofthose differences and the theoreticaland political stakes inselecting amongthem. The intense globaldebates over the contentand signiŽ cance of the Soviet,Chinese, Cubanand similar revolutions were, in large part, confrontationsof scholars,activists, and others deploying different concepts ofclass among their respective entry points.No small confusionand frustrationcharacterized those debates in large part because the debaters wereso often unaware of their differences at the basiclevel ofclass deŽnition. They typically spokeand wrote about class and class con ict asifeveryone sharedroughly the same deŽnition. As weproposeto argue here, becausebasic differences in classdeŽ nition exist andhave important analytical,political, and social effects, they need tobe acknowledged and debatedas such. How anyone deŽnes classshapes how social action is understoodand engaged: the theoreticaldebate over class is a practical matterfor activists no less thanfor social theorists. Beyond the contestingproperty, power, and consciousness deŽ nitions ofclass,further problems bedevil classanalysis. Afterall, the inŽnite char- acteristicsof human beings have providedinnumerable ways to disaggre- gatethem intogroups, quite literally to classify them:into age, education, prestige,status, productive skill, “life chances,”ethnic, and many other classes.Moreover, the multiplicityof deŽ nitions of classover the years has inspiredmany tocombine various singular conceptions into composites. Thus,for example, onecan undertake a classanalysis bygrouping indi- vidualsaccording to the property-and-power thatthey possess,to their wealth- and-education-and-prestige, andmany othersuch composite entry points.For example, these compositesare often rendered, by Marxists, as follows: class isdeŽned in terms of “relationsof production” or as a positionthat a per- sonoccupies within a modeof production. 5 Poutlantzas(1978) offered aparticularlycomplex (composite) class deŽ nition comprising ideological, political,and economic relations. These relationsincluded “ politicaland ideologicaldomination and subordination” and “ real economicownership”

5 Cohen presentsa notionof relations of production as “ effectivepower over persons andproductive forces” (1978: 63) and Carchedi offershis notion of “ productionrelations” asthe “ownersand the possessors”who “ havethe power”to order the economy(1991: 39). ACritiqueof Erik Olin Wrightand Beyond 11 ²

(14-24).Similarly, followers of Max Weber (1922)deŽ ne classin the com- positeterms of a person’s material“ life chances”or economic advantages withinan economy,while those of Giddens (1973) deŽ ne aclasscomposite inwhichpower, property, education, and skill arethe centralterms. Marx andmost other sophisticated social thinkers have useddifferent singularconcepts of classat variouspoints in theirwork and/ orcombined them intovarying composites.To the extent thatthey acknowledged changes intheir concepts of class, these theoristssometimes justiŽ ed them onempirical grounds: contemporary reality hadfractured the social landscapein ways requiring revisions and/ orcomposites of received conceptsof class. On otheroccasions, failed political projects prompted theiradvocates to alter their views. Shifting popular usages of class have forcedintellectuals torethink theircategories, and vice-versa. Puristswith singularconceptions of class deployed strictly have oftenviewed other conceptionsas wrongor primitiveor politicallyobjectionable (or all three). They thereforeclashed with those whose reverence beforesocial complexity madethem suspectany singularconceptions and prefer composites that struckthem asbettersuited to theircomplex objects. Ourconcern is not with the debatebetween singular and composite conceptionsof class. Rather, we wish to stress that both singular and compositedeŽ nitions of class can be analyzed forwhat they includeand whatthey exclude.The classanalyses informedby both must inevitably highlightsome aspects of the reality they seek tounderstand and change whileignoring or marginalizing other aspects. What thoseanalyses include intheirsingular or composite deŽ nitions of classwill become central to the understandingsthey generate. What they exclude willbe seen assecondary ornot seen atall. Likewise,the classchanges pursuedand achieved bysocial movements willdepend in large part on what the hegemonic conceptionsof class among them includeand exclude, what they make visibleand important and what they obscure. Here wewill focus chie y onproperty, power, and consciousness since they arethe three deŽnitions of class central, singly orwithin composites, toalmost all classanalyses. Concentratingon these three facilitatesthe exposureand exploration of the basicdifferences among class analyses. Thisfocus on property, power, and consciousness also enables usto highlightan alternative concept,one we think originatedand deŽ ned carefullyby Marx in surpluslabor terms .Remarkably, thisalternative concept ofclass, which he deployedsystematically (albeitnot exclusively), faded frommost subsequent class analyses (Marxist andnon-Marxist) or else was assigneda very marginal,secondary role (Norton 2001: 23-55). We will assess somereasons for the absenceof this particular class concept and 12 Resnick &Wolff ² alsosome of its consequences. Our assessment willalso serve toexplain ourpartiality to the surplusconcept. Toillustrateand concretize our argument, let usreturn,for a moment, totraditionalMarxism’ s entry pointconcept of classas property-ownership. Thatnotion enables endless gradationsof property and hence ofclass(es). Between the twomain classes –forexample, the propertylessproletariat andthe propertiedbourgeoisie – lie the so-calledmiddle classes thatown less (ordifferent kinds) ofpropertythan the bourgeoisiebut more than the proletariat.Another way to highlightsuch gradations has been toelaborate notionsof class-fractions,sub-classes, strata within classes, and so on. Such traditionalMarxism focuses on class structures and struggles in terms of differentiatedproperty-owners. 6 Howeversubtle and varied the gradationsin property ownership informingany suchclass-qua-property analysis, the problemof power remainsas a criticalopening for the alternative theoristsof class whose entry-pointis the socialdistribution of power. The lattermight stress, for example, the differencebetween a legal rightto own property and the actualcontrol over its disposition (Cutler et al.1977: 243-251). An infant may legally ownconsiderable property, but a guardianmay effectively controlthe useof thatproperty. In thissituation, deŽ ning a classstructure interms of who owns what would give adifferentand, for class-as-power critics,a misleadingpicture of the classnature of a societyand hence of thatsociety generally. The same logicapplies if we substitute stockowners andboards of directorsof modern capitalist corporations for the infantand the guardianin this example. PowerdeŽ nitions of class and their applications to social analysis also displaygradations and subtleties. Between the all-powerfuland the utterly powerlesslie intermediatepositions, the “middle”classes forthe power theorists:people who both wield power over others and have somewielded overthem. While the non-owningguardian of our previous example may controlthe useof the infant’s legally ownedmeans ofproduction (for example, sharesof common stock in a capitalistcorporation), the guardian willnot control most of the activitiesof that corporation (including the directionand supervision of its workers). Power to order the behaviorof workersusually rests in the handsof the corporation’s directorsand/ or the managersthey hire.Typically, neither directorsnor managers own signiŽcant shares of stock. In Berle’s (1959)apt (and decidedly non- Marxist) descriptionof many years ago,such managers have “power

6 Lenin,for example, discusses the “differentiationof the peasantry”in Russiabefore the Bolshevikrevolution in terms ofdifferently-sized areas of landcultivated, numbers and kindof livestock used, and implements owned(1972: 70-187). ACritiqueof Erik Olin Wrightand Beyond 13 ² withoutproperty.” The powertheorist may well pursuea classanalysis focusedon the relatively propertylessdirectors and managers, and only secondarily,if at all, onthe shareowners(or their guardians). It follows that whileproperty theorists would designate as “the capitalists”those who own corporations’shares; the powertheorists rather designate the corporations’ boardsof directors, those with mass of powerin and over corporate affairs. Powertheorists often build their class analyses aroundstruggles over power amongcorporate directors and top state ofŽ cials (a “ruling”class or classes thatgive orders),a professional-managerial“ intermediate”class (that both gives andtakes orders),and the wage-earners(a “ruledclass” that takes orders).Many Marxistshave foundvariations of such a powertheory very attractive,while others have triedto combine together such power theorizationswith a propertynotion into a singularcomposite theory of a corporatecapitalist class (Baran andSweezy 1966: 34-35). Fora class-as-consciousnesstheorist like E.P.Thompson (proudly Marxist),neither the share-owninginfant, nor the guardian,nor the directorsand managers will necessarily forma class.Indeed, he ismost famousfor applying this concept to the workers(1963). They only willbecome classes andtheir actions class struggles if and when they have acquireda distinctconsciousness of themselves asa class.It is, as Thompsonargues so forcefully in his Preface (9-14), ultimately a question ofculturaldevelopment: have socialconditions produced a culturein which distinct class identitieshave been forgedand thus a trueworking class made. 7 ForThompson, struggles among social groups who do not (or notyet) think ofthemselves asclasses arenot class struggles, but rather otherkinds of struggles with different social effects. In the languageof contemporarydebate (where culturelooms even largerthan it did for Thompson),the issueis framed in terms of the “subject-positions”of individuals,how society calls them andthereby deŽnes whetherand how classŽ guresin theiridentiŽ cations of self andother. Toillustratethe inclusionsand exclusions of alternative classdeŽ nitions andto draw out their implications and those of the theoreticalchoice betweendeterminist and overdeterminist logics, we consider a leading moderntheorist of class, the sociologistErik Olin Wright.His work offersa particularcomposite deŽ nition of class. Our discussion of Wright preparesthe groundfor counterposing a basicalternative conceptualization (a surpluslabor class deŽ nition). Throughout, we will consider what may beat stake inthese differentclass concepts. The remainingessays inthis

7 Cohen, whileotherwise praising Thompson’ s studyof the Englishworking class, criticizes hiscultural notionof class (1982: 73-77). Cohen deploysa contendingand alternative property-powernotion of class. 14 Resnick &Wolff ² sectionwill then illustratethe kindsof insights that this alternative to Wrightcan generate. We readEOW asdesiringto remainwithin the broadMarxian tradition andwhat he understandsas its property based notion of class. On the otherhand, his approach explicitly recognizesand proposes solutions for problemspresented by a singularproperty deŽ nition of class. The binary structureof a singularproperty notion of classstrikes EOW asinadequate especially fortheorizing individuals who own little orno property, who thereforesell theirlabor power, and yet wieldconsiderable power over otheremployees inthe enterpriseswhere they work.He isalso concerned thatsuch individuals receive muchlarger incomes than the employees overwhom they wieldpower do. For EOW these powerfuland well-paid employees donot Ž tintoa singularproperty theory of class: neither to itsbourgeois nor its proletarian classes. He thereforeoffers a different concept:a compositedeŽ nition of class. His composite includes property ownershipbut also power and income. He worksthis out theoretically and then applieshis particular concept to elaborate empirical studies of the so-deŽned class structures of various societies and especially the U.S. 8 The propertynotion of class lies atthe coreof EOW’ s composite, framingits other entry points.It merits some special attention to understandnot only EOW’s particulartype ofclass analysis butalso the many othercomposites that, while different from EOW’ s, share the quality ofbeing grounded by a propertydeŽ nition. The shared generic property conceptof classpresents capitalism as asocietyin which a classof workers, separatedhistorically from ownership of the means ofproduction,confronts aclassof capitalistswho through that same historicprocess came to possess thosemeans. A historicallyevolved distributionof property (in means of

8 Astilldifferent issueconcerns the causalrelationship between EOW’ s compositenotion ofclass and all other non-classparts of society (for example,race, gender, ethnicity, consciousness,state, and so forth). Weread in hiswork the rejectionof the ideaof determinationby class alone (1997a: 1 andespecially chap. 13). Instead, he (1997a:388) allowsfor “ reciprocaleffects” “between structure andagency” and, we thus assume in the context ofour argument here, betweena compositenotion of class and all other aspects ofsociety.This welcome move to a causalnotion of overdetermination(Resnick and Wolff 1987:chaps. 1 and2) is unfortunately compromisedin hiscontinued claim that a“class structure”alone determines the “limits”within which other factorsoperate (1985: 29-31; 1979:15-29, 102-108; 1997a: 387-388). As we have argued elsewhere (Resnick and Wolff 1992:131-140; 1994: 41-42), one cannot afŽrm afullnotion of “ reciprocaleffects” or ourpreferred term ofoverdetermination and simultaneously hold onto some kind of last- instancedetermination or limitingpower bestowed upon some aspect(s) in society.To do so istoretreat backonto traditional determinist groundsin which, yet oncemore, one chosen part ofsociety (its “ classstructure” ) determinesin the lastinstance (“ limits”or “ultimately determines”) other effectivities. ACritiqueof Erik Olin Wrightand Beyond 15 ² production)structures the classes ofcapitalist societies. As a resultof this classstructure, the massof dispossessed workers have nochoice but to sell tocapitaliststhe only remainingasset suchworkers privately own: their laborpower. Receiving money inexchange, they canpurchase some of the commoditiesproduced by capitalist enterprises and thereby bothsurvive andreproduce their labor power for sale onceagain. Capitalists purchase laborpower because without it the means ofproduction they possess cannotbe worked up into commodities, the value ofthose means would notthen expand,and such capitalists’ survival would then bethreatened bythe whirlwindof competitionamong them. Marx andcountless other economic theorists (both Marxist and non- Marxist) have developedthis generic property conception of class, variously nuanced,to high levels ofreŽ nement. Thus,for example, the distribution ofproductive property thus drives the propertylessand the propertied intoa marketrelationship (commodity exchange) thatboth enables each tosurvive and also makes them antagonists.The propertylesswant ever morefor their labor power while the capitalistswish always to pay less; the latterstrive to charge more for their commodities while the workers seek topay less. Thisproperty paradigm makes the interdependenceand contradictionsof the twogreat classes effects derivedfrom (determined by) a singular cause,namely the distributionof property in means ofproduction. Withinthis paradigm, Marx’ s specialinsight was to show how the propertydistribution enabled the capitaliststo increase the value ofthe means ofproduction they owned.The hiredworkers added more value tothose means duringthe productionprocess than the value paidto them fortheir labor power. This “ morevalue” – translatedfrom the German as“ surplusvalue” – wasthe secretMarx iscredited with having revealed inhis analysis ofcapitalism. Capitalists appropriate the surplus (whatthey call “proŽt” ) andthereby increasethe value oftheir means of production;the workersmerely reproducetheir labor power and remain compelledto sell itagain to the enrichedcapitalists. The proŽts and hence enrichmentaccruing to the latterthus arise most basically from the production“ relations”imposed by the distributionof productive property. Marx thusridiculed the capitalists’claims that it was the supervisorylabor they performed,the risksthey took,the savings they accumulated,or their marketshrewdness; these weresecondary details explaining differential performanceamong capitalists. All capitalists relied upon the unequal distributionof property in means ofproduction for the existence ofand fortheir exclusive accessto surplus value. Here toolay the revolutionary potentialof Marx’ s insight.If embraced by the workers,it could lead them torefuse to produce the surplusfor capitalists and insist instead that they shouldcollectively appropriatethe surplusthey produced.The logical 16 Resnick &Wolff ² means toachieve thisrevolution would require a fundamentalchange in the distributionof property in means ofproduction: the workerswould have todispossess the capitalists.The tightand compelling logic of the propertytheory of class has remainedan importantfactor in itswidespread survivaland its powerful social in uences. Now,EOW beginswith and from the propertyparadigm as hisMarxist inheritance.However, he presentsthat paradigm with what he sees as the majorproblem of corporate managers. Since managers usually lack ownershipof means ofproduction and therefore must sell theirlabor powerto capitalists, the propertytheory of class would have toplace them inthe classof workers, the proletariat.In short,the choicesthey have (orlack) derivefrom the propertydistributed (or not) to them. Their classposition, as determined by property distribution in what EOW takes tobe Marxism’s classtheory, locates them asexploited proletarians. EOW Žndsthis inadequate. Managers, he argues,display a characteristicthat underminestheir designation as proletarians. This characteristic is power. Even thoughcapitalists control managers, the formerdelegate authority tothe latterto ensure that capitalists acquire the laboreffort of workers. Because ofthe powerdelegated to them insidecapitalist corporations, managersare, for EOW, fundamentallyunlike workers.Their authority endowsthem witha “privilegedposition” there. They obtain,he explains, ahigherincome than otherworkers because they canappropriate part ofthe surplusproduced by those other workers (1997b: 16-17). EOW calls thisa formof exploitation, a “loyalty rent”that he deŽnes asthe differencebetween their received salaries and the value oftheir labor power.He concludesthat managers both exploit other workers and also areexploited by the capitalistswho hire them. They thereforeoccupy a“contradictorylocation within class relations” (1997b: 16). Across his writings,EOW elaboratesthe insighthe has gainedin this theorization of managersand also extends itto other groups who occupy a “privileged appropriationlocation within exploitation relations” (1997a: 23). In theoreticalterms, EOW rejectsthe singularproperty concept of classthat he attributesto traditional Marxism and substitutes a composite conceptin which property still plays akey rolebut which also includes powerand income as criteria. The rationalethat he offersfor this theoreticalchange isempiricist. That is, he arguesthat the managerial realitieshe has observedrequire just the particulartheoretical change his workaccomplishes. He believes thathe has madethe theoreticalchange necessitatedby empirical fact. EOWhas nodoubt fashioned a consistentcomposite concept of class anddeployed it for interesting empirical studies. His is one of the many possiblecomposites outlined above. However, it is not, as he claims, ACritiqueof Erik Olin Wrightand Beyond 17 ² necessitatedby the empiricalobservations he refersto. Proponents of alternative conceptsof class, singular and composite, can accommodate all ofEOW’ s observedfacts about managers and his other “ privileged groups”while theorizing them very differently.For example, asingular propertytheorist could readily argue against EOW. Such a theoristcould claimthat managers (or perhaps only the highermanagers) have hadthe capitalists’property temporarily delegated to them forspeciŽ c, exploitative functions.As agents ofproperty in the means ofproduction, they belong quiteproperly in the classof capitalist owners. Readily spinningan analysis ofhow empirical observation of managerial behavior conŽ rms this class designation,a traditionalproperty theorist of class could logically and consistentlyreject any need todepart from the propertyparadigm. While wewill shortly explore another alternative toboth the singularproperty theoryof class and EOW’ s composite– oneequally capableof producinga consistentanalysis ofmanagers– itis worthŽ rsttreating EOWs composite insomewhat more detail. In ourjudgment, EOW didmore than merely addpower and income to propertyto fashion his particular composite concept of class. His composite ratherplaces power at its center and adds property and income. His interestin Max Weber islikely bothcause and effect of this power focus. Like many othersuch attempts to marry the unwillingpartners, Marx andWeber, EOW’ s produceda compositeconcept of class in which the powerwielded by social groups takes centerstage inconstituting them asclasses. FromMarx, EOW takes the notionof property as of crucial importanceand also the conceptsof relations of production, exploitation, laborpower, and so on. He beginswith the distributionof property as akindof framework, but then the Weberianmoves take over.Whereas propertybegins the analysis, powercompletes it and so dominates the Žnal picture.Managers and others like them holddistinct class positions fundamentallydifferent from those of capitalists and workers because of the particularconŽ gurations of power that they wieldand, derivatively, the incomesthey cantherefore command. Thereare high stakes, analytical andpolitical, in the contestbetween asingularproperty concept of class and EOW’ s power-focusedcomposite. In ourexample, the singularproperty theorist renders managers mere adjunctsto capitalists, their agents, and thus highly unlikely tobe susceptibleto anti-capitalist politics, let alone ever toally withworkers in suchpolitics. In contrast,for EOW, the uniqueclass position assigned to managers– andothers privileged like them –makes theirpolitical loyalties andhence an anti-capitalistpolitics much more open questions. EOW’ s compositeconception of class bears the imprintof the socialmovements ofthe 1960sand 1970s. It deserves specialattention because it is one 18 Resnick &Wolff ² ofthe mostproductive and lasting of the theoreticalproducts of that timein the US.Indeed, EOW’ s workrepresents one of the prevailing conceptualizationof class at least onthe very broadlydeŽ ned left ofthe analytical spectrum.Its focus on powerwhile including property, its further inclusionof income and its openness to still furthercriteria have founda sympatheticaudience for the resultingkinds of classanalyses. However,it is not the casethat contemporary realities necessitate EOW’s particularconception of class, any morethan they necessitate any ofthe many alternatives toit. EOW disregardedthe claimof traditional propertytheorists of classwhen they claimedthat “ reality”warranted only theirtheory. He observedreality differently and theorized differently; his observationsdepended upon and shaped his theory just as his theory dependedupon and shaped his observations. EOW’ s theoryis no more Žnal thanthose he disagreedwith; he canno more claim “ the”match oftheory to reality than thoseothers can. How EOW observesreality isnot how others do or must do. His theory is one way to think aboutclass and apply it to social analysis; there areothers. What is excludedin his approach is included in others and vice-versa. What they stress,he marginalizes,and vice versa.The resultis a set ofalternative classanalyses generatingalternative socialunderstandings and informing alternative politicalprojects on the left andbeyond. Todramatize this last point,we propose to conclude this introduction witha briefsketch ofan alternative andthen contrastit with EOW’ s notionof class exploitation. It is the kindof class analysis thatwe have foundmost persuasive and provocative. We alsobelieve thatits neglect or marginalizationwithin the classanalytical traditionhas hadvery negative results,both theoretical and political. We willrefer to someof them briey below.The subsequentarticles in this special section will then allowreaders tobegin to form judgments about this kind of class analysis ascompared withothers such as EOW’ s. We Žrstreturn one last timeto the issueof managers so important toEOW. Then weturn to his notion and rejection of Marxian value andsurplus value theory.Our reading of Marx Žndsthat his analysis ofthe production,appropriation, and distribution of surplus enables a well-developedclass analysis ofmanagers (Resnick andWolff 1987: 128- 129,174-176). It differs sharply from EOW’ s treatment,both in analytic contentand in the implicationsof that content. In the surplusapproach, the employees ofa capitalistindustrial corporation are divided into twodifferent groups according to their relationship to the production, appropriation,and distribution of the surplusvalue producedin that corporation.Marx calledthe Žrstgroup “ productive”because it is their laborthat generates the surplus,the differencebetween the value that ACritiqueof Erik Olin Wrightand Beyond 19 ² theirlabor adds and the value ofthe wagespaid to them. The capitalists arethose who appropriate that surplus (in moderncapitalist corporations, the boardof directors performs that appropriation). Those who enlarge theircapital in this way – byappropriating surplus in production – aredeŽ ned as exploiters: exploitation is the directappropriation of a surplusfrom its producers. However, Marx wentto pains to point out thatcertain conditions had to be in place for any groupof productive workers’labor to generate surplusesfor the capitalistsemploying them. Forexamples, (1)they hadto work hard and diligently, (2) the right quantitiesand qualities of raw materials had to be continuously available, (3)threats from capitalist competitors had to be monitored and thwarted, andso on. To secure such “ conditionsof existence” of surplus production bythe productiveworkers required the laborof other workers such as supervisorymanagers, purchasing managers, research and development managers,lawyers, and so on. Marx’ s label of“ unproductive”aptly distinguishedworkers and labor aimed to secure conditions of existence fromthe surplus-producinglabor of productive workers. Capital (especially volume3) systematically detailshow and why capitalistswho appropriate the surplusfrom productive workers must then distributeportions of it to the unproductiveworkers (as theirsalaries and operating budgets). The subtletyof Marx’ s analysis lies inhis speciŽ cation of how productive and unproductivelabor depend on each other.The production of surplus (by productivelaborers) presupposes and requires the distributionof surplus (tounproductive laborers), and vice-versa. Aclassanalysis basedon surplus – itsproduction, appropriation, anddistribution – locatesindividuals and groups in relation to those three aspectsof the economy(rather than, say, intheir relations to the ownershipof assets, or to powers they wieldover others, to incomes they earn,or to their social consciousness). Thus, unproductive laborers do notexploit productive workers since they arenot in the classposition todo so; the capitalistsare in that position. Managers receive fromthe exploitersa distributedshare of the appropriatedsurplus to manage, i.e. tosecure certain conditions of existence forexploitation. In sucha class analysis, the salariesof managers, whether above or below the wages ofproductive workers, do not determine their class position, since that israther determined by their situation in the structureof producing, appropriating,and distributing surplus. Likewise the greateror lesser powersthey wielddo not determine their class position. Rather the manager’s classsituation is one of receiving a distributedshare of the surplusappropriated (exploited) by someone else. Finally, whetherthe managerowns property or not, how large that ownership might be, and whatconsciousness reigns in his/hermind do not determine class position 20 Resnick &Wolff ² anddo not therefore qualify managers as exploiters. Like power,the distributionof property and particular consciousness are different aspects ofsocial life fromthe classaspects; the formerrefer to phenomena other than the production,appropriation, and distribution of surplus, which iswhat class refers to. Of course, class aspects interact with non-class aspects;power, property and consciousness interact with class. But that is nowarrant for collapsing these differencesinto identities. Indeed, analysis warrantsprecisely keeping them distinctso as to investigate theirdifferent interactionsin differentsocial settings. Ourclass deŽ nition is singular, whereas EOW’ s iscomposite.We offer an argumentto explain andjustify the singularityof our deŽ nition. In ourview, Marx usedboth multiple singular and also some composite conceptualizationsof class across his writings. However, central to much ofhismature work, especially in Capital,wasa focusupon the production, appropriation,and distribution of surplus. Marx’ s surplusdeŽ nition of classadded something new and different to the mucholder tradition ofdeŽ ning class rather in property and/ orpower terms, even asMarx sometimeslapsed back into that tradition. Yet, that original surplus aspect waseither ignored by other class analysts orhas sincelargely fadedfrom classanalyses, thoseundertaken by Marxists and also by others. Many contemporaryclass analytical projectsignore the surplusdeŽ nition, while afewmention it in passing but usually relegate itto a marginalstatus, a phenomenonderivative of more central components of class(e.g., property orpower) and thus warranting minimal attention. In hiscomprehensive book, Class Counts ,EOW,for example, explains why he thinks the Marxian conceptof class exploitation as the appropri- ationof surplus value isinadequate (1997a: 10, 14-19). His rejection of Marx’s value andsurplus value theoryleads him to a deŽnition of class exploitationin termsof capitalists “ extractionand appropriation of effort” fromworkers (16). While dispensingwith that value theory,he doesclaim thatthis extracted labor effort produces a surplusthat capitalists appropri- ate.What he jettisonsthen isthe notionthat “ the value ofthose products isexclusively determinedby labor effort as claimed in the labortheory ofvalue” and “ thatthe appropriatemetric for the surplusis labor time” (10,ft.#11). Elsewhere, he worriesthat the conceptof surplus value is “ambiguous”(and, he adds,“ openended and arbitrary” ) unless the “costs ofproducing and reproducing labor power” are “ clear”(14, 15). Two providedexamples illustratethis worry, and he consequentlysettles fora conceptof exploitation in terms of workers’ labor effort (14-16). EOWcertainlyhas the rightto questionand reject Marx’ s value theory. He hardlyis alone inspurning that theory. What wequestion is his proceedingas if his is some summary comprehensive statement aboutall ACritiqueof Erik Olin Wrightand Beyond 21 ² labortheories of value orsome imaginary absolute, singular labor theory ofvalue. Thereis no singular labor theory of value. Endless debateshave alteredand differently interpreted at least Smith,Ricardo, and Marx’ s formulationsthereby producinga wholerange of labor theories of value. EOW’s notionis, for us, a strangeone in which no mention is made ofMarx’ s notionsof abstract labor which he distinguishedfrom concrete labor,exchange value whichhe distinguishedfrom price, or howand why the magnitudesof values andprices may interactand change becauseof changedeconomic, political, or cultural environments. He ignoresall of thisand more and hence forus the Marxian labortheory of value he rejectsis acaricatureof what the vast literatureon thatsubject contains. Heseems incapableof accepting and handling a notionof ahistorically contingent,ever-changing andalways-contested level ofthe costsof producingand reproducing labor power. That variation is precisely one ofthe majorcontributions of Marx: the ideathat the value oflabor power(and hence surplusvalue) dependson the swirlof social changes inthe society.Included in the latterare capitalism’ s ever-present driveto increaseproductivity and a “moraland historical element.” The former altersthe unitvalue ofwage goods and the lattermolds workers’ needs anddesires for quantities of wage goods. Hence Marx’s value andsurplus value theoryafŽ rms and its conclusions depend upon a continuallyvarying value oflabor power. That variability enables itto analyze andexplain ever-changing,concrete events ofthe day. In discardingMarx’ s value theorywhile simultaneously holding onto hisnotion of an appropriatedsurplus, statements andconclusions appear in Class Counts (1997a)that directly contradict what we understand to be Marxianeconomics: “ capitalistsmay alsowork – they performlabor of varioussorts – andthus the reproductionof their labor power should countas a ‘costof production’ ” (14,ft. 16); “ highwages can be a sourceof exploitation(15);” “the linkage betweenproduction and exchange whichis at the heartof the theoryof exploitation” (16); “ Empirical priceswithin exchange relations,therefore, can constitute a mechanism forthe appropriationand distribution of surplus to capitalists as well asto privilegedcategories of wage-earners” (16, ft. 19); “ somepetty bourgeois mightbe exploited or even beexploited through uneven exchange inthe market”(17, ft. 21). Ofcourse capitalists may labor.Marx recognizesthis but makes acarefuldistinction between those whose labor produces surplus over andabove their “ costs”and whose labor produces no surplus. Any or all capitalistsmay performunproductive labor involved, among other activities,with the laborof managing,researching, marketing, and Ž nance. Suchforms of labor may resultin handsome rewards (“ highwages” ). 22 Resnick &Wolff ²

However,Marx repeatedlyargues that those rewards are a distributed shareof an alreadyappropriated surplus. That is why he deŽnes thiskind oflabor to be unproductive (of a surplus),even thoughit helps tosecure the conditionsof existence fora surplusto be produced and reproduced. EOWseems tobe unaware that Marx carefullydistinguishes among differentkinds of capitalists, reserving the adjectiveof “ exploiting”or in hiswords “ functioning”only forthe surplus-appropriatingcapitalist. In that sole function,the capitalistperforms no labor but receives another’s unpaid labor. 9 Because EOWdeŽnes capitalistssolely asowners of property (and nowhererecognizes differences in these kindsof capitalists), he then must worryabout how to handle them when they alsolabor rather than being “merely rentiercoupon clippers” (14, ft. 16). When he empiricallyŽ nds them working,he placesthem ipsofacto into the ranksof workers in general, whichfor him means all workerswho are all equally working.He thusignores Marx’ s ownand others’ writings on productive/ unproductive differencesin a capitalistclass structure. As we discussed above, an alternative thatrecognizes these differencesends up with a very different notionof class exploitation than his. EOW’s ideathat exchange couldat least inpart be a source ofexploitation runs counter to one of the mostimportant economic contributionsof Marx. He repeatedlystressed that the appropriation ofsurplus occurs not in exchange butin production – incapitalists’ consumptionof the purchasedcommodity labor power. Exchange isa locationin which value isrealized or already created value andsurplus value areredistributed but not created. This different Marxian value theory rendersimpossible EOW’ s assertionof a link between“ empiricalprices” or“ uneven exchange”and exploitation. For one example, capitalists may usemarket power to establish a “privileged”location in which they sell commoditiesto, say, workersfor a retailvalue thatexceeds theirexchange value. This“ uneven exchange”has nothingto do with a Marxiannotion of exploitation in terms of surplus appropriation. Rather, itinvolves sellers creatingand using market power to redistribute an alreadyexisting value fromthe wagesof buying workers in this case to

9 Any individualmay occupyseveral different classpositions depending on the social processesin which he or sheparticipates. For example, an individualwho participates atone and the sametime in the processesof appropriating surplus value, managing the corporationas aCEO,and owning means of production in the form ofsharesof common stockoccupies three different classpositions. In one,he or sheis a productivecapitalist (whoexploits workers as a receiver oftheir producedsurplus value) and in the twoothers anunproductivecapitalist (who receives distributed shares of that surplusvalue in the form ofa salaryfor securing management anddividends for providing access to owned means ofproduction). ACritiqueof Erik Olin Wrightand Beyond 23 ² themselves asmonopoly-capitalists.For another example, when merchants purchasecommodities from capitalists at wholesale pricesthat are less than exchange values, merchantsdon’ t exploitcapitalists. Merchants receive a distributedshare (equal to the differencebetween the capitalists’unit value andthe merchants’paid wholesale pricetimes the unitspurchased) of the alreadyexisting surplusescapitalists have appropriatedfrom workers. Thisdistribution of value inthe Žrstexample orof surpluses in the other explains the respectiveearnings of the “privileged”locations enjoyed by the monopolyand merchant capitalists. These examples canbe extended toexplain the earningsof several differentkinds of laboring individuals including,as have noted,managers and CEOs of corporations as well as landlords,bankers, patent holders, and so forth (Resnick andWolff 1987: 124-132,174-183, 237-245). Because webelieve the lossof attention – analytically andpolitically –tothe production,appropriation, and distribution of surplus has had disastrousconsequences, we have undertakena classanalysis withclass singularlydeŽ ned in surplusterms. The absenceof attentionto the surplus conceptamong class analysts has served theoreticallyto strengthen the wholegamut of conservative social theorists who long ago committed themselves toexpunging the conceptfrom both scholarly and popular usage. 10 Classanalysts should,in our view, at least beadvocates of the salience ofsocieties’ various modes of organizing the production, appropriation,and distribution of surplus, whatever other aspects of society they wishto stress more in their work. The pointis to build from Marx’ s contributionrather than proceed as if it did not exist. Aseconomists, weare especially awareof the currentglobal hegemony ofneo-liberalism andits formal expression in economics, neoclassical theory: a hegemony foundedon the projectof eliminating all reference to,let alone analysis of,

10 Arecent work (Boss1990) exempliŽ es efforts to expunge the conceptof surplus from economics.Recognizing that Smith andMarx madeconcepts of surplus central totheir work,she dubs them “misguidedClassics” (11). She prefersmost neoclassical economics becausethey seehow all workers exist in awebof interdependence. For her, theories ofsurplus are blind to this interdependence, wrongly pursuinga surplus,gleaned from the economicallyproductive but then transformed tothe unproductive(whom shelabels “parasites”). Actually,Marx’ s notionof surplus and productive and unproductive labor containsno notion of parasitism.When he writesin Capital (Vol.2, 1967:131) that “He (the unproductivelaborer) performs anecessaryfunction : : :”hisanalysis understands surplus not asopposed to interdependence, but rather asa wayto grasp critically the particularly capitalistmode of interdependence and its social consequences. See also our discussion of Marx’s distinctionbetween productive and unproductive labor in Resnickand Wolff (1987: 124-141).Marx’ s goalis thereby toshow how economic interdependence could and should beorganizeddifferently, i.e., by organizingthe production,appropriation, and distribution ofthe surplusin non-capitalistways. 24 Resnick &Wolff ² the production,appropriation, and distribution of the surplusgenerated in production. Politically,the evaporationof attention to class in its surplus sense has likewise, inour view, contributed to disastrous social consequences. For example, socialistrevolutions premised on class analyses (oftenMarxist) have taken moreor less radicalsteps to equalizeproperty and power and to altersocial consciousness in collectivist directions. They didthat believing class,class struggle, and class revolution to be mattersof property, power, andconsciousness, whatever the differentweights they assignedto these componentsof their composite deŽ nitions of class. They rarelyworried much,if at all, aboutthe production,appropriation, and distribution of surplus.They rarelychanged them radically.If an enterprisehad ceased tobe private property and had become property of the statein the name ofthe wholepeople, might the organizationof the surpluswithin thatenterprise have remainedlittle changed?Might state ofŽ cials have merely replaceda privateboard of directors in an unchangedclass-qua- surplussituation? And might that unchanged class-qua-surplus structure have underminedthe viabilityor durability of the changes inproperty, power,and consciousness that the revolutionhad achieved? Suchquestions couldnot be asked oranswered or acted upon in places like the USSR, China,and Cuba – amongothers – becausesurplus had disappeared from orbeen marginalizedwithin the hegemonicdeŽ nitions of class in those countriesand across the left elsewhere aswell. One ofthe key lessons thatcan and should be learned from the collapseof the USSRas well as fromrecent changes inChinaand troubled socialist projects in many other placesis that the costof abandoning Marx’ s insightsabout the production, appropriation,and distribution of surplus far exceeds any beneŽts it may have accrued.Hence ourproject of bringing class-qua-surplus into the frameworkof class analyses andthe adjunctproject of criticizing the claimsof class and other analysts whoexclude ormarginalize the surplus deŽ nition. 11 Ofcourse,all classanalyses recognizethat class is notthe only aspectof societyor of individual identity; non-class aspects are always part of class analyses. What distinguishesclass analyses fromother kinds of analysis is thatthey highlightthe classaspects, foregrounding them togain a wider appreciationof their importance to understanding and changing society. Andwhat distinguishes class analyses fromone another – forexample, EOW’s andours – isprecisely what they include,exclude, and marginalize withintheir deŽ nitions of class. All class analyses recognizeand treat many

11 See,for example, the recently publishedcollection of essays on class conceived in surpluslabor terms in Gibson-Graham, Resnick,and Wolff (2000, 2001). ACritiqueof Erik Olin Wrightand Beyond 25 ² aspectsof society – cultural,political, and economic – butthey doso fromthe standpointof those particular aspects included in their respective deŽnitions of class. EOW’s particularkind of class analysis doeswhat any classanalysis must:it highlights some issues and throws others more or less intothe shadowsof its concerns. EOW highlightscontrol (power distribution) primarilyand property and income secondarily. EOW then goeson to connectthese focito many other,non-class aspects of the societieshe analyzes. In contrast,we highlight the production,appropriation, and distributionof surplus and connect them toa hostof other aspects of the societieswe investigate andseek tochange. The articlesthat follow areexamples froma growingliterature built around the surplusconcept ofclass (Resnick andWolff 1986, 1987; Gibson-Graham 1996; Gibson- Graham,Resnick, andWolff 2000, 2001). We ask readersfor two things: (1)do not allow any particularclass analysis toclaimitself as the only one thatcorresponds to reality, and (2) compare the resultsof different class analyses interms of what they reveal andwhat they occlude.Perhaps the next stepis to rethink andreconŽ gure what needs tobe included in the classanalyses thatwill shape the twenty-Žrst century, given the lessons of the twentieth.

References

BARAN, PAUL AND PAUL SWEEZY 1968 Monopoly Capital. New York:Monthly ReviewPress. BERLE, ADOLF A. 1959 Power Without Property .New York:Harcourt, Brace andCompany. BOSS, HELEN 1990 Theories of Surplus andTransfer: Parasites andProducers in Economic Thought. Boston: Unwin Hyman. BUKHARIN, NIKOLAI AND EVGENY PREOBRAZHENSKY 1988 TheABC of Communism. Ann Arbor:The University ofMichiganPress. CALLI NICOS, ALEX 1982 Is Therea Future for Marxism? Atlantic Highlands,NJ: HumanitiesPress. CARCHEDI, CUGLIELMO 1991 Frontiers of Political Economy. Londonand NewYork: Verso. COHEN, G.A. 1978 Karl Marx’s Theory of History. Princeton: Princeton University Press. CUTLER, ANTHONY, BARRY HINDESS, PAUL HIRST AND ATHAR HUSSAIN 1977 Marx’s Capitaland Capitalism Today. Vol.1. London and Boston: Routledge & Kegan Paul. DAHRENDORF, RALF 1959 Class andClass Conict in anIndustrial Society. Londonand Boston: Routledge & Kegan Paul. 26 Resnick &Wolff ²

DE STE. CROIX, G.E.M. 1981 TheClass Strugglein theAncient GreekWorld. Ithaca,N.Y.: Cornell University Press. ELS TER, JON 1986 MakingSense of Marx .Cambridge:Cambridge University Press. ENGELS, FREDERICK 1976 Anti-Duhring. New York:International Publishers. GIBSON-GRAHAM, J.K. 1996 TheEnd of Capitalism (AsWe Knew It).Oxford: Blackwell. GIBSON-GRAHAM, J.K., S. RESNICK AND R. WOLFF (EDS.) 2000 Class andIts Others. Minneapolisand London: University ofMinnesotaPress. 2001 Re/Presenting Class. Durham andLondon: Duke University Press. GIDDENS, ANTHONY 1973 TheClass Structure of theAdvanced Societies. New York:Harper andRow. HIRST, PAUL 1979 OnLaw and Ideology. Atlantic Highlands,N.J.: Humanities Press. HOWARD, M.C. AND J.E. KING 1985 ThePolitical Economy of Marx. SecondEd. London and New York:Longman. JESSOP, BOB 1980 “The Politicsof Indeterminacy ofDemocracy” in Alan Hunt (ed.), Marxism andDemocracy. London:Lawrence and Wishart. KAUTSKY, KARL 1971 TheClass Struggle. New York:W.W. Norton andCo, Inc. KUCZYNSK I, JURGEN 1971 TheRise of theWorking Class. New York andToronto: McGraw-Hill BookCo. LACLAU , ERNESTO 1977 Politics andIdeology in Marxist Theory. London:New LeftBooks. LANGE, OSCAR 1963 Political Economy. Vol.1, trans. byA.H.Walker. New York:Macmillan. LENIN, V.I. 1972 TheDevelopment of Capitalismin Russia. CollectedWorks, Vol. 3. Moscow: ProgressPublishers. MARX, KARL 1967 Capital. Vol.2. New York:International Publishers. MILLS, C. WRIGHT 1956 ThePower Elite. New York:Oxford University Press. 1962 TheMarxists. New York:Dell. MOSCA, GAETANO 1939 TheRuling Class. Trans. byHannah D.Kahn. New York:McGraw-Hill. NORTON, BRUCE 2001 “ReadingMarx forClass” in J.K.Gibson-Graham, S.Resnick,and R. Wolff (eds.), Re/Presenting Class. Durham andLondon: Duke University Press.23-55. POULANTZAS, NICOS 1978 Classes in Contemporary Capitalism. London:Verso. PRZEWORSKI, ADAM 1987 Capitalismand Social Democracy. Cambridge:Cambridge University Press. RESNICK, STEPHEN AND RICHARD WOLFF 1986 “Power,Property andClass.” Socialist Review 86, 97-124. 1987 Knowledge andClass. Chicagoand London: The University ofChicago Press. ACritiqueof Erik Olin Wrightand Beyond 27 ²

1992 “Replyto Richard Peet.” Antipode 24(2),131-140. 1994 “Rethinking Complexity in EconomicTheory: The Challenge ofOverde- termination.”In Richard England(ed.), Evolutionary Concepts in Contemporary Economics. Ann Arbor:The University ofMichigan Press. ROEMER, JOHN E. 1988 Freeto Lose. Cambridge:Harvard University Press. STALIN, J.S. 1954 Problems of Leninism. Moscow:Foreign Languages Publishing House. SWEEZY, PAUL 1964 TheTheory of CapitalistDevelopment. New York:Monthly ReviewPress. THOMPSON, E.P. 1968 TheMaking of theEnglish Working Class. Revisededition. Harmondsworth: Penguin. TROTSKY, LEON 1987 TheRevolution Betrayed. New York,London, and Sydney: PathŽ nder Press. WEBER, MAX 1922 Wirtschaftund Gesellschaft, Parts IIandIII. Tubingen. WRIGHT, ERIK OLIN 1979 Class, Crisis andthe State. London:Verso. 1985 Classes. Londonand New York:Verso. 1997a Class Counts. Cambridge:Cambridge University Press. 1997b Class Counts, StudentEdition .Cambridge:Cambridge University Press.