Erik Olin Wright was born in Berkeley California in 1947 while his father was in medical school on the G.I Bl He grew up in Lawrence, Kans, where both of his parents were professors of psychology a the University of Kansas. Aside from ayeatspen in Austealia in 1960, he remained in Kansas until 1964 when he entered Harvard University. After graduating with a B.A. degree in socal studies in 1968, he spent two years studying history at Oxford University, working especially with Christopher Fil ‘Upon returning tothe United States in 1970, he spent one year in 1 Unitarian Theological Seminar in Berkeley, during which time he worked at San Quentin Prison asa student chaplin. On the basis ‘of material gathered while working inthe prison, he wrote The Polies of Punishment: critical analysis of prisons in America (Sew York: Harper Colophon Books, 1973). In 1971 he entered {graduate school in sociology atthe University of California where he stayed until completing his doctoral dissertation on Clas Src ‘ure and Income Inequality inthe spring of 1976. Author of Class, Crisis and the State (New Lett Books, 1980), he is currently Pro fessor of Sociology atthe University of Wisconsin, Madison VvErik Olin Wright | Classes Verso | Suesrsh Library (Cataloguing In Publication Dats ‘Wright, ik tin Class 1'Secn cases Boss Tao Fics published 1988 (oF On Wright Vero Esitons 15 Greek Steet London WIV SL “Typeset in Times Roman by Spire Pant Services Lt Susbury, Witaire Printed in Gret Brin by ‘The Thetford Pres Ld, ‘Thetford, Norte say 0.86091-1087, (0-86091-8122 pbk Contents Preface 1 1. Posing the Problem: The Agenda of Class Analysis The Legocy of Mars 6 The Agenda of Contemporary MarisiClass Analysis 8 ‘Part One: Conceptual Issues ” 2, TheBiography of a Concept: Contradictory Class Locations 19 ‘The Lopicaf Concept Formation 2» Steps in the Anayss of the Formation of the Concept ‘of Contradictory Locations 4 The Empirical Seng 2 Theoreticl Consirains 26 Alternative Solutions a” Buldng New Concept 2 Problems with Conceptuaiztion 3 3, AGeneral Framework for the Analysis of Cass 6 ‘Roemer’s Account of Class and Exploltation 6 Towardsa General Framework of Class Analysis 2B Once Again, Unresolved Problems 2 4, Implications and Elaborations ofthe General Framework 108 “Alternative Class Theories 106 ‘Mode of Production and Socal Formation 105 The Theory of History 14 Legtimation and Motivation us (lass Structure and Form ofthe State m (Glass Structure and Class Formation reat (Cass Ainces maWomen in the Clas Structure Conctusion 5. Empirically Adjudicating Contending Cass Definitions ‘Operationaliing the Adjudicatons Reformulaion of Hypotheses Note Statstleal Procedures Empirical Results Conclusion 6, Class Structure in Contemporary Capitalism: ‘A Comparison of Sweden and theUnited States (Cass Distributions Explining Differencesin Class Siructures CClassand Family (Cass Structureand Income lass Structure and Class Consciousness in Contemporary Capitalist Society Whats Clase Consciousness? Causal Logie Operationatizations ‘A Noteon Statistica! Procedures Empirical Results Conchsions Conclusion The Exploitation Centred Concept of Class The Clas Structure of Contemporary Capitalism Cas Structure and Politics Politial Implications Appendix: Practical Strategies for Transforming Concepts “Appendix I: Variable Constructions “Appendix Il: Full Data for Selected Tables Bibliography Index 136 130 135 136 I 17 159 163 181 92 198, 210 as 22 To the life of M. Erik Wright, my father.Preface uate faeho cas know why ty dongs wah yo “ara 198) | began work on this book nearly five years ago fully anticipating that would finish it within a year or So. In the course of these ets several important things have occurred in my life which have iad a major impact on the content and schedule of the project. "To begin with, Thave had a significant change in what socolog- ists call my “efeence group’, the circle of people whose opinions fnd evaluations are in the back of my mind as I type away on my ‘word processor. My earlier work on clas structure, the state, income inequality and elated topics was all basically writen or at least launched while I was a praduate student at the University of CGlitornia i the first part of the 1970s. Up until about 1980 most ‘of my published material was either iitilly formulated in my dent years or developed asa direct spin-off ofthat period IMy relerence group while @ graduate student was a circle of Marxist scholars affiated with the joureal Keptalitate and lore organization called the "Union of Marxist Social Scientist ‘Most of these people were students, most had been radicalized ‘uring the heyday of the cv igh and anti-war movements ofthe 1960, and most were committed to some varity of Marxist approach to socal theory. While many of us considered ourseives {be rather unorthodox in various ways, the basic categories of ‘Marxist analysis, fom the labour theory of value to the theory of| the capitalist state, were more of les taken for granted s points of| ‘eparture. There was 8 great deal of fervour and excitement and ‘we al fel that we had 2 firm pip on the truth “As itis commonplace to say, times have changed. Many ofthe stodents who engaged inthe fevitalization of American Marxism inthe 1970s have subsequently een employed in professional and ‘eademic posts, and @ good number ofthe academics have by RoW Feceived tenure. The fesling of assurance that We had answers tO 12 every question has generally been tempered by 2 more cautious ‘and nuanced stance. In many cases, infact, Marxism has become the object of considerable eritcism onthe academic left, and many radical scholars are identifying with what some have labelled ‘post Marxism’ ‘On a persona level, I became an Assistant Professor and then a tenured faculty member at the University of Wisconsin. And [ have also Become more aware of the problems in Marxist theory and the need fora more rigorous and reflective approach. But f Ihave not, Thope, sifted my basic commitment to the project of “Marist theory and to the fundamental insights contained sithin ‘To sustain that commitment I helped to establish a graduate training programme in the Wisconsin Sociology Department, the (Clas Analysis and Historical Change Program. That programme in tan, has become a crucial clement in my new reference group, Unlike’ my student circle in Berkeley, the Class Analysis pro- gramme in Madison is ideologically much more diverse and eer finly less wedded to a traditional Markst perspective. As teacher inthe programme, therefore, have had to defend actively the core theses of Marsism and male them compelling #0 a sym patheti yet unconvinced audience, In the eourse of doing 50, pat fieuarty in the context of arguing with energetc students in yearlong course on the Theory and. Methodology of Marxist Social Science which I regulary teach, I have questioned, clarified and reformulated many of the basic ideas that I had earlier taken for granted. ‘My role a a professor constitutes only one aspect ofthis change in reference group. Perhaps even more important forthe specific illectual direction which my work is now taking Ihave become ‘very involved witha group of leftist scholars of varying degrees of ‘Sympathy to Martism who mest once a year to discuss one smothers work. Ths group includes G. A. Cohen, John Roemer, Jon Elster, Philippe van Paris, Robert van der Veen, Robert Brenner, Adam Preeworski and Hillel Steiner. The central intel Jeeta thivad ofthe group is what they term "Analytical Marxism’, by which is meant the systematic interrogation and clarification of bisic concepts and their reconstruction into a more cobereat theoretical structure. The discussions within this group and the texporure it has given me to & range of new ideas and perspectives have had a considerable impact on my thinking end on my work. I these reference groups define the postive forces 1 have encountered in the formulation of new idess, other especs of my Preface 3 ‘current situation constitute negative pressures. In the transition from graduate siden to tenuted professor Ihave also become integrated into a-newus of rewards that i very alluring. My esearch on class has led 10a series of large research grants which pay pats of my salary and allow me to take te off from teaching fowrite. As my reputation has grown, [have had numerous oppor tunities for travel snd lecturing in varios places around the world ‘And T have been handsomely rewarded by my Sociology Depart ‘ment andthe University of Wisconsin. Asa Marxist materialist and ‘lass analyst, cannot suppose that al ofthis has no effect on me 8nd that by an act of will 1 can immunize myself from the seduc tions ofthe safe and comfortable life ofan affluent academic ins Hineral-democratc advanced capitalist society "The privileges bestowed by elite universities have, with good reason, made many radicals suspicious of "academic Marnsts Soch suspicion may be particularly acute in the United States, where the absence of a cohesive, mass socialist movement, let tone a revolutionary working-class poiial party, has made it ‘ical for many academic Marxists tobe systematically inked to socialist strugels ona day-to-day bass. Certainly in my ovn ease, TThave not been a polal activist in resent Years. While my work has been informed by socal and polis! events, it has not been forged in direct engagement with popular struggle. Tido not know the ways in which the ideas elaborated in this book have been shaped by these institutional and political realities and choices. I do not even really know whether of aot, in the Dretent historical circumstances, the Work has benefited’ or suf- fered from the particular conditions under which it was produced ‘The time, travel and intellectual stoulation that my present posi- tion gives me may expand the space for critesl thought more than the privileges enjoy erode it What I 60 know is that | have been fare of these issues and T have tried to maintain the ind of fellrefletive stance that might minimize the negative effects of ‘hese material conditions on my Work ‘Aside from these various professional considerations, my life has “undergone one other massive change since I fist began work on this book: the birth of my two daughters, Jennifer and Rebecca, pow aged five and four. I do not know if my theoretical sen ed by the wonderful transformation these two litle persons have Brought to my ie, but Tam certain thatthe4 book would have been finished couple of years eatler if Thad ‘not embraced the joy of liberated fatherhood In the course of writing this book I have received considerable fcednack on specie chapters and arguments from a large number ‘of people, I am particularly grateful to Andrew Levine, Who ltiempted, with some succes, to delay the completion ofthe manu Script by giving me too many difficult comments. Michael Burswoy was very important in helping me to clarity the inital ‘agenda ofthe book during the exciting Year be spent in Madison ‘The arguments in the book have also benefited decisively from & ‘series of comments and discussions Ihave had with Jona Roemer. Robert Manchin, a Hungarian sociologist who spent a Year a the University of Wisconsin, contributed grealy to working through the ideas embodied in chapter three, am aso grateful for writen comments from Adam Przeworski, Goran’ Thetbor, Petry “Anderson, Danie! Bertaux, Ron Amiazade, Richard Lachmann, Philippe van Paris, Robert van der Veen, Trond Petorsen and Sheldon Stryker, and for stimulating discussions of the issues raised inthe book with Ivan Szeleny, Jon Elster, G. A. Cohen, Goran Ahrae and the many students in my courses and seminars ‘who have. pressed me continually on these problems. Various technical issues in the empirical chapters were clarified by Charles Halaby, Robert Hauser, Rob Mare and Tom Colbjorson. T would Tike to thank the research team that worked on the cas structure projecr-especially Kathleen “Cains, Cynthia Costello, David Haachen, Bill Martin and Joey Sprague—for the enormous eon tribution they have made to the empirical investigations in the book. To my wife, Marcia Kahn Wright, I owe a special debt of sratitude for not letting me get too obsessed with my work and helping me to keep things in perspective, Finally, 1 woul lke to acknowledge the financial support for the research and writing that has gone into this project from the National Science Founda tion, The German Marshall Fund of the United States and the ‘Wisconsin Alumni Research Foundation, ‘While this book was being written, four people whom Toved have died. My grandmother, Sonia Posner, whose love of learning and lifelong commitment to revolutionary ideals deeply shaped my life, died inthe spring of 1980. Luca Peroane, whose comradeship and briliance helped me begin my fist attempts at cass analysis, ied later that year. My father, M, Erk Wright, whose nurturance Preface 5 and vitality and curiosity will slays be with me, died in 1981. And Gene Havens, a companero and colleague who showed me how to be un academic and a serious Marxist, died just before the book ‘eas finished inthe summer of 1988. To the memory of these four dedicate this book Erik Olin Wright Madison, Wisconsin ‘November 1984Posing the Problem The Agenda of Class Analysis, ‘The Legacy of Marx As has froquenlly been remarked and bemoaned, Marx never sy tematially defined and elaborated the concept of class, in spite of the centrality of that concept in his Work. To the perpetual rustra- tion af people who seek in the texis of Marx authoritative answers to theoretical problems, in the one place where he promises such fn elaboration the inal chapter of Capi Volume 3, entitled “Classee™—the text slops after only page. Just before the end of| this incomplete text Marx wrote, ‘The frst question to be ans- Wwered i thiss What constitutes & class?. Two short paragraphs later comes Engels's sad comment, “Here the manuscript breaks off, Wihile Marx never systematically answered this question, his work is filled with clas analysis. With some exceptions, most of| this work revolves around two problems: the elaboration of ab tact structural maps of clas relations, and the analysis of concrete Conjunctural maps of classes a-ators The fst of these kinds of fnalyses concerns the Way in which the socal organization of pro- duction determines a structure of empty places in class relations, places filled by people. This structural anslyss of classes is found Partcuariy in Marx's most celebrated theoretical works, espe- ally in Capital where he decodes the structure and dynamics of| the capitalist mode of production. The second kind of analysis, on the other hand, is not concerned wih elas structure as such, but ‘with the ways in which the people within cass structures become brganized into colleciviies engaged in stuggle- This analysis of ‘ase formation is found most notably in Mary's political and his- torical writing, where Marx i ying 10 understand the interplay of 6 Posing the Problem: The Agenda of Class Analysis 7 collectively organized social forces in explaining specific historical ‘ansformations “The images that emerge from these two sorts of accounts are quite different. From the abstract structural aecount of classes omes the characteristially polarized map of cass relations which Furs through most of Matx’s analysis of the capitalist mode of Production in Capial and much of his more abstract discussion of pochal trajectories of historical development: masters and slaves, lords’ and serfs, bourgeoisie and proletariat. ‘While non-polar. ized postions ‘are occasional relerred to in. these. abstract lscussions of class relations, they are never given a rigorous Theoretical status and are. generally treated a having strictly peripheral importance, ‘In contrast to this simple, polarized, abstract map of class rele sions, Mars conjuncturalpolitial analyses are characterized by 8 complex picture of clases, factions, factions, social categories, Strata and other actors on the political stage In the Eighteenth ‘Brumaire of Louis Bonaparte, for example, he refers to atleast the following actors in soil conflicts: hourgeoisie, proletariat, large landowners, artstocracy of finanee, peasants, petty bourgeoisie, the middleclass, the Tumpen-proeiaria, industrial bourgeoisie, high ‘igaitaris. No attempt is made by’ Mars to presenta sus. preoccupation inthis text is with understanding the relationship ‘Between the steuggles among these actors snd the state. In particw- lar he tries to explain the patterns of victories and defeats in these struggles, the effects of those vietores and defeats on changes in the state, and the effec of changing regimes on the patern of alliances and struggles among these actors. He is not concerned With elaborating a rigorous map of the concrete sock! structure inhabited by the protagonists inthe drama, Thisis characteristic of ‘Mars politial-conjunctural writings. While he gives uss itt of descripive categories correspending to the actual acors in the confit, he doesnot provide a set of precise concepts for decode Ing rigorously the structural bass of most of those categorie, ‘What we have then, in Mar’s own work, isa polarized abstract ‘concept of the ‘empty places" generated by las relations and descriptively complex map of conerete actors within class strug fles, with no systomati linkage between the two. Marx of course fei hat the historical tendency of capitalism was towards inereas- ing concrete polarization. ‘Socety 28. whole,” he ‘wrote With5 SEL cide cee ly aaghenescare Saag cy eed th ae oa Sentiewineerarene imate cerca ip pear inayimaeestiorcomg iter sine ‘own vr mst ey Bupa ten SP ara aor ghee Pastime eirgepotee geaerete nica tiataa aerate sae ae ‘The Agenda of Contemporary Marrst Class Analysis ‘The historical record of the past hundred years has convinced ‘many Marxists that this image of a pervasive tendency towards fadical polarization of class relations within capitalist societies is incorrect. To be sure, there has been a steady decline in the pro: Portion of the population owning thei own means of produc: tion the se-employed in advanced capitalist countries at east unt the recent pas” But among wage-earness, the growth of professional and fechnial occupations andthe expansion of tan “gerial hierarchies in large corporations an the state have at least Posing the Problem: The Agenda of Class Analysis 9 “Toretical Objects and Level of Abstraction a Marxist Class Analysis Teter ‘Aiton Thera! Obj of Ants wooe or Pst Cs eons ——_Egch agheteen oer. CCoasitenseofcanestaned Cis lanes Fomancn nee maces of Tage vlpmest ts laos in Bee fobe pate Shop foot created the appearance ofa considerable erosion of simple pla Sted srecure “Given that itis nooner generally aozeped tht the cas struc ture within capitalism ncreaingl polarized thas become more Aitfcut to sidestep the theoretal problem of te pap between the abstract polarized concept of das elation and the complex onctete puters of ass formation and cess strug. Tne fongerssamed that story wl gradually eliminate the concep problem, Resolving this problem hasbeen ane ofthe ental con- Exmso thevenurgene of Marat class analysis inthe past twenty To understand the theoretical agenda of this new body of Marx- ‘xt work on das t wil be helt to distinguish formal two ‘Sienstons of class anal that have been impli i out discus Sono fr: st, whether the analysis focores primary on cas Structure or om caw formation, and second, the level of abstac ttn at wich cass are analyned. This els the sx possible fk ot class anal lltrted ia Table 1 ‘The dsunction between cas sucture and class formation i a tas olen imply dstoeio nea anaysis Cn strture {eter othe suture of socal relations nt which iavidal (or, Insome case, ames) enter which dtermin ther das interes Wel haves great desl tay about how these relations sould0 bbe detined in subsequent chapters, The point to emphasize here is that elas struture defines a set of empty places or position filed by individuals or families. This implies that with respect to class Structures we can tlk about "vacant positions (positions which are ‘ot currently filled by actual people), about an ‘absolute surples Population’ (an excess of people with respect to the places within fhe class strueture), and “incumbents” of ease postions (people actually located within a given clas structure). While this does not imply that class structure exists independently of people, it docs mean that it exists independently of the specific people who ‘occupy specific positions” ‘Css formation, on the other hand, refers tothe formation of ‘organized collstvities within that clas structure on the basis of the interests shaped by that class strutare. Class formation is 8 variable. A given type of clas stretute may be characterized by & range of possible types of class formation, varying in the extent and form of collective organization of elses, Clase-based collec tivities may be organized, disorzanied or reorganized within 8 given lass structure without there necessarily being any fundamental transformations of the clase strocture itself I class structure is defined by social relations benween classes, clase formation is defined by social relations within classes, socal relations which forge collecisiies engaged in strugle inetons among levels of abstraction of clas analysis is somewhat more complex iste, ‘Tree levels of abstraction typi characterize Marxist discourse on class: made of production, ‘social formation and conjuncture The highest level of abstraction is mode of production. Classes are here analysed in term of pure types of social relations of pro- duction, cach embodying a distinctive mechanism of explotation ‘When Marx talks above ofthe “pure form” of elasce in capitalist sosty he etering tothe ana of Claes at is highs level of abstraction, In many discussions ofthe ‘mode of production’ level of abstrac- sion its absumed that no variabiliy within a mode of production is admissale at this level of abstraction: all eapitalsme are eq tent when discussing the mode of production. Thi, 1 think, is 8 mistake, Without shifting levels of abstraction, i is sil entirely possible to define diferent forms ofa given mode of production, Indeed it has been one ofthe central themes of Marxist theories of the capitalist mode of production that this mode of production itself hasan intrinsic logic of development. This logic of develop- Posing the Problem: The Agenda of Class Analysis 11 7 ge ‘ment means that the capitalist mode of production itself has an intrinsic tendency to passthrough different ‘stages each with & distinctive form of capitalist socal relations (primitive accumula: tion, competitive captlism, monopoly capitalism, ete). Like all tendencies, of course, this tendency may be blocked by various mechanisms, and the investigation of the acwal processes which may facilitate or impede tis trajectory of forms does require mov. ing th analysis to a lower level of abstraction, But the analysis of the developmental logic of capitalist relations such must be theorized at the level of abstraction of the mode of production ial! “The term ‘social formation’ has come to derive its meaning from the analysis of societies a8 specific combinations of distinct modes of prodedon or types of relations of production.”"The analysis ofthe Presence of pre-capitalist classes. within capitalist society, and more rarely the analysis of post-captalistclasees within capitalist society, are examples of analysing class structure atthe level of Abstraction of social formation. The analysis ofthe specific ways in ‘which diferent forms of capitalist relations are combined within a {iven society fe alzo a problem of the socil formation level of Abstraction. For example, analysing the specific combination of competitive, smallscale capitalist production with large, eoncen- tated and ‘centralized capitalist production in a given society ‘would be a social formation analyse. The problem of alliances ‘between elses and fractions of casts ie the principal object of the analysis of clase formation at this level of abstraction, ‘Conjunctural analysis involves the investigation of societies in terms ofthe concrete institutional details and contingent historical {actors that enter the story." The analysis of specific forms of labour-market segmentation within the working class, o the legal practices which define the powers of managers over workers, or the creit relations that link petty bourgeois to bankers, would all be instances of conjunctural analyses of class structures. The analysis of unionization, party formation, casebased socal ‘movements, etc, would be the analysis of class formation at this mort concrete level, “The conjunctual level of analysis is also the level of abstraction at which the most sustained analyses ofthe relationship between lass and non-lass relations and practices usualy oceur (eg. class land race or clas and gender). This isnot to sy that in principal Such sues cannot be addressed at higher levels of abstraction, but the conceptual apparatus for such more abstract investigation =R rather underdeveloped and when attempts are made they tend be reductionist. For example when the gender-cas relatonship explored atthe level of mode of progustion, most marist analy. Ses ellecively end up reducing mate domination to lass domina- tion. Typically this redeton occurs in some sor of functionalist Iman the existence and form of pariarehy is explained by the ‘sential freon flim reproducing the base as lations ot eats In there forms, many debates can be interpreted as disagree- ments ove the aproprat evel of abstraction fr adesing cer= tun problems gender and class have completly contingent rela trons between thent—that the aus inereonnections between them occur simply Because they alect the same people but not because they presuppose eachother in anyway then thet rela tonship ean ony realy be analysed a the conjunctural level. tom the ote hand, there are stractral properties of these two rla- {tons which are intrnsialy elated, then a mode of production nays may Become possible. To take another example, some theorists suchas Nicos Poulanzas ave argued ha the relation- ‘hip between the form of the state ad social lasses can be analy Sd at the level of abstraction of mode of production and ths iesas him to ty 10 constructs general concept of the capitalist, Sate, Other theorist suchas hed Skocpol argue that he sate annot legitimately Be theorized at this level of abstraction {sist on sry historic! (Le. conjonctiral) investigation ofthe ‘lationship between sates snd canes” ‘An analogy may help fo clanfy the dsinetons being made be. twoon these levels of abstraction. Inthe scientific study of the hemsty of a lake, the highest level of abstraction involves ‘Spelling the particular way that the base elements hat go into Taking water, hydrogen and oxygen, combine to make water, TO. The study ofthe tlre forme of water-—ce, ligula water, tvaportion, ete—-would all be at this mos asta level THe nile level of abstraction corresponding 10 soil formation nals invohes investigating the ways So which this compound, IO, interacts with other compounds i lakes. Finally he con jucturtl evel involves iavestenting the myriad of contingent ac- {ors-ittogen washed down from farms, chemical waste dumping ‘fom factories, eto—whicn conectly distinguish @ gen ake ‘hemicaly from ll dhe lakes in time and space in terms of Table 1 the bul of Max's tales of esses is ‘onoentated ia the upper left hand cel and the 1owertWo ght Posing ihe Problem: The Agenda of Class Analysis 13 hhand cells. OF course, Marx had something to say somewhere about every eel inthe able, but he never provided a systematic theoretical exposition of the lower two levels of abstraction of class structure. Nor, as altoady stated, did he ever provide aso tained theory of the causal linkage between elas seuctire and lass formation, of the process through which position within cass Structures analysed at different levels of abstraction. become formed ino organized collecvities. ‘Much ofthe recent development of Marxist theory and research fon clases can be viewed 3s attempts a bridging the gap between the abstract analysis of class stuetute and the analysis of cass formation. This new cass analysis has had two principal thrusts fis, filing in the undertheorzed cells in the structural side of the ‘typology; and second, much more systematically analysing the problem of the translation of this strdture of rclations into the formation of collective actors The work that has focused on the problem of cass structure in advanced capitalist societies, the pre-eminent preoccupation has been with the embarrassment ofthe middle clases” The evidence ofthe existence and expansion ofthe “new middle clas’ has been 8 the heart of most critiques of Marrs class theory, and Marxists have found it necessary to respond to those ertiques in one way or ‘another, However, the concern with the middle ass, or, equiva ently, with specifying the conceptual ine of demarcation between the workingsclass and nonsworking-class wage earners, is not ‘simply a defensive response to bourgeois attacks. Resolving this ‘conceptual problem isso seen a essential ifthe clasieal concerns ‘of Marxism—understanding the development of the coatradie tions apitalism and the conditions forthe revolationary transfor: mation of eapitalist socity~are to be analysed in a rigorous Way, “The parallel problem for the structural analysis of clases in third-world capitalist societies is the "embarrasiment of the peasantry’, whith at least according 0 many earlier Marxist analy- Ses was thought tobe a cass in rapid decline. The introduction of the concept of "articulation of modes of production which attempts to give specificity to the relationship between peasants ‘workers and capitalists, and the elaboration af the world ystems approach to the study of third world societies were both important Strategies for rethinking the class structures ofthese societies." ‘As we will ce inthe next chapter the result ofthese attempts 10 salve the problem of the middle cesses and the peasanty hi been a range of alternative conceptualizations of clas structure at1“ the middle levels of abstraction. Inthe course of building these new concepts, the more abstract mode of production analysis has itself been subjected to scrutiny, and various elements to that analysis have been challenged and altered by diferent theorists The fll ramifications of these various conceptual innovations at sill working themselves ou. The second general thrust of recent work attempting to bridge the gap between the abstract analysis of class structures and the analyst of lass formation has forised on the process of class formation. The starting point of most ofthese analyses has been & firm rejection of the view that particular kinds of class formation can be deduced directly from the class structure. In ts place i the feneral view that the process of ass formation is decisively Shaped by a variety of institutional mechanisms that are them Selves ‘relatively autonomous’ from the class structure and which determine the ways in which class structures are translated into collective actors with specific ideologies and strategies, Some of this research has focused primarily on the political mediations of the process, showing how the process of class formation is shaped by the forms ofthe sate, the strategies of partes and other polit- cal factors." Other research has dealt primarily with he role ofthe Inbourprocess and the organization of work in structuring the process of class formation." Nearly all ofthis research has been {oncemned with showing the complex and contingent character of the relationship between class structure and class formation. ‘Neither of these kinds of contibutions—contributions £0 the conceptual map of empty places in the class structure and con teuttons to the theory of the formation of colletve actors from those empty places—is entirely new in the Marxist tradition ‘Theoretial discussions of the mide class can be found in set. tered places and certainly by the time when Karl Kautsky wrote about the middle classes around the turn of the century, it was ized as a significant problem.” And the casical Marxist theory ofthe state and partis, particularly as elaborated by Lenin, is pre-eminently concerned with the poitial mediations i the formation of clas actors, prety the revolutionary working “But while the themes in this recent work are rooted in classical Marxism, the new Marnst cass analysis is distinctive in two specs Fest, much of this work has attempted a level of self ‘conscious conceptual precision that was only rarely encountered in ler Marxist discussions of these problems. Secondly, i has sys Posing the Problem: The Agenda of Class Analysis 15 tematcally tried to develop concepts and theories at the “middle level of abstraction, less abstract than the exploration of modes of| production but more abstract than the concrete investigation of fhe concrete situation. Increasing attention is Being patd to the ‘theoretical dimensions of variability in ‘actually existing capitae jem While the more abstract debates ofcourse continue, there is fan emerging recognition that it not enough to have good abstract concepts of the capitalist state, of bourgeois ideology, of the aptalist lnbour-process and of the capitalist class structure; We tlko need a repertoire of concepts capable of specifying the Vat- {ablities in each of these at mote concrete levels of analysis. ‘This book will tempt to make a contibution to these debates on | class structure. Part One will evolve largely around concept fssues, Since these debates on class centre on the production an twansformation of concepts, chapter two will begin with a brief iethodological discussion of the problem of concept formation and then ‘continue. by exploring in considerable. detail the development of one pardcular conceptual solution tothe problem ‘of the middle clas, the concept of ‘contradictory locations within {tase relations’. The’ chapter will end with an inventory of internal inconsistencies and theoretical problems with this conceptualiz tion. Chapter three wil then offer a new general strategy for “analysing class structure which avoide the problems posed by the ‘oncept of contradietory loations The essential argoment i that the eoncept of contradictory locations, like much neo-marxst las analysis, has effectively displaced the concept of ‘exploitation’ from the core ofthe concept of clas structure, replacing it withthe concept of ‘domination’ The strategy proposed in this chapter attempis to specify the concep of exploitation in sucha way that it ‘in he reinstated a the central basis for defining clase structures in ‘general, and solving the conceptual problem ofthe ‘middle classes’ {in particular. Chapter four wil then explore the theoretical impi- cations of this new approach for a wide range of problems of interest to radical scholers: the theory of history, the problem of| lass formation and elas aliances, the problem of legitimation, the felationship between cass and gender, and a number of other Part two ofthe book will deploy tis new conceptulization of class structure in a series of empirical investigations. Too often ‘conceptual debates are carried out srctly in terms of the internal logic and consistency ofa conceptual apparatus with at Best ance-16 otal reference to empirical research. Chapters five to seven, therefore, will systematically explore a range of empirical prob. lems using quantitative operationalizaions ‘of the abstract con cepts elaborated in chapter three. Chapter five will attempt a tematic empirical comparison of the mers of the definition of the working class based on the framework elaborated in chapter three With two other definitions, one based on the eriterion of produe- tive labour and one onthe criterion of manual labour. Chapter six wil use the new conceptuslization to compare the United States and Sweden on a variety of issues involving class structure: the Aistibution of the labour-force into cass locations, the relation. Ship between this distribution and a variety of other structural properties of the society (economic sectors sate employment, firm size, etc), the relationship becween elas and sen, the lass sructures of families, the effects of class on income, anda number of other problems. Finally, in chapter seven we will examine ‘empirialy the complex problem ofthe relationship between class Sructure and class consciousness. Marx asked on the final page of Capial Volume 3, “What consti- {tesa elas?" This isthe Basie question this book hopes to answer, ‘The answer which will be developed in the course ofthe analyse will undoubtedly not be the one which Marx would have given i he had finished his chapter. Not only have there been s hundred years of theoretical discussion ofthe problem of clas since Mares ‘eath, there have also been a hundred years of history, and if Marxist theory is at all entific one would expect conceptual advances to have occurred in such a period, Nevertheless, the fnswer which I will propose wll try to be faithful both 10 the theoretical agenda forged in Marx's work andthe pollical goals that agenda Wes meant to promote tes 1, Kes Mar, Capita wo 3, Landon 1974 #8. xin pr ol mide Sse tind dein ss Feces a Theos of Supa Vatu Man ots Wha be Reo foes ‘Senphuin he sos rene fhe ee dee wind ne me cis nr not par ppred nelyou ef reaswhe ea Posing the Problem: The Agenda of Class Analysis 17 buen on te aon onan ah nent the td cy an per ‘th oper ten howsng ots In Marga Ncsuay,‘Moleratuad Midoe "5. The dit Sem to ate at nay aps counties et employment iets ace stout 972 aleve soar nae pss ae [Soar force (eoving toil US Goverment fare) See then te ‘omen nerd wen yr eat 1 ‘human nseienesn sotlflsbons has een the objet ot prota ne fe ‘Shee demerin wang teen srgued a ance sod eons nol ot ‘Eat he homo nae inte aioe cnet es eretre ‘tes 9 ese stgush the str rm te cae wth te to. {ie The formal fave sped doce na ge scl atone an ence Sniper of pop tisha ore hem a cece mc ot ‘Punlr psn Sed fren sou can chine theta neue {"fgor te couse enertion ad et he ds cae of he ton ‘ould remain te me. Zi 7 "Undeniable formation inter of te naan ‘ronal Adno Praeworsa. See input "For Prostata ino Cast Toe Fes fas Some fom Kat Kn Te Car Sane Hee Debates Pals Soin, 07s 00-8197 “nites Mar ior wet be eelpmentl age of captain in some srs inte to the ge of be plist mode of pci that ese ‘geome yas a thee! of baton he mode produto Tht {enayned tthe sme fel of abc st he genera siete popes of eee tee Shire eran racometn cae ates Seca ea ernst ea pea (REA ora gore ws {Se opera sneer aad en pie poses ta 2 bata i ti ee sie ne ea ee eee Bebe ire melateaipeirgioms oneas Eee mesa i ne germ nc tt een Frenomeson’, New Lf even, 123,198, Neral iets Css CompromisePart One Conceptual Issues| | | 2 The Biography of a Concept Contradictory Class Locations Inthis chapter we will examine in some detail the process by which 4 particular concept for solving the problem of the middle classes incapitlism was produced, the concept of contradictory lations ‘thio cage relations’. This wil not be a literal chronological faovount of the development of that concept, but rather kind of| Togieal reconstruction of the process. The actual history of the foncept was not quite This wasasfaras the development ofthe concept of contradictory“6 ‘opt Ge {Sl Boe | eee ee Sesto Eapoyes “op Becntes t Sepersion re Working Gass | ontatictory Locations within Css Reations ‘Basic Class Map of Capitalist Society locations had gone by 1979, AF that time I embarked on a lage empirical project on clas structure, cass experience and cas com Sciousnese, The heart ofthe research involved developing a survey ‘Questionnaire which opertionalized the class ertenie In my pro- posed lass mapof advanced capitalist societies together witha wide ange of other variables (measuring alternative concepts of class, lass consciousness, class biographies, ender attitudes, and other things), This questionnaire yas then given to random samples of the working population of a number of advanced capitalist societies." At the end of al of my previously published empirical ‘work I had always bemoaned the fact that the data used ia mY ‘Satstical investigations had been gathered by bourgeois sociolog- Jsts and economists, using non Marxist categories. While this pro ided! me with a convenient excuse for problems in my own “analyses, [felt that t would be useful to generat a substantial body ‘of statistical data explicitly gathered within a Marks framework ‘Once began the task of tying to formulate spec questions to The Biography of « Concept: Contradictory Class Locations 49 operationalize my proposed class concepts, it became very dear thatin certain important as they remained vague or incomplete In particular, the "semautonomous employee locaton wat Impose to operationalize ina rigorous maner. Ths practical Aifclysuimulted a rehining ofthe loge of hs category. ‘The fethinking ofthe category "semFautonomous employes incised ith my intial work on paper on posteapiast Societies, eventually pushed es ‘Captim's Futures At the heat of the analysis of that paper was a discussion of what 1 tcrmed the intr-penctation’ of modes of production, 3 forme ‘of production relations which combine spect fom dstigt modes ‘of production in a jstematie way, This concept was important for gorously spocfving the. tendencies towards postcapitalat ‘Otitis generated within capital Hall “The concept of nterpenctation of distinct relations of produc- tion ls had a bearing on the persistent problem of propery ‘etning sem-autonomous employees. In al ofthe ear work L ted specited a st of ereria for cass relations and then defined partchlar eas locations and contradictory locations by thei a. {es on this common set of siteria. But i certain Cases te ‘efined by diferent types of production relations (mods of pro ‘lucon), then diferent ertefia are clearly needed. Feudal sri, could not be defined by values based on capitalist ‘rterionspectying relations of personal bandage ‘would be needed a xiteron which is absen from the speciation Uf any class nthe capital mode of production. In other words the global concept of ‘contradictory locations within clase relation needed to be formally ifeentated into {wo distinct sub concepts: contradictory locations within amode of produstion, and contradictory locations between modes of produc: ion" Inthe former ease, contradictory locations ean be spectied tithin a single set of ester; Inthe later, the contradictory haracer ofthe lotion reques two distin es of eel, each rooted in diferent production relation “This re-conceptuaization meant that to define property the category of sem-astonomous employees me had 0 seciy the appropriate erteria forthe petty bourgeoisie, ie. forthe eas ‘Rtermined within simple commodity production. The necessary tarification to accomplish this task ame out of my debate with Solin Rosmer over the roe of domination fn the eoncet of cas. ex Ava result ofthat debate Us convinced thatthe central sing cterton forthe socal relations prdveton, wich a“Sine Commodi Prodan Anprorion : | j Developed typology of lass Structure ‘Canis Moe of Production ‘Bourgoie: Bue Gam Loaion ‘Top manugere Empoyes The Biography of a Concept: Contradictory Class Locations $1 tur provides the basis for defining classes was the unity of appro- priationreladons anddomination elations."Thsledtoasimpliica- {ion of my orginal extria for capitalist clas relations from the to two, Low fet that control over the operation ofthe physic means of production and direct control over work should be treated es two allemative mechanisms of domination of workers, rather than two dimensions of clas relations with equivalent com. xptual status to contol over investmens. Clases, and accord- ingly contadictory locations, ate therefore to be defined by their Bestion within prea pe f apropos amination In these terms, the problem became one of specifying appropris= tion and domination relations within simple commodity produc- tion. [took the appropriation relations tobe unproblemate, defined by individual appropriation of the product of one's own labour Ge." seltemployment)" Domination relations within. simple commodity production were, ina parallel manner, defined as sel ontro, (Le. the individual sel-direction within the labour proces) ‘Sach ‘sel-iretion’ in operational terms was the ability to. put ‘one’s own ideas into practice within work, orn tadiional Maraian language, the "unity of conception and execution’. Tis meant that semi-autonomous employees were now defined as postions which dnt involve se-appropration of the pro: duct of labour (Le. they were capitalsieally exploited) bat did involve set-direction within work (ie. they were not capitalist cally dominated in that they retained an effective unit of concep. tion and exceution). It was still «difficult task to operationalize this eiterion but the concept had more precision than earlier ver- sions had “These modifications lead to the final version ofthe dass typo! ogy of contradictory loations represented in Table 2.2. Thiss a Tong way from the inital, simple four-fold table which began the ‘ory of contradictory locations. And, a8 we shall see, there were Sulfient remaining problems withthe conceptual framework that fventually I became convinced that it in tum needed to be Superseded, Problems With the Conceptualization “The concept of contradictory locations within clas relations was, I believe, an advance over the alternative ways of dealing with the problem ofthe ‘middle clas in advanced eapitalist societies. Bothsivas. Yet, fom the start there were problems Some of these were apparent quit early others became clear only inthe cour ofthe development and suc ofthe concepe, partic lent in the coment of my empirical vestigations. Fou ot these roblems were particulary significant the claim tat contradictory [bcatons ae contractors the status of autonomy” asa rteron for ean; the relevance ofthe concep of sontradtry locations for postcapitalist steer the marginslzstion ofthe Concept of xpos inthe concept of Clas. (1) The Contradctoriness of Conradiciory Locations. From the fist publication using the concept of contradictory locations, the tse of the term contradiction’ has boon eriized In the case of| ‘managers a plausible story cam be told. If we aeept the character- {Eatin of managerial positions as combining relational properties of proletarian and bourgeois clas locations, and if we accept the general Marxist thesis that the objective interests of workers and apialsts are intrinsically antagonistic, then at 2 minimum it takes sense to describe the interests of managers as interally inconsistent. Because ofthe systematic character ofthis inconsis tency, it would not Be unreasonable to characterize ita contradic- tory a5 well, ‘But why in the word should semi-autonomous employees be viewed! as having internally inconsistent interests? To say that ‘emi-autonomous employees have contradictory (rather than sim ply heterogeneous) interests so imply thatthe proletarian pole of| theie class location generates interests that contradiet those gener- ated by the petty- bourgeois pole of thei location, Presumably this petty-Dourgeois pole defines interests inthe preservation of aut ‘Snomy within the labour process. By virtue of what does auton: ‘omy within the labour process define objective interests that con tradiet working-las interests? The only answer I could provide ‘was to say tht workers had interest inthe collective contol over the labour process—colleetive autonomy if you will—-which was ‘opposed tothe individualized autonomy of semi-sslonomovs emp- Toyees, This, however, was unsatsactory since collective control ‘over the Taboue process aot necessarily opposed to significant ‘Spheres of individual control over one's own work ‘A silt problem exists for the smallemployer contradiciony location, te location which combines petty-bourgeois and capital The Biography of a Concept: Contradictory Class Loctions 53 ist lasses. While it may be that small employers have specitic immediate interests opposed to large capitalists when those large ‘apitaliss compete with them, it isnot obvious that they have any fundamental interests that are necessarily opposed." They may face various Kinds of dilemmas in competing successfully ia world of large corporations, but this doesnot obviously imply that {hey nave internally contradictory basi interests. ‘What I have called ‘contradictory locations within class rel tions, therefore, may be "ul or “heterogeneous locations, but except inthe ease of managers and supervisors, they are not obvi- busly “contradictory” locations. The term could therefore be Fetained for what I called contradictory locations within modes of production, but seems less appropriate for contradictory locations Between modes of production (@) Autonomy ar a Class Criterion, A second problem with the tlaboration of contradictory locations centres on the category ‘semi-autonomous employees’. Three, issues seem especially troubling: de claim that autonomy is a‘petty-bourgeois property of class relations; the relatively unstable ‘or underdetermined Character of autonomy in certain work settings; and empitcal fnomalies inthe use of the concept, ‘Even if we accept provisionally the idea that autonomy is an pect of class relations, dors it make sense to treat autonomy as having 2 ‘pety-hourgeos" class character? There are both situe- ‘ural dd historieal objections to this characterization. Structurally, the characterization of sslonomy es “petty bourgeois" ress largely on what may bea rather romantic image of the pety bourgeoisie as independent diret produoers character- ized by & ‘unity of conception and execution’ The contrast be treeen independent producers (celhemployed artisans, eaftsper- Sons, shop-keepers, farmers, ete) with such autonomy and pro- letarian wage-abourers without such autonomy may simply be incorrect. On the one hand, ora variety of reasons, selemployed petty-bourgeois producers may have litle choice over how they produce of, in some circumstances, even over what they produce “Their options are constrained by markets, by credit instittions, by long-term contracts with capitalist enterprises, and so on. On the other hand, itis easy to exaggerate the extent to which workers in ‘modem capitalist firms are indeed fully separated from “concep tion’, since in many factory settings the actual operation of produs- tion Continues to depend heavily ona wide range of accumulatedse knowledge on the shop floor, knowledge which must coastanty be applied in non-routinized ways.” Such autonomy, therefore, may ‘ot have distinctively “petty bourgeois character tal. The only thing which defines the petty bourgeois is ownership of certain Kinds ofassets—land ool, afew machines, perhaps in some cases ‘skill’ of credentials and selLempoymeni, ut not work auton- ae "The characterization of work autonomy as pety-bourgeais is also very problematie when looked at historically. The semi ‘autonomous employee category contains wo quite distinc sorts of Positions: highly autonomous. cfaft Wwageearners, and chnical wagecarners. The former could plausibly be considered combinations of pety-bourgeois and proletarian ‘sees, since the independent artisn san historical antecedent to the modern craft worker. It makes lest sense tO see 8 research ‘sclenst, a university profesor, an indusral engineer ora social ‘welfare counsellor a having a pell-bourgoois character combi Ing elements from the capitalist mode of production and simple commodity production, The kinds of autonomy that occur within ‘contemporary bureaucatically organized institutions cannot be treated as remnants of simple commodity production’, But this ‘what is implied by treating semi-autonomous class locations as Combinations of proletarian and petty-bourgeois clases "A second problem vith semi-autonomy a 4 cass criterion is what could be called its strtural underdeterminaton. Whether fr not a given job is ‘semi-autonomous could easly be a conse- ‘quence of rather contingent characteristics ofthe work seting. For ‘xampie, a research technician could move from a job where the Scientist in charge assumed that technicians were incompetent and ‘hus gave them no responsibilities, to a laboratory in which the Scientist was lazy and left a great deal of discretion and decison- ‘making up to the technicians. Inthe second job the technician ‘would probably be classified as semi-autonomous inthe former as proletrianized, Should such a shift in jobs be viewed as change In the clase character of the techniian poston? Ts the former position purely. working lass while the latter, semipetty bourgeois? The concept af class is meant co designate fairly sable land structurally determinate Properties of locations within the ‘Social relations of production, Ata minimum, the seemingly con Tngent character of autonomy in certain jobs a weakness the claim that autonomy isa class eiterion, ‘A final problem with autonomy as a class eritrion revolves The Biography of a Concept: Contraditory Class Locations 55 stound a numberof empirical anomalies that have emerge inthe ouse of the empha esearch involving the concept For exam lo, if autonomy i defined in terms of contol over what one Jrodces nd how one prods i then many jeitore in schools ‘tho also perform a varety of handyman’ tasks wil end wp being Imore autonomous than ain plots. Now, one could regard this ‘St deep discovery about the ature of the ces location of pilot, in spite ofits apparently counterstuitve character. Ii more Dlable tat it inciates she problematic status of the claim that, Eutonomy shout be viewed ar base criterion for cls 9) Classes in Posteaptalst Societies. Casiat Marxism was Sbooltely unequivocal about the historical prognosis for expt. {Em socal and ultimately communism—-was the fture of capitat societies, The bearer ofthat necesay future wae the ‘working class. The polarized” clas structure between. the bourgeesie and the proletariat within capitalism thus paralleled the polarized historia alternatives Benen capitalism and soil. “Tne actual historical experience of the swentith century has A General Framework forthe Analysis of Class 89 <éepending upon the particular combinations of exploitation rel tions in a given sorely. The historical pattern of principal con- fe 3.4. In feudalism, the ‘iitieal contradictory location is constituted by the Bou he fsng clas of the successor mode of production." Within eapit {sm the central contradietory location within exploitation relations js constituted by managers and state bureaucrats. They embody a principle of ease organisation whichis quite distin from capital- {aan and which potentially poses an alterative to capitalist ela- tions. This is particularly true for state managers who, unlike cor- porate managers, are les likely to have their careers Erafed withthe interests ofthe capitalist class. Finally, Societies, the “intellegetsa” broadly defined constitutes the pivotal contradictory lnetion.© asic Canes and contradictory locations In sucesve modes of roducion Node poduon Pe case nan ipa Serge Mangen/breucats Oncofthe consequences of thisreconceptualizaton ofthe middle class is that Ht no longer axiomatic thatthe proletariat isthe “unique or perhaps even universally the central, alt the capial- ist class for elass power in capitalist society. That classical Marxist ‘sumption depended pon the thesis that there were no other ‘asses within capitalism that could be viewed a the bearers’ of historical alternative to capitalism. Socialism (as the transition to communism) was the only possible future to capitalism, What ‘Table 3.4 suggests that there are other class forces within capi talsm that have the potential to pose an alternative to capitalism ‘Alvin Gouldner and others have argued tht the beneficiaries of social revolutions in history have not been the oppressed clases of fhe prior mode of production, but ‘third classes" Most notably, it twas not the pessantry who became the ruling class withthe demise90 of feudalism, but the bourgeoisie, a class that was located outside the principal exploitation relation of feudalism. A similar ag ‘ipitalism and experts with respect to state bureacur ineach ease these constitute potential rivals tothe existing ruling las, Inthe case of capitalism, it might seem rather far-fetched t0 claim that managers and sate bureaucrats constitute potent ‘hallengers to the lass power of the bourgeoisie. At least inthe ‘sevanced capitalist countries, corporate managers are so closely integrated into the logic of private capital accumulation that it coms quite implausible that they would ever oppose capitalism in favour of some sor of statist orgatization of production. As rites (ofthe ‘managerial revolution’ thesis have often argued, whatever Special interests or motives corporate managers have, the realza tton of those interests contingent upon the profitability oftheir firms and they will therefore adopt strategies consistent with the interests of capital And even for state managers, who arguably have a power base that i atleast partially independent of capital, ital seems very unlikely that they would ever become conss- tently antrcapmtalist becatse of the multiple ways in which the interests of the state are subordinated to and coordinated with the interests of capital, Since in capitalist society state revenues ‘depend upon privately generated profits (because the state islt des not organize production), the state is systematically con- ‘rained to actin a way that supports the profitability of capital and = captalist exploitation. Regardless of their personal prefer- fences, therefore, sate managers cannot capitalist ways managers and bureaucrats as even potent bourgeoisie “Behind each ofthese claims about the effective integration of ‘managers and Bureaverats into the capitalist social order is the “sumption that capitalism is successful asa system of exploitation tnd accumblation. So long a firms, in general, are able to make Droits, they are able to integrate their managers into a logic of {xpita accumulation; ind s0 Tong. as capitalism reproduces fevenue base forthe state, state managers will have thei interests tie fo the interests of eaptal. But what happens to these interests land strategies if capitalism permanently stagnates? If profits ean fo longer be assed inthe long-tun? TE the career prospects for large numbers of managers became very insecure and precarious? A General Framework forthe Analysis of Class 91 Would statist appeals for greater direct state involvement in con: tolling investments and lows of capital Become more attractive to corporate management? Would statist options be seen st more fealistic for state manager? I do not want to suggest that statist, ‘Solutions that undermine the power of the capitalist lass would Automatically be pursued by managers and bureaucrats under such ‘esonomic conditions, There would also have to bea range of polit fea and ideological condition to make such strategie viable, and there is no necessary reason why such politcal and Hologic conditions would be forthcoming even In situations of chronic Stagnation." The important poiat in the present context is not that there be any inevitability tothe emergence of such conditions, ‘ut that one can imagine historical conditions under which mana {gers and bureaucrats even inthe advanced capt je third world countries) would find ant-capit tions attractive “The historical typology of contradictory locations in Table 3.4 oes not imply that there is any inevitability to the sequence feudalsm- ‘Sprints New Let Reve, Mater 198, p= Sey ety hens Cue 9 Marian fo acne of why te ser af prc can play be ewed ging bir econPart Two Empirical Investigations5 Empirically Adjudicating Contending Class Definitions In this chapter ne wil stomp to adjicate empiisallyberneen contending sfinons wth the Marts theany of cane Av T seed in shane ny he robin of he mil cs as een ithe conte the sontenportytthinkng of Marat cones of else nroctre The empra avian nti caper wil therefore Toss onthe debates ovr the lin of demarcation be- {ween the working css and “nidleclas” wage euner More ‘pesca Tw proposes strategy forthe cna sent he reaine mero the apronch to specping he working Class elaborated in chapter thee and two tmportan lemanes the ample seatifestion ofthe workings with mane Wage ltbor and the more somple conepttalation of the workag Chane proposed by Pouantas These ae not of couse, te oy esising terns. Many Mart slop adopt tose Skfnston of the working clas that inca ll nowsupetiory ‘manual nbourers pls 'proletaranze” white colar workers (er ‘Gat worker expel} Sich»defaion comes extemal Sine tothe expioatonsenred concep | have propored in chapet three, and practical terms they ae amet ietingushble have’ desided to focos om thew wo" parc aes, therefore party Besa an mpl inenenton the debate Itty to produce reltvely box ond erpetable rests: Inthe it seon ofthis cpt wl yout he fas age of snyempircalsaty, Ths wl be followed by a dcsson othe Praca! task of operationalizing the verbs fo hese in ths Steg, The inal section othe chapter xemine the esl ot shal stay ig these opertonlznony. 136 Empirically Adjudicaing Contending Class Definitions 137 ‘The Empirical Strategy Definitions of specifi casos canbe regarded as a particular kind ‘of propoition. Ail things being equa al units Gndidls andor fans, depending pon the specie sues under discussion) bitin a gen se should be more like each oter than ike units Iother clases wit respect o whatever ti hat clas s mean fo xplan, The proviso “wth respect to whatever is tha ls fheant to explain’ ie equivalent (0 saying that these kinds of {elinionsl-propsitons are always with respec to given theoretical object, The dupes in question are aot over how best {o use nord aough sich sven may e important (0 ¥Oid Confson n thot! Sasso, The debatnarcovr how best {o define concept given that designates a theoretical objec that I sujet fo tase agreement, Gat ot ase) meant © fxplan dietary preferences, for example. There i therefore no feason to believe that individual in the same lass tin ferent ‘tinc groups wil be more like cach other with respect to such preferences than they mill be like people who are inthe same Ethnic group but in eiferent lasses. On the other hand, class Structure fs neat to explain (along with other mechanisms) class Confit A particule deiniton of te working class thusa propo Sti abut the nes of demarcation i the confi) promi process. This fs nor equivalent it most be emphasized (0 sing {htall workers wil actin identical way, ince the lai snot that {las location isthe only mechanism affecting clas action. There Shay be ele or gender or other mechanisms that vary among, ‘Torker and produce empirically heterogeneous outcomes in $e Gre homogencous cass determinant What s being claimed how ‘rents thal ther things Beng eqs, two people who fal within thes lines of cass demarcation wil ave a higher probability of teheving ina similar fashion within class confiets than wll (0 people filing on cifferent sides of the Tine of demarcation ‘Retecingly each contending definition of the same elas i an implct propostion about the homogencity of eects generated by Ine structure which the defsiion attempt to spect i defintions are propositions about ines of demarcation for homogenvous effects, then this suggests that the appropiate SiatepyforaGadiatingUsputes over deiaionsof cass to foras ‘on thane eas where oe dfinion paces vo posit138 treats them as homogeneous. These are the cases where the df. ‘ences in definitions have diferent empirical Implications “These disputed cases can be identied by a simple cross: tabulation ofthe two definitions, This is ilustrated in table $1 for two contending definitions of the working class, copie tet ag at ea se Bm | antl ie ‘ovings | Steger 1 anes 3} a Call Lin this table consists of positions which both definitions define as working class. Cell, on the other hand, consists of vagelabour positions which Both definitions se as "mile clas. Calls 2 and 3are the disputed categories, Denton A argues that cell3 should be much more like eel than tlie ell and Gell 2 shouldbe much more like cll 4 than tisk cel 1, whereas Sefnition B argues that el 3 should be basically similar to eel and cell 2 should be basically siilar wo cell The empirical adjudication of these contending definitions ofthe working las Consists of secing wheter the disputed categories ae closer tothe ‘gred-upon workers o othe agreed-upon middle’ class in terms ‘of eriteria gn which both definitions agree the working cass and the midale’ as should fer a should be noted that the logic ofthis strategy for adjudication oes not imply thatthe disputed category shold be indatinguish- ble from the class in which a debnition caimet belongs. Take the problem of the identifstion of the working clas with manual labour: Even one rejects the claim tht this an appropiate way of detring the working class, one might sil elev that fora “arity of feasons the menual-non-manual dsnctionconsitutes Empirically Adjudicating Contending Class Definitions 139 an internal division within the working class. This could imply, for ‘example, that white-collar workers would be less ideologically prowworking clas than manual workers and yet would stl be ‘shin the working lass. The hypotheses, therefore, are not that the ‘disputed category is indistinguishable from cither of the ‘agreed upon categories, but chat i signficanly closer to one oF the other. "nthe specific comparisons we will be making inthis chapter, the precise formulation ofthese hypotheses difer somewhat from ‘the Simple model in Table 5.1. There are two modifications. Fist, the exploitaion-centred class concept includes» specific acknow- ledgement of certain kinds of marginal’ class locations, parieu- larly wage-carners with marginal credential sects and wage- ‘carers with marginal organization assets. Such postions should not be simply amalgamated with either workets of non-workers in the adjudications, since this could conceivably have significant ffets on the interpretations of results. I is more appropriate, therefore, formally co include sach marginal workers in the adjudi- ‘ation typology. This i lustrated in tale 5.2. While the bulk of the adjudication analysis will focus on the corners ofthese tables, ‘enprodactve abou concepts ipoeaon Ral acon cred ‘scion Workings “Mideta 0 7 Magna DB) @ an {Cal(5] is emp n he somprson th Potezass ein son of te working dns wa unproductive10 explicitly including the marginal categories will enable us to pursue amore nuanced analysis where necessary ‘A second. modification of the simple adjudication table in table $1, is that in the ease of the comparisons with Poulaneas's efinition of the working class, there are no eases in the lower: leftcand corner ofthe table cell 8) there are-no postions Which PPoulantzas would consider working lass but which would be con. sidered unambiguously outside of the working cass. by an texplotation-centred concept. The debate with Poulantzas’s defini tion, therefore, is stcly over his allocation of certain postions hich are working class by the framework elaborated in this book to the ‘new petty bourgeoisiy, especially unproductive wage ear= tis important to emphasize that even if one definition i unam- biguously shown to be inferior to ancther in this procedure, this does not definitively prove that its incorrect Its always posible that some independent mechanism is at work which confounds the resulls. Let us suppose, for example, that women ate predomi ‘antly subordinate clerical employees, and we were to use income asthe dependent variable in our adjudication. And futher sup. pose, a5 the ease, that there is systematic wage-dscriminaton “Against women in general. In this case, the relationship between gender and the contending class eiteria could have the effect of Sepressing the overall average income of cell Zn table 5.2 and thus making i much closer to cell 1, even if among men or women taken separately the manual-mental distinction was 4 sharp on Sand cell? was much closer to cell 8, The defender ofthe manu labour definition would then beable to show thatthe adjudication was confounded by the effects of gender-mechanisms. The verdict Of the initial adjudication, which ignored the effects of gendet mechanisms, would therefore be overturned in this subsequent nals. Empirical ciferences must therefore be viewed strictly as { provisional bass for choosing between the contending defi “These adjudication hypotheses wil form the basis for the empir- ical analysis inthis chapter. ts ofcourse nots tivial problem to specify precisely what ts meant ether theoretically or empirically by more like" in these hypotheses. To do ths involves bath defin ing the content of the object of explanation (es. consciousness {rms of collective action, income, etc) and the appropriate stan dard for defining similarity. Once these tasks have been accom” lished, however, the empirical tes is straightforward. It sto this Empirically Adjudicating Contending Class Definitions 141 task of transforming these general propositions into more concrete esable" hypotheses that we now tur. ‘Operatonlizing the Adjude {tis on thing to map out the loi ofan empirical ajudieation of Contending delintons and que another to generate the nee Sry Kinds of data and stata procedures cary out the exer Sez convnengly The dificulty Us hat mos exiting soso fv wnich woud be relevant to te task were gathered within & hen Mart Conceptual framework for quite erent purposes ‘There are vey few sock surveys which contain eter the Roses ‘ry intormation (0 operationalize the explotation-entred eon- ‘Sora orto aes the periment Kins of ey which make {So ofthe Maran concept of das consciousness. ita for eto tet 107 red oh as become a large comparative survey research projet on css tue {ure and Gas cnstlousness Tis invlved fr developing 2s ‘ey questionnaire which adequately measreda varity of aetna {he Maps and non-Mardet cast concept along wih range of Sher tous, ond then adminierng Wt ‘atonal samples of Sits na umber of counties The dats fom the United Sats Sney inthis projet wll provide the bau for the empial SSjlaton ote deste wich we ae comer We will first examine the ‘dependent variables in terms of which the comparisons of concepts wll be assessed, before turning tothe problems of specifying the various class structure categories inthe {ro adjudications and th statistical procedures to be employed While the details of these operationaliations may seem rather tedious, they ate nevertheless important, for the cogency ofthe final compatisons depends largely on the persuasiveness of the ‘operational choles made in setting up the analysis wll therefore fb through each ofthese stepe quite carelully. Readers who are [patient to see the punchline of the story could skip the res of thie secdon and turn directly tothe statistical reslls which follow. DEPENDENT VARIABLES FOR THE ADIUDICATION In many ways the most deicate part ofa conceptual adjudication lies in the specification ofthe ‘dependent variables in the analysis.12 ‘The adjudication proposed here only makes sense because the rival definitions are meant 10 explain atleast some of the same things. Tes therefore crucial that appropriate variables for making the comparisons are selected “This task poses rather substantial problem for the definitional adjudication at hand, In Marxist theory class structure is above all ‘meant to explain a range of macro-socal processes: class forma tion, class alliances, social conflict, historical trajectories of social ‘change and so forth. Ofcourse, Marxists frequently make claims about the consequences of class location for individuals (ep. in ‘explanations of individual consciousness), but such claims are typi= cally undertheorized and in any event are not the core of the theory within which the concept of clas strutute figures. A defender of Poulantzas’s general stance towards sss structure could therefore argue, wit some justification, that the micro level variables which Twill investigate are at best of secondary impor tance within Marxist theory, and therefore cannot constitute & ecisive basis for comparing the definitions ‘Nevertheless, Twill propose a number of individual-level vari- ables tobe used in the adjudication of the conceptual debates we have been discussing, Ido this for two reasons. Fst, even though Marist class analysis is, above all, 2 macro-theory of social rela- tions and social change, that theory must be Tnked to a mlcro- theory of outcomes for individuals itis to be complete, For class siruciure to explain social change it must have systematic effects ‘om individual action. This does not prejudge the question of the {extent to which the practices of individuals are explainable by cass felatons or by other determinants, but iis hard t0 imagine how lass structure could explain cast strugale and social change if individual behaviours were random with respect 0 lass. This sug- {ests that individual-level variables are appropriate criteria for ‘Comparing cass concepts, even if they are no sufficient basis for 2 definitive judgement of thes relative ments” Second, to engage propery in an adjudication of cass structure concepts sing macro-historieal data requires a very broad com Parative analysis ofthe relationship between cas strocture onthe ‘ne hand and clas formation and class struggle on the other, The logic ofthe adjudication using macro-stuctural data is that one way of specifying the variations in elas structures, both acrors time and eases, wl etter explain the variations in elas formation and strugsle than the rival speificaion. This cleanly much ‘more arduous empirical enterprise than the more miero-centred Empirically Adjudicaing Contending Class Defitions 143 proach being used here, This is not to say that such a task is f, but its beyond my present research capabilites, judication of contending clase definitions, wil there irctly ted in particular class-oriented attitudes and personal income. In wat follows I wil briefly jusify the use Of these specific variables and explain how they willbe measured. Aides: Justification “There ae two crt! objections which could be rise against he ropostl ous attitudes ta eifrion for adjudicating sas etn Tou: fn, that atiudes cannot be comsideret vad indicator fof eas consovsnes, and secondly, tha even i atitudes per fect reflected clas comsiousaes, consciousness itself is only loosely elated classaction, and this the only appropriate trteron for assessing cas concep "Atte responses ona survey, however wel designed the ques sion, are at best loosely eisted t0 the Maras concept of ‘clas consciousness” As erica of surveys have often pointed oat, the opinions indiduals express are heavy contex-dependent, Sand the peculiar context of & questionaife ineriew™an 0 inted individual talking to a representaive of scientific tesitton-~undoubedly shapes the patra of responses. 1 could well happen, for example, that workers wil exprese much more Conservative views in response tthe. questions posed. in such isterigws than they would in a conversation ‘ith their work: ‘Nothing inthe questions which we will use to construc atitude variables svois exe potential bans and such ses might wel Influence the conclusions which we draw from the data, But is Important o remember that biases in data Jo notin and of them Selves invalidate the use of such data totes hypotheses, since ‘ues may be neural with eapct tothe expectation ola propos ono tenet more rer tha em itt vba the plasty of the hypothess Inthe esse of ou dentinal sdjueations, the eal empirical fests are alvays of the differ ines Between trios esteporics. Unless the bioes dilfer 70s the cas categories being compared in ways which infenee the ricl tests, then the adjudication comparisons ean be perfectly Sound even ifthe datas quite Stored. Thu, for example, We ‘would adeed face problems the biases nthe responses operated16 in the opposite direction for prolearanized clerical employees, they Gd for mansal workers. This could ave the effect of making appear thatthe Two caiegories had similar class con- Schusess (ar measured bythe atitede question) when f fat thei real consciousness sas ite different. Inthe absence of ‘rong reasons for believing that sich compex interactions of bis with ease occur, I vil sesume that though there re ceri nificant bises of warous srt im the responses tothe atte ‘gestions, these bass ae rendom with respect tothe comparons wre are making “A more serious objection to using attitudes asa criterion for adjucicating contending case definition that cas conscious neta leat if this enderstood as cesignating sable forms of Consiously understood bets by individuals, is only very weakly inked to etal cae behavior. Thos gute apart fom the problem, of using attitudes to meanire class consciousness, they ae 0 Spproprate as adjudication eer, snc the contending pea Catto of claw structure re meant to explain sass proces) Serugle, and since consciousness nota very important deter= Iant of actual behavior Thee not to sugges tht cas actor: te automatons, aconsousy paying out seit im 8 ramos implies merely that clas action fe much more heavily dotermined bythe conerete choices and pressures that people face in given ‘cumstances than by aay sable or endurng pater of com Scioumness (bli copnitve structures, values, et) which they bring to thoe chokes The only appropriate adjudication ete tion therefore, mould be the actual colces made, that i, the patterns of elas behaviour ‘My assumpion in adopting atts a a eiterion i thus that they are notin fac “eppenomena’ that they have real conse {uenoes for clas action, and hat hey ae, to some extent at est, ‘etermined by cas location. This implies that behind my se of fades 8 2 causal argument about the relationship Between fem fens nbs, ca sctn ania ca “Cas location sa basi determinant of the matrix of objective possibilities faced by individual the rel alteratives people face tn making decisions. At one level tis concerns what Weber rele red to atthe ingvidols "ie chanes, te overall ajctory of posits inidualsface over theif eee Ina more mundane fay oper he dy shies peop hes aout what 0 ban Empirically Adjudicaing Contending Class Definitions 145, “The objective alternatives faced by individuals, however, are not siretly transformed into actual ehoiees or practices. Those objes- tive alternatives must be perceived, the consequences (both mater- {al and normative) of diferent choices assessed, and a specific Sltemative chosen in light of such assessments. This proces is partly the result of conscious, active mental evaluations and Eakcalations; te partially the result of what Giddens refers to as ‘practical consciousness’ the routnized ways people negotiate and Understand ther socal’ word; and itis partially structured by largely unconsciogs psychological determinants. In any cae, thi subjectivity mediates the ways in which the objective conditions of ‘lass locations are translated into the active choices of lass factions. While the objective social context of choice is clearly {important in this explanation, T would argue thatthe subjective Imediation of choies-the actual process of choosing—is an ‘ssental pat ofthe process as wel ‘For our present pirposes, the critical link in the argument is between cass location and forms of stable, casscelevant subj: thiy, It could be the case, for example, that although forms of consciousness mattered great deal for explanations of class Struggle, the mechanisms which determined such consciousness ‘were not located within class felaions as such (or at Teast, not Ioceted there in important ways). Schools, churches, the family, the media and so on, could sl be much more sigifeant determin: fants of forms af consciovsness than Tocation within the class Structure, If this were the ease, then elas consciousness—let alone Atitudes’ which “only indirecly” derive from such conscious fhes—would not be a very effestive ertrion for adjudicating ‘debates over definitions of clas structures, Tam assuming, therefore, that one's location within the struc ture of clas relations is an important mechanism determining forms of consciousness, This assumption is based, atleast in pat, ‘onthe viow that class locations objectively structure the interests ‘factors and that people are sufficiently cational to come to know thon interests, There shoul, therefore, be a least a tendency for these aspects of eonsciowsnesé which revolve around clas interests to be structured by das location, fone accepts this kind of reasoning, then class consciousness ‘an be tretted as an appropriate criterion in the adjudication of Contending cass defintons and responses to a survey question- fate can be viewed as an approprite indicator of cass conscious- mess, Again, 861 have said previously this nt to claim that lass146 ' the sole determinant of consciousness, but simply that it gene tes suiciently systematic effects that consciousness can be used asthe bass for evaluating contending views of clas Attitudes: Measurement ‘Most of the attiude questions used in tis analysis are what are called “Likert items. Respondents are read statement and then ssked to indicate whether they strongly agree, somewhat agree, Somewhat disagree ot sttongly disagree with the statement. The lass structure survey contains a large number of such questions anging over many different topies, For purposes of the adjudice tion of contending cass definitions [wil retit the analysis othe surey items which have the clearest class content, since these should be the kinds of attitudes most systematically shaped bythe individuals location within the age structure 1. Corporations benefit owners atthe expense of workers and 2. During a strike, management should be prohibited by law from hising workers to take the place of strikers, 3. Striking workers ate generaly justified in physically prevent- ing strike-breakers fom entering the place of work. 4. Big corporations ave far too ‘uch power in American society today: 5. One of the main reasons for poverty is thatthe economy is based on private property and profits. 6. Ifgiven the chance, non-management employees atthe place ‘where you work could ruin things effectively without bosses, 7. Tis posible for a modern society to fun effectively without the profit motive, ‘An eighth item was the following 8. Imagine that workers in a major industry are out on strike ‘over working conditions and wages. Which ofthe following out= ‘comes would you like to see oceut: (1). the workers win their most Important demands; (2). the workers win some of their demands 8nd make some concession; (3). the workers win ony a few of ther demands and make major concessions (4. the workers 0 buck to work without winning any of thelr demands Empirically Adjudicating Contending Class Definitions 147 Fach ofthese items was coded +1 ifthe remondent (00k the Sorting cas postion it they took the proepitli cass fonton and 0" thy tad that hey ide ow oe vero, 9 {peor tem they pave reponse extegory (2) These eight Feaponses were then ade p,geersting sacle anging rom 8 {Gntnmaliy propia) 10 8 (maximally provvorke) that {Mtosures ane po-morer or prepa oFetton on hs SXror question anegave vale many tat he raponden (ook {he pubcapiue postion more frequently than, the poorer postion, a poste vale indeates the oppo in addion fo wsing this consrcted omeeovsnes ele, we il ako examine the relalonship between das structure qn fail onventional variable meastring slat Wentieaton." Respon ‘fs were fist asked the following question: ‘Do you ink of Youre sr belonging to a-pariclar socal ast” Wr they Fesponded "ye they were che asked in an opewended fashion {Gien cis that I they su “oon dont Know, they were asked & lose-eategory follow-up "Many people sy that they telong tote working cl, the midle cat othe upper-middle Gass Hf you had o make choice, wbich would yousa) you belong {OP We are thar ale wih th set of guestone 10 tingunh je wit strong das idenlfstion (hose who answere yes {De intial question and thore witha weak cas Setieation Since we will be ung ths variable adjudicate contending stu {ual deinitons ofthe working case wil code simply a8 4 ‘Morking cae wentifcation variable Income: Justification In some ways income is less satisfactory variable than atstudes for adjudicating contending class defiations within a Marxist frameworks, even given the problems of using attudes and con- Sclousness dicinied above. While Marxist theory has sstt things to say about the distribution of income between cy labour, in general the theory ismuch less elaborated inthe analysis fof income mequality among wage-eares, Since the disputes in {Question all concern definitional problems among categories of Mtage-earner, income could therefore be considered sfatrly weak jon fora definitional adjudication ertheless, | have chosen to adapt income a6 a secondary criterion. Allo the class definitions we will be exploring typically Characterize "middle-class wage-carners as’ a privileged social1468 category. Furthermore, since the adjudications being considered all involve comparisons with an explotation-centred concept of lass, and sucha concept surely has systematic implications for income differences, its appropriate to use income inthe adjudica- Income: Measurement Respondents in the survey were asked what heir total personal income before taxes wis forthe previous calendar year. This figure ‘vas meant to include income from all sources —wages and salaries, ‘ate transfers, interest on savings an investments, ete There are three sources of potential eror inthis variable which may conceivably influence our analysis. Fist, a in most surveys, there isa relatively high rate of refusals by respondents to the income question, about 1S per cent. Second, the income datas for the previous year, rather than forthe respondents current jb, while our class-assignments are based on data forthe espondent’s present position. Third, the fact thatthe variable includes non- braze income mans tht it snot srcly a measure of the income ached to positions, but of income which goes to individuals, Whereas the’ adjudication logic directly concerns the postions {hemssives. Inthe use ofthe income variable in he adjucation of ‘contending clas definitions we must assume that these posible ‘measurement errors are random with respect t0 the critical Categories sed in the analysis. Ift should happen, for example, that there was much greater income mobility in one of the lass ‘categories used in the adjudication than in the others, withthe ‘esult thatthe income for that category was based dowaward, this Could conceivably affect the soncusione we might draw Ida not think that these biases are in facta problem, but they should be kept in mind, OPERATIONALIZING THE CLASS STRUCTURE VARIABLES The Explotaion- entred Concept of Class Srucure “The conceptual map of clas relations adopted in this book is fxirky complex. It is based on three principal dimensions of exploitation felations—esploitation based on control of capital, organization tnd credentias/skils~ combined in various ways. The essential Empirically Adjudcaing Contending Class Defintions 149 task of consritng class typology’ consists in opertonaliin ‘ich ofthese dimension, and then combining them. _ “The sratogy Ihave adopted isto casa the relation of each respondent othe relevant sets int three categories (1) clear 4m expliter wid respect to that set; (2) lealyexpoted with ‘espect o that ast; tnd (3) ambiguous The ambiguous caesar), in ths context, is ambiguous forone of two reson either the Fespondent geuinely appears to occupy a marginal postion within the lations of exploitation with reapet fo that act, or we lack Sufenlyprecse data to dearly define the respondents ation, ‘The ambiguous caves are_thus a combination of intermediate’ pesitions—pestons which may be nether exploiters. nor xpoted wih respect tothe abe in question and measremest ror. Throughout most ofthe analy therefore, our atenton ‘Il focus more on the polarized locations than the ambiguous "The basic operational riteria used foreach of the thre dimen- sion of explotaton ae presentedin able 5.2. Without going into ncesnve deta, few comments clarifying these operttonaliz tions wl be hep 1, Assets in Means of Production. Differential ownership of assets in the means of production generates two principal classes in ‘capitalism: workers, who by virtue of owning no meane of produc tion must sell thelr labour power on a labour market in order to work, and capitaliss, who by virtue of owning substantial quan lites of means of production are able to hire wage-earnes to use those means of production and need not themselves work at al. These two categories constitute the traditional polarize classes of the capitalist mode of production, ‘These polarized classes, however, do not exhaust the cass posi- tions generated by unequal distribution of capitalist asete Three ‘ther sors of clas postions ae also potentially important. First of All there are people who own just enough metns of prodicton 10 ‘eproduce themselves, Dut ot enough to hire anyone else. This it the traditional ‘petty bourgeoisie". Secondly, other persons own some means of production, enough to provide for some of theit Subsistence but not enough to reproduce themselves, thus forcing ‘them to als sell their labour power on labour market. This is he classic ‘semi-proletarianized wage-carner of early capitalism (and the parttime peasants of many third-world countries todsy). Anditeriateroperaonnaiont plains ene ssc anit ee. oo oer iam het Sed i ety Soups wang wo oe Dey mae Sari wit pel y deo eonoer Sort ommcanon sonar Tilo s apn wa can ey mae pty ens Espira os claro eras ee Apes le wapein ete! debate 2. Magl——_ Slten reg ved a hip acon degree Empirically Adjudicaing Contending Clase Defiitions 151 finally there are people who own enough means of production to hire workers, But not enough that they eally have the option of ‘not working a ll Tis isthe small employer employer artisans, fall farmers, shopkeepers, cie.—who. work tongside.theit emloyees, frequently doing much the same kind of work a8 the people they hire inthe data used in this study we cannot rigorously distinguish all ‘of these categories. In particular, the only data avalabe to ds {uish small employers from proper capitalists is the number of| ‘employees ofthe respondent, and this tat best a weak indicator, Since it does not really measure the amount of capital owned by 3. For present purposes, therefore, T wll adopt & tary convention, and define al employers employing ten or more people as fully-ledged capitalists, and employers em Ploying between two and nine employees as small employers. The petty Bourgeoisie defined as any self-employed person employ. Ing no mare than one employee. We will not, in the present dala analysis, attempt to distinguish fly prolearianized wage-earners {rom seine-proletarianized workers, although on the bass of data about second jobs and jobs of other members ofthe household we ‘il be able 10 Introduce this distinction in subsequent work 2. Ansets in Organization. Organization asets consist in the ctfctive control over the coordination and integration ofthe dive Sion of labour. Typically, such assets are particularly salient in ‘defining the explotation relations of management, although not al jobs which are formally labelled “manager involve contol over ‘organization assets. Some “manage” jobs may simply be technical experts who provide advice to the effective controllers of oryaniza- tional planning and coordination, In the terms ofthe explottation- centred concept of lass, such "managers’ might be eredeatial- ‘exploiters, but not organization exploiters ‘Wi respect to organization assets, we will distinguish three basi postions: (1) Managers: positions which are directly involved in making policy decisions within the workplace and which have effective Suthority over subordinates, (@) Sipervisors: positions which have effective authority over subordinates, but are not involved in organiational decision:132 ‘making, These postions I shall reat as having marginal organiza- tion aes "G) Non-management: positions without any organization assets within producti, 3, Ase n Credents, Assets in credential are quite gift onaie ins nuanced way. On theface ot, might seem thatsimply sing formal academe quaifeatons might be sasac tory. There are two basic problems wih such a seategy ist tchuse of the sap expunion of educaon over the ast (Wo feneratons and the chaning formal education requirements for Ertan kins of jobs, any formal credential arable Would have (0 ftvole enor spot eredentals and some provision for histor {ll devatunon of redenals over tine Second formal ceden- {ial only becomes the base for an expovation flaion when it Inatehed with job that quires such credentials. person with octorate im Englsh who drives tania not @ credential ‘Seplover, What hr inpis is that in order o spe property the ‘aplotaton relations Bit upon eedental aces we must neude information on the actual Joo person hols and not simpy on that persons formal academe cercates" “Ts immediatly pores an adionl problem: many job sides and ocupational designations ae extremly vague wih expec 10 the credema hey demand Tis nox partly problem for professional occupations, but certaly i forth wide range of Rage” jobs snl even forsale and rea jos. Some sles jobs requir engineering degrees an ae, a praca tems, more {ike an engineering consltany than a simple slesperson jo some manager jobs, onthe other hand, require no particular cre ‘Eola a hey mip sal conta explers with respec fo ‘rzanization ase, but notin relation to credentials of kl. Sine jos erouped as sea” occupation involve high level of training and experince, others require very hile.” Even the etal occupational tos donot always dings these cru Sances ina stsfotry wa. ‘We wil solve this complex of issues by vsing a combination of ccrupatona tls formal credential and jb fats a bas for Secingushing people certain credentials are mane Gstory—and' hus postions involving credential sset exploit tion from thove notin such jobs As inthe other assets, we wil tuo define an intermediate Staton in which its ambiguous Empirically Adjudicating Contending Class Definitions 153 exactly what credential assets the iadviualsctaly controls. This yields the following thre catepories: (1) Esper Tis incudes (2) al professionals; (0) etnicians aunt manager (ty oesopational te, nt by the eile used to sine the organizational ssets speciied above) with colege ‘egres. Ske employees: (3) shoot eacers and eratworkers (0) manages and echilans with les than colloge degrees (Cle. permon or eis wih clege degrees and whose jobs hve Teal Sitonomy. (G) Nosed: (2) ceria and salespersons not sting the extdenil or autonomy criterion for sked employes; (b) non ‘aft mancal occupations and sevice oocpatons “Taking these three sts of exlotation-asset criteria together gen- crates the overall map of ea Tocations dplaje in table 33 thapter three ‘ur focus inthis chapter i aot om this ent marx of elas locations, but onthe deintion of boundary ertera distinguish the working dls from the "mide dats"? As mentioned carer, this ates the question of how to tea these oargina categories in the adjudictons-partealany those cells ta table 3.3 deg nated ‘marginally credentialed workers” and “uncredeatalled Sipervisors= The procedure I wil adopt wil Be 0 include these {wo marginal categoria a dsint category in the analysis Ci thins out tha for practi purpones these eategris can Bete {ted ss basally similar to workers, then in subsequent analyse ‘ould be jstiable to morge them with the more restctvely ‘efined working eas. Manual Labour Definitions of the Working Class Even though manual labour definitions of the working class are certainly the simplest, it is not trivial to define rigorously the {appropriate criteria for distinguishing ‘manual from ‘nonmanval’ Inbour. ‘The conventional approach is to equate this with the purely ideological distinction between “bive-ollar” and "white olla? occupations, as defined in popular discourse. But this has the effect of puting a variety of highly routinized clerical jobs—key-punch operators, typists in lange semi-automated oices, et which in real ferme involve less "mental labour’ thta4 ‘many siled artisanal jobs, into the “mile clas 1s precisely ‘because of thi kind of ambiguity that many theorist ae hesitant to adopt simple mental-manval distinction as the basis for defi Ing the working cass “in spite of these reservations, I will adopt the conventional biue-0 112A. The difrence Jo dan sade between the doped eatery and ‘ctr oes it pny teas Han cn ee Feat an 3] Tec? ~ cen 6} <0 1U2R, Te diferent io cl sie Setwces the depute’ category snd oeetsapen wor woe spay mare han betwen he 2 PENT ey Tat atne) > 0 1, Addin of man lbow defo ype TILIA. The itfernce in come beeen dpe separ 1 (2) an the ares apo wore wil ei the Detwen them and the aetapen mid cs (ena cal] “Teet'2 el 6 <0 TIL1B, The diernce in inne Deven, dipted clean) band the ‘Bensapn cre a anonym en [ata] Seat ean 6] >0 Empirically Adjudicaing Contending Clas Definitions 159 TABLE 55 omtnued) 128, te direc in incomes between ped category 2 (cs) a he ‘scedspon worn wb Siguanty more tas betwee ih sae teedpon mde'dasn Feet ean es 3~ cet 6| =0 N28, Toe ieee incomes betes dipued aegry? snd be ‘don werk eile nay let hn Setter be de (Eaten stots cue) <0 ma me eer ay a te ts ee > ers en ret ae ats (aif Py seo won gc tee et ey ay arene sn ins ee et westerns as NEAT esca 2A es w We pti to he ion pay Stat tte pastoral ype i abe $5 thes Ryo ei tated) epee pein om he extn ceed coop ve ‘Ssondhypethess(eaarted 8, represent he pecan fo he a ea [Note on Statistical Procedures ‘The data for the United States was gathered in a national tle- phone survey conducted by the University of Michigan Survey [Research Center in the summer of 1980, Respondents were sam pled on the basis of @ conventional two-stage systematic cluster Sample of telephone numbers inthe cotermins United States, The fist stage consisted of sampling clusters of telephone numbers. I160 the second stage telephone mbes within sters were random eegHindy. wtin howshold, lle respondents were eSeY atrandom The resting sample consis ofa foal of 1850 adults vere age ofsxezn working in the labour force, 92 ator force and 170 Howseives for a toa of Wet capondents. The tesonte rate way about TS per cet, & {ay peal at for hs Kindo survey. Throughout tis book w= Sil ef analyse the working labour force sample “Tne Sweunh seme (not ted nth ehaper) consnts of 1145 adult bosween the ages of eighteen and sive selected 0+ ony fom a nation Ist of tbe population. Respondents wet arty ated's qcstonnie, and then, i they did not send eters contacted by telephone” The overall response rate tas about 76 per cen. ween need to be sal about telephone interviews, since dle nfs with survey esearch may be somewhat cepa EOP hai of such neriews, Research which has con 20% tlepnone td personal imerview has shown that ther are pasta ferences m respons fo questions sing che 60 Teaiqucn there mc, however, corti advantages an dk anna eat One the one han, personal interviews allow fr anspor completed questions priclay questions that race wnal ade of rariovs srt Telephone interviews tend (0 TEAS he gucnonnle oly simple questions, On the ott Sod eal ere re vy moe expe hn nerve, require much more eusteng nthe sample rooms stander espace (aleat i the Uaied Sats) thay SEREE: Ynote based sample since many people willbe wing see asoungt on the eephone who would not be wing 0 {Ecihem ito tne hows, In any event, for beter or worse, the Rist ue Unted Sate in ths stad come ently fom ile fone interviews weionts For reasons which are not entirely clear to us, the educa- ‘Son-by-occupation distibution of respondents inthe United States ample fb somewhat more biased towards higher status occupe ‘Wont nd higher leves of eddcaton than one would expect fom a furve) ofthis type, Some of this i typical of telephone surveys, ‘Snve only eround 95 percent of individuals inthe United States live indwelings with elephones and non Since son bats could aft some of the Crowe-natlonalcompatisns we wil be doing In chaplers sand Seven and ice they might aloha effects on te dat analysis in this chapter I have aplied a et of poste weights tothe data hich fave the consequence of reproducing the 1980 census dicaion-by-ceupationdstbutions inthe data we wil be wing ‘The weights ave dengnd in such way that the foal Nin the ample & not effected by the weighting sytem, Throughout the nals inthis book we wil ue the weighted dat. STATISTICAL TESTS “Throughout hs anys Tvl ely on fy smpe statistic tts Wel prmary be enamine aiflerences im mesnsFetween broaps, and therefore wil tse conventional ‘ess ex the Stal sinficnee ofthe dfeences observed. Sine not ll ‘eager il be familar ith ten rit Word neds fo Be thade about how they shouldbe ittpreted and how they are Geist Tatu suppose we record informatio frm samp of workers and superviors and om te bas ofthis infrmation, we exmate that workers havea meen income of $1,000 an superiors ave ne of $6,000, What we want to text a wheter the diferencia thew observed incomer $500 significa Signin, in ine Sona senct st ho Sit wee dee nerve tence seal dren fom 20. Is alway pose ‘heater all that so group cing compared coun reaitybave {oie ncomes, ot ht Bea of random vara n gator ing the dats we might obser ilerence, When we stat he ‘Sheree ference spitcan athe Oi level what we mean tha bused on ceransntcal spon, or bet ges that Inky oe ot ofa hundred sarveys could th ages ference ‘be observed when the real difference was zero. rc “he ten proce defor performing this Kind of tt nvives calling what nada ste: Te cleult he we have 0 ive or estimate of the sifernce in means Setwecn the tro froups by what ale th stander error of thi iference. The Siac he sandra error relative tothe diferens inte means theless kely west be ery concen tha the observed ier162 ‘ence in means reflect a true difference. How isthe standard error itvelf measured? Itis based on two pieces of information: st the Sample size on which the observations have been made, and Sec- ‘ond hat is called the ‘standard deviations of each ofthe means ‘A'standard deviation ie basially« measute of the dispersion of| ‘alues around the mean, If everyone in a sample had identical Incomes, for example, the standard deviation Would be 2er0; where incomes are quie dispersed, the standard deviation wil be large. The larger the sample size andjor the smaller the standard deviations (relative to the lferences in means) the smaller wil be the standard eror. "in more technical ters, the statistic used to test the signii- cance of differences in means between Iwo groups i calculated 2s follows: ean of roup 1) ~ (Mean of 4p 2) (Sita dovaion of aroup 1, Sander deinen op ‘Gomplese ingroup)” ample sein up 2) “The larger the valuc ofthis satstic, the more confident we can be that the observed differences between groups reflect tre difler- fences in the world rather than chance diferences in our measure ments. From the formula itis clear that there are two ways in Which our confidence in an observed diference between means am be high: first, if the standard deviations of each group are Small, and second, if the sample sizes are larg for each group, ‘With avery large sample, even ifthe values within eech group are quite dispersed we may be quite confident that a relatively small iterence in means is not just a random res ‘tests ca bo used in what ae called one tests A two-tailed testis used when you simply want difference between two means exss, but you have no pr 7 {ations about the direction ofthe difference. A one-taled test, on the other hands designed to text whether the mean of one group is greater (or smaller) than another group. In general we wil use ‘one-tailed tests in ur analyses since we have sttonga prior expoc- {ations about the directionality ofthe differences in question. "Most ofthe hypotheses we are exploring are not simply about the differences in means between groups, but rather concera the bitferences in cfferences between groups (he hypotheses under I ‘and Ill in table 5.3) In such eates the use of the t-test becomes Somewhat more complicated, Thies because the usual assumption Empirically Adjudicating Contending Class Definitions 163 ‘of a testis that the groups being compared are independent of teach other. This assumption holds or the comparisons of the dis- puted eategory with workers andthe disputed category with non- ‘workers in our adjudication of the definition ofthe working lass, but it does not hold forthe comparison of the two differences, since the disputed category appears in both of these. What this ‘oan technealy i that when we eleulte the standard error for the difference in differences we have (0 include a term for the ‘covariance’ of the wo diferences. Tis is accomplished by the following formula (in which se. means standard error) Soxiologists ae often prone to ftishze si more attention to them than to the substantive meaning of tats ‘al results. Significance tests ae sretly measures of one's conf ‘dence that the observed results sre not random, but its stil the results themselves that should be of theoreteal interest. While T ‘wll ely fairly heavily in places on the statistical tests to add per- Suesivenes to particular arguments the real burden of the diseas- sion wil be on the substantive results themselves and not on Spnificance Tevels as such, Empirical Results ADIUDICATION OF THE PRODUCTIVE LABOUR DEFINITION OF THE ‘NORKING CLASS ‘The basic results comparing Poulantzas's definition ofthe working class andthe exploitation-centred definition appear in tables 5.6, 577 and 5.8, We will bevin by examining the two hypotheses held in common by both contending definitions of the working class and then turn fo the substantive adjudication using the empirical pre- dictions of each definition Common Hypotheses ‘A precondition forthe adjudication strategy to work is that the ‘agreed-upon workers and the agreed-upon ‘middle class difer in the expected ways onthe dependent variables which are to be used164 TABLES Adjdicaton of productive labour deiniton of the working cas: income ri Tate (standard deviation) Netter eer (on) oe eee Te wen eS a ; eas A 7 we | og] as sag whe [ah | ase = ee” | | a : waco | tedlen | a | ange was | Baas i as 185 Empirically Adjudicaing Contending Class Definitions 165 Adjnicaton of productivelabor dtnition of the working clas: ‘dacratitude sale (Geanaatd Berton) santero se (wee) gen Peter atin Sills voimar awe Se tte a we] ow 2 i % % sa B Ps ‘ae FD & (empty ce -127 snow |e & ge seats sv an ne aint Set net Cour Sean te "EI wo ot ie inden he wo pnp aca a eS ae ae oe SO Sa lh hcl tet Peg a al a ag ee ee Oe Cea Se ela eso each eet SOERIRCAY ih ita Neier “ase a an ciaty WED (aia posal Seige =e 1s be 9; Mal wont cis ~ ls 9. ‘observed differences are not due to chance. If, therefore, one ‘grees on theoretical grounds that these are indeed appropriate ‘iteria for adjudieating the contending definitions, then we can Shsume that there i at feast a presumptive empirical case that our fonorete measutes ate appropriate a8 Well Income Adjudication ‘The rests ofthe adjucaton using the fcome variable provide Te support for Povlmta's Seto othe working cass, wile they Se gute consent wih te denivon Ihave bee vancng. Wie duped categny shoud in realty be cessed with he {jecdupon idl! dase then we would eapet tha, Hike oxer166 TABLE St ‘Test of ajudeation hypotheses: pradactielbour Yrs exptaton defn ‘Seance Enpiset let Hype rete" ¢ fone) _Concion Troe -e>o S61 72000 suppor Promoting lander Penteo -290 76 000 spond Income Mik |= 21-12-51<0 suis no 090 ‘EAA. tony a }23|-(2-[eS0 Soper tL3B Proworking ls atid! maa (i [lz si poyecs (the hear of the dapated eatepor):sexsm, both inthe Unions themssies (preference for organizing men instead Of women) and in the employment situation (greater vulacrabilty of Female employees various Kids of contol by employes); the Ffapmentaton and dispersion of hie solar employes oie legal costes om orpaninng the state sector? and so on. Non ‘anual subordinate employees could be ful nthe working css, tnd yet because of sch factors, have dramatically diferent levels ‘fumonation. The fact that i some counties, ach as Sweden, the rate of unonization among white cll non supervsory emp. loyees virtually the same ass for mana! workers supports the ‘tions in levels of unionization between non: Siperiory manual and aon-manial wage-camers fs more the Tetult of polical and ideologies! determinants than of possible Siercnes in thelr lass location, tis interpretation ofthe unionization results is correct, then te level of dnionzation ix not avery satsfoctory adjacation ttiterion. Acording, while the snioneation ress do introduce Some ambiglty ito the analysis, neverices the overal weight tthe empiri findings lade litle suppor to definitions of the Morkng lass but around the extern of productive and no: ‘uete labour ADIUDICATION OF MANUAL LABOUR DEFINITIONS Let us now turm othe comparison ofthe definition ofthe working class as blue-collar-manual wage-earners withthe exploitation: ‘centred concept. While this definition is both conceptually and ‘operationally very simple, the adjudication is more complex than inthe case of Poulanteas's definition of the working dss. Ta the analysis of Poulantzis's definiton there was only one disputed ‘ategor)—positions which Tcaimed 10 be inthe working class but Poulantzas claimed to be new petty Bourgeois. Inthe case ofthe ‘manuel labour definition there are two disputed categories: posi. tions which I claim are working class but the rival definition claims to be middle’ class (mainly proletaianized white collar jobs) and positions which I claim are “middle” class but simple manual Ibour definition would regard an the working lass (maaly blue sola wagectacn in sperviory and deciion- mating js) The former I wil refer to a6 dspuied category 1 and the liter as disputed category 2. Out task, then, as charted in table 5.5, itovm np Aig Coen hs Deftins 13 mass pee santa men i tin es me at et iin tin a sau Sbodsed Deviations) (Standard devistions) eines wareitiee ci sun ea seuitierttim Tle atneee— ateam a The otga nea a 3 oe cgese | Bete i "3 = ar | 2 bb & | & S a ge | 2 3 7 vert | abe | ale | ag set a | ow, = er | oe | Be oa sy & | 5 @ 3 a Sees | Set eins | SoS Fe Hee | “HSS | tas, a fn i | & | 8 it | Ss cm 1s cast > a oe i okey Se = GER BAC Noe ning cs = is Pod ‘explore the range of hypotheses for both of these adjudications. “The basi results are presented in tables 5.11 and 5.12, andthe sata et forthe various adjeation hypotheses spp in Common Hypotheses As in the adjudication of the Poulantzas definition, the agreed- ‘upon workers and agreed-upon ‘midle" lass in the adjudication ‘ofthe manual labour definition dtr inthe appropriate Ways the TPS Ea Se sag m= fo0 8 ima pene Buchner = Ce 13 Table 3.3; Mag wring c= es 9 ond 1 ‘nei Ge = oe 8 an 1 agreed upon ‘middle’ class eatn on average over $10,000 more per year and sore, on average, 273 points lower on the pro-working- ‘ass attitude sale Income Adjudication ‘The results for the income adjudication are essentially the same for disputed category 1 28 they were in the adjudication of the76 Empirically Adjudicating Comtending Class Definitions 177 ‘Test of adjudication hypotheses: manual-labour versus exploitation definitions i — a pn tal te ae fren ot pple eT aad ate ale nae caerercean! “iets Dina peel Beet jones leet, Reet, ene Beere eset peeet cara eees | [== Prooring seas 4. Corea ben rote eae iao aa eee ome pordaatesr sth oscsast? ieee era eee at pecans ea eae NNO fear U1) pottitttoor itt age tae HQ Hg|saed woe oe ne Seo aokee Seeeeeeeeeeag ttt teeeeeeCas foment ama cpespepeecere EER ddd eee escape area UE el aah et MR |L-SI713"6128 one au an emergpnes ERM a ous oka mas {to3{=]828]25 cient \ ie 0 {Te Hypoespambes eespond io he aunts o Tale 6 "The nin pretest the sth Toles 80 and 1, productive labour definition: this category is clearly much closer to the agreed-upon workers than to the agreed-upon non-workers, The results for disputed category 2, however, are inconsistent with both definitions under scrutiny: the average income of this ‘eategory falls almost exaetly hallway between the incomes of the agreed-upon workers and agreed-upon ‘middle’ ass Aniude Adjudication ‘The attitude adjudication miors quite closely the income adjudi- cation, There is substantial support for the hypothesis that dis- puted category I is significantly closer to agreed-upon workers than agreed-upon ‘middle’ class (hypothesis LSA), and no Sup” port for either of the two hypotheses concerning the second di puted eategory. Again the result falls almost exactly between the two agreed-upon categories, When we look atthe ilem-by-item breakdown ofthe atitide scale ia table 5.14 we se the same basic fee NO 08 oo 02-038 Patmos" psy soe 04-0 Ina tite, tis orden 1s. oor ons 002 eneanin 3p dag NS 367 ao mss Speonorter 1 pro-sotgens,D = dont kaw). Sibi be noted dirham ut Yewenn 6 po cnt and 2 pe cant of he ‘oor ne in nha eg aca edn prone a178 pattern. The fist disputed category is clearly much more like the reed-upon workers than the agreed-upon on- workers on five of the items, its loser to the agreed-upon non-workers on only one item (item number eight) and it falls fairly much inthe middle on three of the items. The second disputed category, on the other hand, is closer to the agreed-upon workers on two items, closer to tReagreedupon nonwrkerson thes tem ad ight ithe mi The Effects of Gender and Union Membership ‘Table 5.15 presons the results forthe adjudication ofthe manual- labour definition looking at men and women, and union and non union members separately, These results are rather complex in certain respects. For the income adjudications, disputed categ- (01) 1 is closer to the agreed-upon workers in each of these coat- parisons, except for men for whom this eategory iat the mid-point between agreed-upon workers and “middle” class. For disputed category 2, on the other hand, the results are quite inconsistent cross comparisons: for men and especially for women taken sepe rately, this category & closer to the agreedsupon workers; for Union members, this category is identical to the agreed-upon "id= ig” class; and for non-union members it fall in between the ‘agreed.upon categories For the class-aiitide adjudication, the results are perhaps even more indecisive, While among women, the pattern i pretty much 48 expected (disputed category 1 closer to\agreed-upon workers tnd disputed category 2 closer 1o agreed-upon "mildle” clas), among men both of the disputed categories fall in the mic. ic, Between the two. agreed-upon categories. Among union ‘members, asin the evaluation of Poulantss's definition of class, ‘there sno clear patter—while disputed category 1 hes the high: ‘ext value on this variable al of the other categories have roughly the same values. Among non-unionized employees, on the other hand, both ofthe disputed categories fall around the middle Overall Assessment ofthe Manual Labour Adjudication ‘What sense can we make ofthese seemingly inconsistent findings? ‘Two things should be noted: fst, the dificulty mainly involves Aisputed category 2. In general the results support the propos tion that disputed category 1 i closer tothe agreed-upon work Empirically Adjudleaing Contending Class Definitions 179 _Adjoicaton comparisons fr set and union members categories taken ‘moeratly: mancatlaour sj ‘Aevatapon Dipund Dial Agrcapon ‘ining etry? tango mec Women 99, “ass “125 “oat Seite Creed et i ee ieee fetes ttt tag ttt aces tlaopes a essere eee tage eee Une 106 aT 9 a ee UU TA eae wed ers than the agreed-upon ‘mide’ class, Although there are some Instances in which thie disputed category falls close to the mide point between the agreed-upon eategories—for example the eon iousnese results for men--there is n0 instance in which itis closer to the agreed-upon “middle” class, With respect 10 pro- Tetarianized white collar employees, therefore, the data offer 80 ‘pport to the claim that they are "middle la and considerable ‘ppor tothe elsim that they are part of the working lass. Sec- ‘ond, with respect to disputed category 2in early every case the resuite are completely indecisive. In terms of the problem of “djicating beoween contending definitions, therefore, they sup- Port none ofthe hypotheses we have been entertaining, and thus180 {do nol allow us to dstnguish between the two definitions under investigation, (My guess is that there are two principal explanations for the results involving disputed category 2: fist, problems with the ‘Operational criteria adopted in coastrucing” the expoitation- ‘centred clas vatlables, and scoond, the issue of class biography. Tis, of course, easy to blame contradictory results on problems of measurement and operationalization. The universality ‘f measurement problems is one ofthe things about sociological research which encourages fsearchers to tlk themselves out of| ‘ifcultes. Nevertheless, 1 do tink reasonable case can be made that some ofthe anomalies we have observed are linked to meas- Uirement issues, Some indication ofthis ean be seen by looking at the marginal working cas’ category, clls3 and 4 in table $12. ‘The manual-non-manval demarcation within ths category shows ‘a sharp difference in sores on the attude scale: white-collar mar- final workers (largely white-collar supervisors and semi ‘redentaled white collar employees) havea value of ~0.50 0m the Scale while manual marginal workers (largely craft workers and ‘manual supervisors) have 4 mean of 1-44. This seems to indicate that many of the people in category three really belong in the agreed-upon worker category, and perhaps some of the people in ‘category four belong inthe agreed-upon "midale” clas. “Two measurement jsues are implicated in these posible as- sifcation problems. Fist, there is the problem of distinguishing between supervision that is really part of the management spporates and. thos partakes in at"Teast’ marginal Tevels of ‘otpanization-exploitation, and supervision whichis nominal, which {Emainy a transmission helt for orders from above. We have relied fom a series of questions about what supervisors can do to their ubordinates in order to specify this maypinal level of organization fst exploitation, arguing tht the ality to impose Sanctions on ubordinates isthe important in of differentiation. This may not, in fact, be a satistactory way of specifying the probiem (assuming, ‘of cours, thatthe basie conceptual statis of organization assets find exploitation is acepted), Some kind of minimal participation in cocrdinative decision-making may alo be necesary. With 2 ‘more stringent criterion for organization aset exploitation, most ‘ofthe blue collar supervisor that we have included i the "mari- ‘al working cas category would no longer be treated as proper fupervisore at all and would thus be placed in the agreed-upon worker eategory. This would also lead to a reclassification of a large part of sisputed category 2 into eels 1 and 3 ofthe table Enpirically Adjudicaing Conending Class Definitions 181 A second problem of operationalization concerns the treatment ‘of craft labour as having margint levels of skiljeredentia assets, thus placing them outside ofthe pure working clas. If they were also Supervisors, they would be placed in one of the unambiguous ‘pon working-lass categories. If craft bout had not been treated inthis way, then most of the individuals eurcendy in cell 3 would hhave been in cll 1, and many of the people in disputed category 2 would have bees in ell 3. This again would have substantially aifected the values of the depeadeat variables in these cells and Potentially affected the conclusions reached from the adjudication nals. Ts ualkely thatthe results displayed in these tables are entirely the arfacts of errors in judgement on the operationalization of concepts. They probably also reflec salient properties ofthe real ‘mechanisms at work in the relationship between class structure tnd cass consciousness. In particular, I suspect that atleast some (of the results are sgniicandy affected by the problem of cass biographies. Most of the incumbents of positions in disputed eategory 2—blue-ollar employees ia managerial and supervis ‘ory postions—have biographies that ace tightly bound up with the ‘ageed-upon working class category. In many instances they are in careers which bepin incall | ofthe typology and move gradually through cell 3 to cell’, and their socal es through family and fiends are likely to be closely linked to the working class. Simi- larly, many of the people in disputed category 2—protearanized ‘white-collar employees —are likely to have biographies tied to the {agreed-upon “midale-lass' locations. Indeed, this may be why thelr ideclogical stance is signfianty Jess proworking class than the agreed-upon workers even though they are much closer {0 ‘agreed-upon workers than agreed-upon "middle" ass,” Class con- Sciousness does not emanate from the relational propertcs of the Positions people fil at one point in time, Rather, i i formed {rough the accumulation of class experiences that constitte person's biography. To the extent that such biographical trajec- fories vary aross the ccs of adjudication typology they can com Found the adjudications themselves ‘Concusion ‘The exercises in this chapter have been designed to provide @ systematic empirical intervention into debates over the concept of|ae class structure, Two basic conclusions can be drawn from the investigation: 1) In the debate over Poulantzas's conceptualization of class structure, there is very litle support forthe view that productive labour is an appropriate eriterion for distinguishing the working ‘ass from non. working-class wage-carmers. Except forthe rates of Unionization of different categories in the adjudication, the is- Dated category was closer to the agreed-upon workers in every Instance. Atleast at far a this speifie adjudication is concerned, there is much more empirical support fr the stuctural deaition ‘ofthe working cass as uncredentialled non-manageral employees. 2) In the debate over the manual-labour definition of the working ‘lass, there i almost no support fo treating this division as a class Gistincion. Proletarianized. white collar workers are_generally more like proletarianized manual workers (ie. the agreed-upon ‘Workers in the analysis) than they are like nonproletarianized ‘white-collar employees. Iti less lear how non-proltaranized anual wage earners should be treted, but in any event the data {do not support the thesis that they are part ofthe working cas, ‘These debates will hardly be setled, needless to say, by the data and analyses we have explored. Defenders of the positions Ihave ‘etcized havea variety of avenues open for reply. Fist, ofcourse, they can reject the ene enterprise ofthe empirical adjudication bf contending definitions, arguing that definitions are strictly con ‘ventions and that their adjudication is therefore strictly a matter of ‘helt togical coherence. ‘Second, the nced for empirical adjudication canbe accepted, but the miro-individual lop of the empirical investigations of this chapter ean be viewed as inappropriate for adjudicating contend. ing lass concepts. If those concepts are meant to explain historical Urajectoris of struggle and change, then it could be argued, the ‘ata explored in ths chapter are radically unsuited tothe present ‘ask, This ba serious eiticism, and it cannot be dismissed out of hand, The rejoinder to such ercims i that even if the concept of| class structure is centrally preoceupied with such maero-historcal 1d dynamic problems, there ae, ate all, real people in that class SHructue, eal people who are systematically affected in various ‘ways by trtue of Being in one cass rather than another. Unless ‘ne is prepared to argue thet the effects of class on individuals are Empirically Adjudicating Contending Class Definitions 183 cop s.oaia a rg “sot "ee SLES SA anal Sat cine Soman te Sa corte Mee ese et eee Shc gil ss Si ai eaten RE TRE ts i pt ft icon of tops Ses en stat SMaeGton amie nie sponta eer ‘ersten el i reo Se ee ee sere Sas wea tae Hole sma teats pre pecans te i of npn ce ne, orm ams cu suet terete ate Same R's wey Sul alee ages son Sn lees naka Say Seot fet auSy ee oe pl Sipe ni oe uate184 less sharply differentiated from non-workers in general. The dis puted eategor, then, could be viewed asthe part ofthe new petty bourgeoisie to whieh workers are most closely drawn. The histor cl context ofthe data could be taken a8 the Diss for explaining hhow the results might be consistent withthe eass concept in ques- ‘These kinds of alternative explanations suggest the need for bistor- ical and comparative research to deepen the conceptual adjudic tions we have explored. Ifthe adjudication results were essentially the same in countries in which the working clas ie more cass conscious, more mobilized and organized than in the United States, for example, it would undermine the kind of ertique sug- gested above. If, on the other hand, the adjudications look very Gifferent in socicties with different historical context, then this ‘would sugges that the conclusions I have drawn need to be mod- Ie [As Ihave stresed throughout this analysis, there is no possibilty Of the absolute validation of a concept, adjudications are always among actively contending concepts, rivals which attempt to cap- ture the same theoretical objects. The conclusions established tn this chapter are therefore of necessity provisional, both because the defenders of the concepts T have criticized may effectively tespond in subsequent research and argument, and because new Alternatives to the conceptualizations T have. proposed may be produced in the future. One final issue to be dseussed, therefore, 'S whether the adjudication analyses we have explored suggest any ‘iretions for such future conceptual elaboration. What are the ‘anomalies in the data? What results point tothe need for further conceptual work? These ambiguities and loose ends fll under two ttegories, First, the question of specifying the criteria for the ‘working dass, and second the choice between concep of class bated on trajectory and concepts based on position. ‘Specifying the Working Class ‘While feel that inthe debates over the definition ofthe working lass, the empirical evidence is. most supportive of the explotation-centred concept, 2 number of results in the analysis Empirically Adjudicaing Contending Class Definitions 185 suggest that some further refinement is needed. Above all, there is indiation thatthe logie of the credential-explotation criterion ‘needs further work. This issue played a pariculaely important role {nthe ambiguities in the adjudiation withthe manual-nonmantal definition of class siueture, especially around the westment of ‘raft labour as marginal credential exploiters. ‘At the hear of this problem of the stats of skil/eedentil sets in the analysis of cas structure isthe lack of clear relational ‘iter linked to the ownership of credentials, The ownership of ‘capital corresponds to a socal relation between employers and employees; the ownership of labour-power assets in feudalism cor- Fesponds tothe social relation between lords and serfs; the efee- tive contol of engsnization assets corresponds to the authority relations between managers and workers. There is no such rel tional correspondence to credential aset exploitation. This fx one ‘of the reasons why the precise allocation of people to class pon Hons seems much more arbitrary with respect to tis asst than the others, and why there are particularly sharp problems in the Uteatment of erat labour. In order to reduce this arbitrariness in the operational use of the concept of credential aset exploitation Inclass analysis, additional theoretical clarification is needed. Class Trajectories I of the adjudications we have explored in this chapter have been between postonal definitions of class structure, that 5, fle te diernce betmcen he erence (Theron fe thesia ich ‘Seite we pen ate we ire ice eee ‘Byte appro corte fr tele ay The snd er of ee ‘Se rd Chase Hay fer Swipe spy a pe apc gl tl aes ote “f Toe ator he Sifeeocebewcee workers and he pte eng ‘be diterence he cterene 24, whe ara ebay 0s ta ‘onze the are ores 6 percent cattery itso mac renter for men tr omen foe ae 1) spore thst, sine arouse protaraned fr wre a ‘Seley feces ssc in hem were Shien ane utc the sre mehedlgy a vali ectgu for ote igiiornionce desc nr nega Se a) itera esi, se rer ave inp, then re hold not be ‘cn song decor beeen oe rac ctegoercmplyes ia jt Enplrically Adjudicaing Conending Class Definitions 191 cin nA mano elton tie a todd soya Sus toate arta ints igo ings beeen cy i nee as Ria cc csr cerca Spee prs tins nonin ptr st emg Sabina i Aaa et pera res STieetedae opine ans Tarpon peas Sr tagcat oe pte ay ai ae Seahena ie cereus ae haar ATM ast Say na toe a33. Erik Olin Wright, Classes (Londo! Verso, 1985) Class Structure in Contemporary Capitalism A Comparison of Sweden and The United States In this chapter we will explore a range of empirical problems oon- cerning the cas structure of advanced capitalist socetes using the txploation-centred coneeptalization of class relations. Although ‘ve will test formal hypotheses in a number of places, most of the Chapter willbe largely desriptve in character. There have been ‘ery few systematic empirical stues of class structure from a ‘Marxist perspective, and none using the exploitation-centred con ‘cept of cass elaborsted inthis book. Tis therefore of some impor tance to improve our descriptive maps ofthe class structure, since the concept figures in so many dliferent kinds of problems Marsists study. This willbe our basic goal here. The data analysis wil revolve around systematic comparisons between the United States and Sweden. Within the family of highly developed capitalist countries, Sweden and the United States represent an importent contrast On the ope hand they are inmany ways rather similar economically: they have roughly the Standards of ving very lle state ownership of indus tion. On the other hand, politically they are in many ways polar ‘opposites. According to one estimate, 6a fesult of state policies Sweden has the lowest level of real income inequality after taxes fnd after transfer payments) of any developed capitalist country, ‘ile the United States has one ofthe highest twe take the ‘ofthe teal income atthe ninety-fifth percentile (othe real income tthe fifth percentile, this figure was only about 3:1 for Sweden in the early 1970s, whereas in the United States was 13:1." Sweden has higher proportion of is ein labour fore dirty emp- Toyed by the Hate than any other advanced capitalist nation, well ‘over forty percent, while the United States has perhaps the lowest, lunder twenty per cent. Politically, Sweden has had the highest 192 (Class Siructure in Contemporary Capitalism 193 level of governance by social democratic partes of any capitalist county; the United States, the lowest. We therefore have two ‘counties with roughly similar economic bases but sherply differ. ‘ent politcal ‘superstrtures From 2 Marxist point of view this is fertile terrain on which to explore the problem of elas structure and its consequences. In chapter seven we will focus mainly on the consequences of ‘lass structure for class consciousnes. In this chapter the focus of attention will be on the class structure ite. The investigation wi begin with an examination ofthe basic distribution of the labour force into clases. Attention will be paid to the relationship be- tween this cass distribution and sex, race, industrial sector, she of employing organization and the state. The following setion will then aitempt to explain the observed diferences in cas structure by decomposing these diferences in various ways. The next se tion will shift the focus from individuals as units of analysis to families, The basi question willbe how families are distrbuted in the class structure, with particular attention to the problem of classheterogenety within families Finally, the chapter will con- ‘clude with an examination of the relationship between cass struc {ure and income inequality in Sweden andthe United States. Since the concept of clas used throughout this analysis is rooted inthe concept of exploitation, there should be a direct relationship be- tween our matrix of class locations and income, (Cass Distributions Before looking st the data,» brief word on operationliation is ‘needed, In general the eri we wl employ in opeationalizt- Ingconcp te ds src ee he wo canes sts are studying. The one exception involves the specication of ‘credential sats. Because ofthe nationally spose meanings ot sven academic qualifetins and the historical evolution of the Felstionship betweon diferent Kinds of eredenis and the labour ‘marks, it docs not make theoretical sense to adopt mechanically the same formal academic depres in the specication of this exploitation generating aset. On the other han, there te on Ssderable kein undermining the stick comparability of the ata if diferent credentials se adopted ss eniersn for diferent coun tries Ideally one would like tome dive messure of the sil Searcy of labour power isel, but I do not have & lege Wea Bow18 this could be measured, and there certainly nothing in the data ich ean pls be ase to measure tis Ta brdarto balance these various concerns, it seemed advisable to hit the optational eter for cedental eases for Sweden Tescad of vg sole depos the salient etetion for ai {erotiaing diferent loves of credential ast for thse occupe- tins for whi ste necessary (ee tale 5.3.0 9 the previous hapten) a highschool degree adopted a the erterion, While Scr inay ben the process of becoming more Ike the United Slats tn hs regnrd, unl recently highschool degree Was @ much more subset nd importa eetlcaion In Sweden than Ine United States, A much smaller proportion of people went on {0 the Uniersiy, and'a universiy degree way pot considered smeesay Yor range of highly sled postions “One fal preety comment on he expontion ofthe ress Plouphing through masts of etl stata tables can often be 2 edieusand cumbersome ali The problems compounded in the present ese because the complex ofthe clas pony beng medrol eaegoris tm sltvand the logical suctre’ of the Ibpology a matiirsan make te tabular peseotation of ess ‘Miho Uawseldy have therefore adopted the flowing sae. omlet data bles for ihe vanous substantive topic inthis chap {erippear in appends It Inthe body ofthe chapter 1 wi follpes end sinpiy the class typology in various ways, tlloring thetabls to the desrptivegenelizaons I wish to emphasize in Ihe text. (OVERALL. CLASS DISTRIBUTIONS Let us now turn to the data analysis. Table 6.1 presents the di tsution of people in the labour force ino classes in Sweden and the United States, In broad contours the two class structures are ‘ery similar In spite of their vast diferences in levels of soi Inequality and patterns of clas formation, the basic distribution of people inthe cass structure does not vary dramatically between neve two countries In both counties the working cass s by far the largest class numerically, around 40 per cent of the labour force. If we add the contradictory lation with marginal control ‘over organization or sil assets, this increases to about 60 per cent {nveach country In both countries the bourgeoisie and petty (Cass Siructare in Contemporary Capitalism 195 istibtion of he labour fren the clas matrs wang the xpotaton centro’ ‘oncept of clase" “sss he mans of proton vee Nonomess (eage boues TBowgedie] [eigen [7 semi | 10 Une, vs iw] lus aoe] us. eae lus, 23% superiors | Superior |>0 arom us em} lus arefus. cae Jus. om | 3 Peey 6 Bape | 9 Semice- [12 Protein vs em] lus aae}us. aaylus soe Seosen Sim | [Soden 6a | Sede 173% | Sedan 13.5% Stu rdei e ‘Psion of role wei ear fe, th eg noe “For operation of he eter for asin thi tae, ee Tle 3. bourgeoisie constitute a very small proportion ofthe labour force: around 5 to 7 per cent are pure petty bourgeois and another $ to 6 percent small employers bot countries, and less than 2 per ent fully-fledged capitalists ‘While the basic outlines of the class structure ae similar i the to societies, there are some diferenes tat deserve atenton. First, although approximately the same proportion ofthe labour foree in the two eountries11 t0 12 per eent—oceupy proper196 managerial positions (positions involving organization-policy decision-making), there are significantly more supervisors (n03- ‘decision-makers with sanctioning authority) inthe United States than in Sweden: 17.4 per cent compared to 10.1 per cent. This ‘contrast is particularly striking for people without substantial re ‘dential assets. Non-expert supervisors constitute 13.7 per cent of the US labour force but only 6.3 per cent of the Swedish labour force. 1 will offer some interpretations of this difference in the level of supervision in the two countries inthe next section ofthis chapter where we try to explain the diferences in the two class Structures. For the moment the thing to note is that work isthe United States appears tobe significantly more supervised than in Sweden, a ‘A second point of contrast between the two counties isthe working class: While the working oumtries, iis somewhat proletarian with sem-credentaled workers (cells 9 and 12 in he typology) the Swedish working class is about nine percentage points larger than the American (61.3 compared to 52.1). Most of| this diference, as we shall see later is attributable to the higher levels of supervision in the United States. Third, if we look only at people with high levels of eredential/ skill aseis~experts of various sorts~a considerably higher pro- portion in Sweden are completely outside of the managerial $pparats: 43 percent of Swedish experts have no organization assets, compared to only 31 per ceat of experts in the United States, The reason for tise aot that there are fewer expert ma ets or expert supervisors in Sweden. To the contrary in Sweden {here are slighly’more people i such postions than inthe United States. Rather, the reason is that there are more non-managerial experts—about twice the frequency of the United States. Finally, while in both the United States and Sweden the vast majority ofthe labour force are wage-Tabourers, there are slightly ‘more seltemployed inthe United States: 14,7 per cont compared 10 10.9 percent of the labour force in Sweden.” If we add to this those wage-labourers who have at some time inthe pat heen sell: ‘employed—13.8 percent ofthe labour foee in the US and 6.7 per tent in Sweden the proportion of the labour force with strong petty: bourgeois experiences is considerably larger inthe United States than in Sweden: 28.5 per cent compared to 17.6 per cen. While becoming a capitalist remains largely a fantasy for most people in both socictis, there are mare people who have atleast Class Structure in Contemporary Capitalism 197 tried being st-employed in the United States, snd this may have Important ideological ramifications, [As one would expect, the class distribution among men ‘and women diferssharpyimboth Sweden andthe UnitedStates Gee {able 62), In both countries women in the labour force ate dispro- Portionately in. the working class, while men are dispro- Portionaely i exploiting class positions, particularly the capitalist ‘ass and managerial positions. The result i that in both counties ‘women constitute a lear majority of the working class just over 160 per cent of all workers are women. Even if we add the ‘marginally-cedentialied employee category—which includes a fair ‘numberof highly skilled craft positions largely fled by men—to the pure’ working lass iit the case that & majority of work isrtin clases within sex ategore, Unie States and Sweden? United Ses Swen Ud Seren la cars “Men Women Mon Women Men Women Men Ware TEmines 7 303 G1 469 Wl S262 2 Y payee $03 ae? 87 WS Te 78S do SMocges tg et 5 BS ass hea 3 2p foo del GS er S78 Iniapemios 627 362-710 290-43 BB 267 129 Stones $77 23. S61 89 30 396m ga o Rlisltren Me tet A Mee 68 Td 202s Sone 93 as Soe 2 80 ak hao Se THIS ns 2s te 4 as6 99 str wa Orca iol 54345756040 Fee eT Ge ee REET SMa Sey Gane amet tacit198 crsare women in both countries. The image which stil present in ‘many Marxist accounts thatthe working lass consist primary of tale factory workers simply docs no hold true any longer (if one adopts the concept of lass proposed here) ‘Looking at the distributions the other way around—the class isuibution within sexes—approximately one third of all men in both Sweden and the United States are clear exploiters (managers, experts and employers), compared to only sbout one ‘ith of women. Over half of all women in the labour force are in the working class, compared t0 only about 30 per cent of men, “The one thing that was not anticipated inthe results in table 6.2 hat the degree of sexual difference in class distributions i rea= ‘erin Sweden than in the United States. In virtually every position ‘of exploitation privilege, women are more underrepresented in the Swedish elas struetore than in the American, In Sweden, the percentage of men who are employers i 3.9 times greater than the percentage of women who are employers, whereas inthe United Sates the figure is only 1.9 tines greater, and the percentage of ‘men who are ether expert-managets or semivcredentialled mans: fers is 4.8 times greater than the percentage of women in these positions in. Sweden, whereas in the United Stites the over- Tepresentation of men is only 2.8 times. Following most of the popular prejudices about ‘enlightened’ Swedish social democracy, Thad expected thereto be less sex-bas inthe class distribution in ‘Sweden, but this is clearly not the eas, atleast not sczording #0 thie data ‘While its beyond the scope of the present analysis to investi- gate in depth the actual process by which men and women ate fitferentialy sorted into classes, we ean get a first glimpse atthe process by looking atthe clas distributions for men and women Within age groups. As table 6.3 indicates, in the United States the Proportion of women who are working cass does not vary sub= Stantally indifferent age groups between age twenty-one and sixty-five, Among men, on the other hand, thee is 8 clear age pattern: the proportion in the working cass declines until middle lige and then rises slightly among older men, The age distributions among managers fer even more sharply between men and ‘women inthe United States: an increasing proportion of men are managers as we move from early stages in careers (0 mid-career, ‘whereas for women there is a monotonie decline in the proportion in managerial positions as we move from the 21 to 25-year-old ed proup. Ta Sweden the contrast (Class Suture in Contemporary Capitalism 199 between men and women is not quite so clear-cut asin the United ‘States, but a basically similar patern exists: men appear to have & much sharper age profile for managerial postions than do women, Fising from 7.8 percent in the 21 to 25.year-old group to 19 per ‘ent in the 36 40 45-year-old group, compared to virtually n0 thange for women, about 5 percent in both groups. "These various age-class profiles within sex categories suggest that men have mach greater probabilities of promotional mobility from working-class positions in(o managerial positions than twomen doy particulary during the early and middle stages of ares. Of obure, the patterns in table 6.8 area complex result of the intersection of career patterns, transformations in the class Clas stebation within agesex eateries? ores Manager arnt 367-0 a 33 aa aS Bi ios oerss aD BS ate was" a3 Tso ia 3a Sas HS os Bo 38 "For compe da se Tate 12 pend200 structure and changes in rates of tabour-foree participation, For ‘example, how should the curvilinear relationship between age and the proportion of men in managerial positions be interpreted? Ix ‘nlkely that tis the result of demotions of managers at the end oftheir eatees, Rather, one would suspect, that tis reflects the Intersection of two causal process: fist the carer trajectory proces in which promotions into manager positions occur nthe Fist hal of men's career, so that by the later part of one's career itbecomes relatively rarer to be promoted from non-management to managerial ponions and sscond, a historical cohort dynamic, in which the probabilities of becoming a manager have increased overtime (asthe relative number of managerial postions has expanded), The first tendency would mean tht the proportion of men in managerial positions would increase with age (although at 8 decreasing rate beyond mid-career; the second tendency would tmean thatthe proportion of men in managerial positions would Aecrease with age. The combination of these two tendencies pro- ‘luces the ciiinear relationship tn table 63. Given this ind of complexity tis nota simple statis tsk o demonstrat conel- sively that he eilferenta outcomes for men and women obscved in this lable are primarily the result of gender discrimination in Romans, Nevis a provional cols, saps: thle hypothesis that ths a substantl contributor tothe gender Siteronces in las tributions, . CLASS AND RACE, Because of the cal homogeneity of Swedish society, it is not possible withthe data at hand to explore the issue of class and race In Sweden. Table 6.4, therefore, only presents the data for the United States. The pattern of racial difference incase distribu tions is if anything, mare pronounced than the pattern for sexual differences. 59 percent of blacks are inthe working class, compared twoonly 37 per cent of whites; at the other extreme, about 16 per cent of whites are employers or petty bourgeois compared to less than 3 percent of backs. These racial contrasts become even more marked when we break hem down by sex: Neany 70 per cent of black women inthe labour foree ate in the working class, com- pared to only 27 per cent of white men, with white women and black men falling Between the two at about S0 per cent, Class Seucture is Contemporary Caplalism 201 Distribton of race sd ls nthe United States ‘iso of des win ace nd ox es cunarge! itn Tosa ode Tat ite. BEB fe f Dineen we OR soon ao 4436428 Teiedwoker™ 441566 9.727824 a ‘The envio collped fom ‘Me fal cle splogy le Table 61 SEE See ey eat, sn © 5 en SST Spr samen st ome = 5 ren together withthe gender rests, we can daw two stg contusions fom these dats, Fist tals se dey ins fighyprivtigd poston in clas terms. About one white man in 2 diners capitalist or an expert manager, that in cas. iocations which re ether part ofthe dominant caso closely ted to the dominant clas. we add to this other managers and hoes, vera third ofall whe en inthe labou fee aren Sohal exploiting lass postions ‘Secondly, the working as contemporary American capt ig cond analy hy momen an mvt As already note, 605 percent of the working elas in the Uni Bie ae women, Hf we add Rack mem to thy the ure approaches twothnds. Ay poital strategy fr the mobilization oF the working sass hor t9 fake this demographic structure into ‘Sonsderation202 CLASS AND ECONOMIC SECTOR® Historically, Marxists have tended to identify the working class with industrial production. As we have seen, this identification has ‘been canonized in certain definitions of the working cas, sack as Poulantas's, which effectively restiets the working cess to indus. tral (Le. productive) labout. ‘The conceptualization of class structure proposed inthis book doesnot link the working cas to industrial production by defn tion. Yer it remains the ease tht in both Sweden and the United States, industrial production, or what I term (Following the usage Adopted: by Joachim Singelmann) the transformative sector, remains the core of the working cass: in Sweden nearly 42 pet cent and in the United States 41 percent of all Workers are emp- loyed inthe transformative sector (see table 6.5) Tf siled work: fers are added to this to constitute an extended working cass, the figure inereases to 45 per cent in Sweden (although it remains ‘esentilly the same in the Us). This, it should be noted, i not ‘dramatically disproportionate representation of industrial produc tion among workers, since 36 per cent of the US labour force and 40 per cent of the Swedish labour force are in the wanslormative remains the case that industeial prodvtion coneti- tutes the core of the working ease The situation is quite different for experts, whether they be ‘managers, supervisors or non-managerial employees. These class locations are highly concentrated in socal and poitial services in both the United States and Sweden. Whereas only 22 percent of the entire labour fores isin this Sector in the United States, and 36 per cent in Sweden, 42 per cent ofall experts inthe United States and 59 percent in Sweden work inthis sector. As we vil see below, the core of this service sector employment is nthe (CLASS STRUCTURE AND THE STATE (On the face of it, itis simple matter to study the statistical relationship between state employment and the class structure, since with few exceptions itis faity unambiguous whether or not 8 given person works for the state. On close examination, how ‘ver, the problem is more complex, since many firs in the private sector may be closely linked fo the state without actually being legally par ofthe state itselt. Tiss certainly the cas, for exam (Class Structure in Contemporary Capitalism 203 le, for military contractors. Should a worker employed in amit Sy weapons factory be treated as located within the tate sector {he private sector? In certain rapes atleast, employes in such firme may have sntrets more ike thos of dest sate employees than oer private-sector workers. For example, ayn the ase of {ullysledged state employees, people in state-dependent rms ave cree interests in the expansion of state budge. “To:map ot propery the relationship between he tie and class steucture, therefore, we would wealy like to distinguish among, Private sector firms on the bass of the nancial inks othe sat Recess to soy, not an easy empl tak to get reliable information on such ten The best we have been ale 10.00 to the private sector respondents onthe survey to sie anes. Tor the fm for which they work ofthe percentage ofthe business that done with the state. These estates are tntkly tbe very accurate, but they ray give Us some Ver rough ‘es of these indies inks to the site “Table 6.6 presents the distribution of state linked employment within various leatons im the ean tuctre. The table indeates Sn interesting pater of cilferences and similarities between Swe den and the United States Mon striking, perhaps, te the ama {ely ferent levels of direct state employment in the two cou tries: 175 per cent (20.6 percent of wagecatnes) in the US in. por cent (466 percent of wage-earets) in he ‘Swosh sample are state employees. This siference occurs throughout the lass strctre, but st is pardcultly notable "mong experts (63 per ent sre state employees in Sweden com pated t 205 percent inthe Unite Stats), On the other hand, {ppeats tht 4 higher proportion of Americas workin private Sori that en mal en the wate, Ai te ae of experts appear tobe the mos striking” 39 per cent of exper Ste suc Re inthe US compared to eny 17 pet contin Ses den “Te result of thes wo patterns tha in both countties,experts the category inthe cas srutae wih the closest inks, ect tinct ih he satay 3 percent nthe Us ad fs han Sb pe cen im Sweden report that they are in private sector firms ‘ht do no busines at all withthe state Tn contrast, in Doth Se don and the United States, workers are the category of wage: sree with the last employment ties Tote state 56 pet cet of UUs morkers and 43 percent of Swedish workers. Also, 8 would be ‘expected, in both countries the cass lations that are MostClase Structure in Contemporary Capitalism 208 stops poe vas pa SSNS HIN Roar IRONED JO Uo soa 208205 Class Structure in Contemporary Capitalism 207 | lated from the state are the petty bourgeoisie and small emp- | Toyers—around 90 per cent of such individuals in both countries | {do no business at al with the state "These data onthe sate and clas structure indicate the impor. tance of the state for various categories of ‘contradictory class | locations As [have argued in an earlier study (using my previous ‘oaceptslizaon of class), much of the expansion of what are | “uslly thought of as midde-clas’ positions can be rect ati- buted'to the growth of state employment» Between 1960 and | 1970 virtually all of the growth of ‘semi-autonomous employee” positions in the United States occured within the state orn those Private sectors (euch as hospitals) which are heavily state: Ffependent. tn the est ofthe economy there was acially an over- sl Gedine of such locations during the period. Managerial loca Tons, while less dependent upon slate expansion than semi- fnutonomous Toeations, nevertheless also increased considerably | ‘de to the expansion of the sate. Politically, the fact that workers and other uncredentaled emp- ) loyees ae Under sepresented in sate employment and in sate- linked employment probably one ofthe reasons tht here tends to be a certain amount of antestatist sentiment in the working ‘Gass, particularly inthe United States. The absence of economic ties tothe state also probably contributes othe ant-statism inthe petty bourgeoisie. OF course, workers may sill receive material ‘engits from state redistributive and social sevice policies, but {hate Iivelthonds are less likely to be divecdy bound up with state Expansion, and this creates @ context for antstate sentiments to ‘evelop, ‘One other observation about the clas-by-state distributions i should be mentioned: although in the society at large there ae { Somewhat more capitalists and petty bourgeois in the United States than in Sweden, when the analysis is restricted tothe private ctor itself there arerif anythin, slightly more self-employed in ‘Sweden thn inthe United States (18.5 pr cent compared to 17.8 per cent) ts asf the market sector ate given level of technical JRevclopment generates a certain level of se-mployment oppor- fUnidee, Although the total sodal space for such self-employment 'S'smaller in Sweden because of the very large amount of state employment, this does not seem fo dampen the impulse for self- } employment in the private sector itself spl seltenpoye = ‘egos ofS aii atledwen beauce oe soy (has srwtare and the state in Sweden and the United States? ces ole ‘Dirihton of eaetnke® empoymen win camer Ieprvas fms cou of Sam, Fees aoe tr eater208 CLASS AND SIZE OF EMPLOVER Marxists have generally characterized the present er sth ee of ‘monopoly captalism’ To be sure, there ino doubt that the growth and power of the multinationals is decsve feature of Sitanced capitalist soles It shaper the polite posibities of ‘orkers and the cconomiemanoewverabliy of sates n pervasive ways ‘Nevertheless, it sa mistake to concide from this that most, workers ae employed direct within such ian capitalist enters Drives. Table 6.7 indicates the Gsibutions of clas and size of Employers Only 1.8 pr cont ofthe working sass and 19.3 per ent of sill workers nthe United States, and about 10 pos cent ‘ofeach of hese in Sweden, work in gigante corporations, corpore: tions employing above ten thourond employecs. It we exclode state employment from the calculations here figure ise to 18 fang 25 per cent of private employment in the Unite States, end 18 and 22 per cea in Sweden, bit are stil fr from 4 majority of workers. Indeed larger proportion ofthe working class in both Counties work for fms wih ss than fifty employees than firms employing over ten thousand people: 22 percent of US workers fnd'17 pr cont of Swedish workers work it such sl firms (or 26 Ber cent and 32 pr eet rspesivy of private sector workers) ‘This may be the Era of monopoly capita bat hie does no imply that monopoly eoeporations avec organize most wage labour ia these ssc “The data table 6.7 point to «second set of interesting find- ings. In both Sweden and the United Stats, mediomsized com: Panies~those ranging in sz fom 500 to 10,000 employees—are the mos proltarnized: 52 percent ofthe postions insach cor oration in oth countries are inthe working Gass. The pian Eoxporations inthis respect look rather more Hike the state, with Under 40 pe cet of thei employees in working-class postion in the Us and around 85 percent in Sweden Tn one respec, fr which I can fer no interpretation, the us and Swedish data are quite diferent. In the United States, the ge corporation has by Tar the hie proportion of supervisors Ins abr tre: 27. percent When Contined withthe nest 1 pet cent managers, this brings the total employment in the ‘managerial apparatus in these corporations to over 40 percent in the US Ths i considerably larger than in ether the sat ut unger 33 pr cent) or the mile size corporation (36 pr cen In ‘Cass Siructre in Contemporary Capitalism 209 Clas itebtion by eof emt Se of enti frm 10000 10.00 Stet ctascweroict <5 Sts ee ee PE BR: UY oR S Mebmumges 24/68 34/52 63/16 StS 10248 cHerne, Ht ie Gus us gage hme net ape Fach pla in Tne 6 te rs amnc emoyets= Ley ety urges 3° manger = 47.10 oaperor BRI eget nar-nangen "8 led wonton = Sr mothers = 1 Eftligdes we ea he sake ech pr ar the pana f people in he ‘sag abo are sper ash clu the iar eg of Ebac oe the peony af peopl 8 gre car who ea Sie ‘ihe prensa se employment ier sigh or Tale G6 esas of ising Sweden the pattern is quite differen: the proportion of super Niuors Is quite low in the largest corporations, even by Swedish Nandands~under 7 per cent of their labour forces—and overall the size of bureaucrat apparatuses does not vary very much fetons organization size (23.7 per cent in medium sized corpora- tions, 24-2 per cen in large corporations and 25.5 per cent inthe sat).210 We have explored a diverse st of findings in this section, Four ener observations ae worth keeping iting. is insite of the various ferences, « numberof important charsteatis of the lass structures ofthese two counteles are relatively imi: the ‘working cls is the largest class the working class and those com {radctry locations that are marginal explotrs conte & sa Stan! majority ofthe idour ores of oth counses; the petty bocrgeai an eptat ca re gute sal and ede exploiters in partulat, and contradictory lations more genet: aly, are paulrly ied tothe state i both counties Second, women ae dsproportionatly proletaranized in both ounties, although slighty more sn Sweden th nthe United States. The result is that women consttte a majority of the work ing class. “hit, in the United States works more heavily supervised than in Sweden: there ate considerably more supervuorr inthe cass SSrotue, paricslany in large corporations Fourth, Sweden has a higher proportion of non-managerial experts than docs the United States, The pomession of credentials nd the contol over organization assets seem fo be less nim Inked in Sweden than in the US Explaining Differences in Class Structures"? In the previous section our fous was on desrbing the smarites and dferencs inthe Swedish and American clas sructute, Ip this section we wll pr ategy for ndersanding a east some ofthe structural eas for these dilerences. In patclar, tre wl explore two principal hypotheses: fist that the ferences 4m dase sroctures are the rest of diferenoes in the mie ot economic activites in the two counes (le ferences inthe Isourdore hts acne econo so) ana second, at they ae the esl of the diferences inthe sie ofthe state the wo societies 7 “The st ofthese hypotheses corresponds tothe view hat vari ions in las strates ae largely tobe explained by technoocl Factors of various sorts If we astume that within given types of conomie acti Iechnologes ae gute smi inthe ‘ited Sates and Sweden, then the principal way in which technological (Class Structure in Contemporary Capitalism 211 factors could explain the diflerences in lass cstibution would be athe diferent mixes of economic activites in the two countries. For example, the manufacturing sector is somewhat larger in Sweden than in the United States and this is precisely the sector in which there isthe highest proportion of workers. This could help to explain why thece are somewhat fewer workers in the United States than in Sweden, "The second hypothesis corresponds to the claim that the state ‘constitutes the essential bass for 2 non-capitlst mode of produc tion If this isa satisfactory formulation, then ‘caval, the relative size of the sate should have a considerable impact on overall clase distributions. At the very minimum it ‘should help to account for difference in size ofthe capitalist class fang the traditional pety bourgeotsic. Tn order to explore these hypotheses we will have to elaborate a statistical strategy for structurally decomposing diflerences in class Structures. This wil be followed by an examination of th extent to which the observed differences in class distributions in the 6¥0 Countries ean be atribated to differences in sectoral distributions, the sizeof the state, ofthe structural ink between authority and credential. | STRATEGY FOR DECOMPOSING DIFFERENCES IN CLASS StRUCTURE The base statistical strategy we wll adopt inthis analysis is based tn the ‘shiftshate’ technique commonly sed in eoonomies and Semography.” The purpose of this technique i to decompose the Uiferences in lass dstibations between the two countries into 3 ‘number of different structural components. In the case of the hhypothess about the effect of sectoral cistibutios, for example, sme would be interested in two primary components, one indicating how much of the foal difference in lass distributions between the two countries ie atibutable to differences in the lass distributions ‘within economic sectors, and & second indicating how much is Ihtributable to the differences in the dstsbution of the labour force across economic sector. (A thitd component, referred (0 as fan interaction term’, which indicates how much ofthe difference betncen counties cannot be wniquey attributed to either of the other components will also be calculated). ‘The technique for decomposing the tual diferences between the two countries into these components involves playing a kind of22 counterfactual game. Tn the case of the hypothesis involving sec- {oral distibutions, we bepin by asking the question, what would the overall US class structure look like ithe United States had (a) {he Us distribution of elasses with economic sectors Dut (8) the Swedish distribution of employment across sectors (or what is technically referred to as the marginal distribution of economic sectors, of more succinctly, the sectoral marginal)? Tis would tell ts how much the Us class structure would change if ts industrial structure changed to match that of Sweden, wil its cass structure ‘Within economiesectorstemained constant Thiscountrfactal est mate provies the bass for calculating the part of the total die ‘ence between the two enuntres attributable to differences in sec- ‘oral distributions. Weil all this component of the total diter- ‘ence the "Swedish economic sector distribution effect on the US class structure (or the Swedish sector effect for shor) ‘Once this counterfactual distribution hasbeen estimated, we ask ‘4 second question: what would the overall US class distribution Took lke ifthe United States had (a) the Swedish distribution of lasses within economic sectors, but (b) the US distibution of employment aero sectors? This tell us how much the US class Structure would. change if its sectoral distribution remained Unchanged, but the cass distributions within sectors matched that ‘of Sweden, This counterfactual enables us to caleulate what will be referred to as the ‘Swedish within-sector sass distribution effect fn the Us cass structure” (or more succinely simply the Swedish class effec) Finally, after calculating these two components ofthe total dit. ference between countries, we can compute what is termed an “interaction” effect Mathematically, the interaction effect is residual term: it isthe diference between the totl difference in the class distributions fr the two countries, andthe sum of the two fcomponents discused above. It reflects that part of the total dit ference that eannot be uniquely assigned either to differences in sectoral distributions or to differences in clase distributions within fectors, It impli that there isa correlation in the way the #0 ‘countries differ in both thir sectoral distributions and the clase Structures within sectors." ‘We have these thes components in terms of the counterfactual effects of Swedish distrbutions on the US cass Structure. We could alternatively, have expressed the decompost: tion as effets of the US distributions onthe Swedinh class struc. ture. When the interaction term ie zero, we would get identical (Glass Structure in Contemporary Capizaliom 213 answers in ether decomposition; where the interaction terms are large, however, the decomposition wil look different from the ‘vantage point of each country.” Tn the tables which follow we will tive both sels of decompositions. As it turns out the interaction ferme are quite small in neatly every case, so the conclusions are largely unaffected by the specie decomposition whichis chosen ‘Using this basic strategy we wil examine three different decom. positions: (1) by eeonomle sector, (2) by state employment, and {3)a more complex decomposition involving the linkages between credentials and authori ‘One final preliminary methodological point. Because of the compesity of the data analysis in these decompositions, both the ‘exposition and interpretation of the results quickly become lnwiely if the numberof categories involved becomes too lage. For this reason its necessary to collapse some of the distinctions ‘made in the fll clas typology. Table 6.8 indicates how this will be ddone andthe class dstoutions for Sweden andthe United States ‘associated with the collapsed class typology. ‘Clas strectare typlogy and astibtion for decomposition of dferences Caceres for seu a ceminmiet le yas Sadar FER SS se Zain eee SE ceo RE [Although in terms ofthe range of economic sector ditrbutins ‘toe ll Countries inthe wor the Uaied Stes and Sweden five rather ser economesrctre, there are nevertheless Selkng ferences tetmeen tem, AB table 63 insted the fied states hase mach ger proportion of total employment in wat could be tomed the wanna captlnensarke series2s (Gistibuive services, business services and personal services) — S377 per cent compe to 182 er et for Sneden— whereas ‘Sweden hat correspondingly larger proportion of {otal employment in soial and political serviees~36.2 per cent ‘compared to 25.7 per cen inthe US. Thus, while the tia "tertiary ctor’ inthe to countries is of roughly similar size inthe to ‘aunties $9.5 per cent in the US compared to 54.2 per cent in Decomposition of iterence in slaw structure by economic star 1% Brine of oma bins 2 ” (Bh USsacor Set ste BS Sedo els Swadsh marl SUS (st rion at ron let Diartatons —‘winaeaor” tn eer 2 Upurea See IS 247s 19 SSrerton’s 198 209 ua ‘renee or sg 9s 106 6 Sitmpoyet 181108 Be 2 9 6 oa Sout San Sr Iron Airset afc OE fey te Bem ime 434 2h sao } nose 36? -0s fewer H7 ye cae 6 Sittings TRS RETA (Class Structure in Contemporary Capialism 215 TABLE 69 (cus) @, 4 gy a Aeron fee a Be ee Bau) ae 28-05 480 ce Table 8 for operonandom of hv cs gai ‘om the acl Swan dstbuons he nracon effet term seo then ‘2 Mo dttmposton wi be the same sap at nh sec ane Sweden—the detailed activities which contibute 0 these totals fare quite different, It might be expected, therefore, that these Uiffetences in sectoral distributions might contebute tothe overall tllferences in cass structures. ‘Table 6.9 indicates that this i notin fact the case. This table should be read a Tollows: column 3 indicates what the clas dis tribution would be in a society withthe United States marginal “ectral distribution but the Swedish distribution of easses within economic sectors column 4 tells us the complementary counter facual in which the Swedish sector marginals combined with the Us clas distributions within sectors. Columns S and 9 ae the gross Uiferences between the to clas distributions (the signs are oppo- Site because in coum § the Swedish figures ae subtracted from the Us figures whereas in column 9 the US figures are subtracted fom the Swedish ones) All ofthe other columns are calculated by Subtracting in different ways columns 3 and 4 from the original ‘istribations or each country. Columns 6 to B give the decomposi-16 tions in terms of imputing Swedish distributions on the US lass ‘Structure; columns 10 012 give decompositions in terms of imp. ing the US distributions on the Swedish elas structure Tf much ofthe differences in class structures in these two coun- tries could be attributed to the differences in economic sector Gistributions, then the sector effect in columns and 11 in luble 6.9 would be large relative to the cass effect in columns 6 and 10. This is not the ease. Except fr the sett Tig ieee “There te, nevertheless, some differences between men and 1 soumise| [4 Eger [7 emis [10 Une women that are worth noting. In general the degree of cass polar= ar pe ation among men i constderabiy greater than among women. | fue-amin |] mospies | wr rzosian” | w sso ‘Male proletarians and expert managers differ by 2.8 points in the ) See ere eer eee United States and 3.6 pnts in Sweden, whereas tie women Seaaciol eset wiser oer on || stnhon | "Spain|" Seton | onan ranter3 wmizeae || massa | a soeacn | a eazien [°° Se is ates clacton wine atgre, Uned Stts BST || MOG |S [HESS Ase i he et of pousion Setbeaue || © ffir |” Sar oe Nonownes mage bose aesoxecan || ac oracusy [ac eg2ey | 270000 © Z | WOT] AEST) ARE GP | AE BS iors | ime |" ges | stent ee w-sian| | w-iseaey | -asscn | scsossean {* | mencan, = 509 NOG" [MaRS [Woo GD | WER Nesey N= Famw | [Semen fe somaca [11 umes “Nene irr 6G Ne Empires || eprion |Spon |" Sects | onan M-118 | | M1029 | M0216 | v0.77 cx) [7° Satan NedarGo| | Wooat Ga | Woo |W So3ste) counterparts ditfer by only 1 point in the United Sats and 1.9 } pois in Sweden, Mow ofthis omer degre of polarization comes, > ae || Eas | ems | heen From the fot that women expert managers ate considerably les rurale | | sonmanager | workers s provapitalist than men expert managers, probably reflecting thei w-oisisn || atom an | soatcan | 4 037 eu Eoncetation in tower evel of management wears | |W rosstan |W ome) |W 1osc655 Th one other respect th abe differs between men and women: - en | tne uncredentiaies manage cll dos not behave’ proper for ‘Steel se Men (0; N= aor ‘Women Gi: N= 80 proletarian cll and certainly does not follow the presribed mono fonie pattern; among American women, on the other hand, tithe least pro-working class ofall the wage-earner categories. cannot ‘offer any explanations for these specific results. Ta any event, itis268 czrainly not the case thatthe overall class structure patterns in table 71 are artifact ofthe sex compositions of clases ‘When we expand the possible sources of spuriousnes to include ge and clas trajectory and calculate the adjusted means in table 74 we again see that there is no evidence thatthe observed relations in table 7-1 are artifacts of the personal atibutes of the incumbents in lass locations. While clases certinly do vary con serably on these variables, they are not the source of variations across classes in class consciousness. (2) Intervening mechanisms in the consciousness formation pro ess, Table 7 examines the patterns in table 1 separately for tunion members and non-union members in each country, ‘Table 7.6 then examines the adjusted means, adding to the con. Avi the mean of proton owns Non-omner (wage kone Pom | seinen [s semeces | 1 Uns ptoers | | somemon | Ropertors | pears | Oran 005 -0m som +080 Stren eee (Cass Structre and Class Consciousness 269 TABLE 74 (anu) ‘Asset nthe men of prodestion ners Nomonne (wage oer { powgotie] [4 expen [7 Smiced. | 10 Unc ed 146 oss suas si90 2 Seat Seapen | 8 Sener | 11 Unces ce Envi sll are adja an vale onthe woring cae coco ol ‘slsied fon multi regomn eatsonconnng te css darn salen ‘Ses secand ca ejector Ses Tae 77 equation (2). trols in table 744 @ number of intervening variables: personal income, unearned income, home ownership, unemployment ‘experience, working-class networks and union membership Unionmembershipislikelytobeamong the most important inter- vening factors in the consciousness formation proces. Tt is cot tainly closely tied to class location, particularly in the United States where the legal system prohibits certain class locations among wage-curners from becoming unionized—management Positions are generally not allowed to be in unions—and one Would expect that unions ought to have atleast some impact on lass attitudes. In these terms, che results in tale 7.5 are quite Interesting. Fst ofall, they clearly indicate the mediating ole of270 Unions: in every cel, union members have emphatically more Dro-working-latsatdiudes than aon-union members in both the United States and Sweden. But equally iteresting& the fact that atleast in Sweden, the same basic patter of polarization and monotonicity is observed among union and non-union members lke. (In the United States there ae so few union members inthe ess not simply via the effects of clas structure on cass formation (as measured by union membership), but also because ofa direct, Impact of class foation on the incumbents of positions, ‘When we calculate the adjusted means controling for ll ofthe mediating variables, there i, 2s would be expected a substantial change from the values of the unadjusted means in table 7-1, in general reducing the differences across cells in the typology.” tapers ‘Cas atts by elas cation fo union and non-union members, ‘nite Sater and Sweden Ae nthe mn of prodaston Now omer [wage labore] ‘tigen [7 Semiaed. ] 10 Une veosar |usisiay | v-ois@ N=152085 | X“os9 Gq) | Naar Sitio |" Seton | ewan | ae veu@ |ususag | ysssrae [°° St Noastsh [x “oaetes) | N “odo Geg yr96 |v sie | v erescey | Nioietty | 8 ORGR, [8 LoSGS} (Class Sructure and Clase Consciousness. 271 TABLE 75 comin) Are inthe meme of prodestion Non-owner [wig houes) seen |? semared | 10 Unces feomges: | 7 surges |" managers veoora9 |uvussin |v +2900 Ni25G9 [x lang) | N2oa36) v-oian |usomes | y-2san Rion [|X a6G) [None v1 6) | vss0608 | u 299099 Niassa [Ntaa0e8 | Nt1290I8) statedeni es "Monn prethosse wep However it stil the case in both the United States and Sweden that wth afew deviations, the basie monotonic velatonship be teen class and consciousness retained ‘Several principal diferences between table 7.6 and able 7.1 are worth noking” fist, when the varfous controls ate included in {able 7.6, the adjusted mean consciousness fo expert manages in ‘Sweden is no longer pro-apitalst. What this means is thatthe ugregate pro-ciptalist stance of Swedish expert managers gem trated by the ink between their class location and heir incomes, ‘inion membership and other intervening process. This is not the fase in the United States, In fact, in the United States expert managers are more provcapialistFelative to the bourgeoisie in {able 7.6 thin they were in table 7.1. My interpretation of these Fesults is that in Sweden the labour movement has been able (02m oie eit seen of management an re line of demarcation between upper evel managers and the lk ot managerial employees. Mont ofthis change inthe adjusted means a the expert-managercallcan be atibutedto the operation of the union variable. Av table 74 indicates, the overall pro-aptalist, Stance of expert managers in Sweden comes from the ety pro- aptalistpontton™ more procapitalit infact tha their American ounterparisof nonsinionzed expert managers in Sweden, This division beqween unionized and non-unionzed exper managers "undoubtedly corresponds to division beoween top management 8nd other managers. What we are observing hee that he union ‘movement is ale to pl lower and mide levels of management into at least a passe coalition with workers. In part because Ave inthe mens of proacion omnes Nen-omers (wage toues) Tr Boureese] {4 Exper [7 Semiaed | 10 Uncea a ee plore || sapemwon | sopesion | Spero | Orga (Class Structure and Class Consciousness 273 TABLE 146 comin) Asti te mean of proucon omnes Nonowes [wig towers ‘ste or pe repress gu ostaning he cay amy ables ae ‘Swen dn bf wi sewer oer ems ern snare of legal obstacles to unionizing managers, and in part because ofthe general weakness of the American labour movement, this has not happened inthe United State, and asa result the rank and-file of management is firmly integrated with the bourgeoisie ‘Mdeologcally. ‘A second point of contrast between table 7.1 and table 7.6, i that the difference i the degre of polarization among wage eat- ners between the United States and Sweden is no long as striking fs in the orignal table. Tn table 7.1 expert managers and pro- letarans differed by 3.3 points in Sweden and by 2.24 in the United States; in table 7.6 the respective differences are 1.S1 and 1140 (not statistically signiicat), Most of this reduction in the274 Uiference between counties in degrees of polriation between clases can be attributed to the inclusion of unionization as an intervening variable, This Supports the interpretation suggested in ‘hypothesis 2 thatthe degree of polarization is mediated by organ ‘zation and political factors ‘A third, and related point, i that the difference between non- unionized workersin Sweden and in the United State is somewhat less than between unionized workers. This suggests that i is ot simply the fact of unionization that ats at a mediating process in consciousness formation, but the strength and socal weight of the Tabour movement. Finally, in one important respect, the patterns in able 7.6 iffer from those in table 7-1: forthe United States, the bourgeoisie itself i now less pro-apitalis then nearly any of the wage-earmer categories that are pro-apitals.. My expectation had Been that the Weologcal stance of capitalists would be more directly ted #0 their class poston than would be the case for wage-eamners, and thus their adjusted means would be less alfected by the incision of intervening variables inthe equation, This i indeed the ese in Sweden, but notin the United States. donot have an explanation {or this result. The intervening variables which most sfected the regression coefficients for the bourgeoisie dummy variable were the income variables, particularly the unearned income” dummy variable. Since this variable iso closely tied to their cass locetion, fetmay be inappropriate to consider ian intervening variable 3 in thelr ease So far we have looked exclusively atthe relationship between class Structure and consciousness. In thi inal analyse will examine tte relationship between the other independent variables used to generate tables 74 and 7.6 and consciousness. The results are Presented in table 77, “There are a number of striking properties of these equations. First, clas and class biography variables (working-class tajee- tory, unemployment experience. and” working-class networks) consistently have bigger effects in Sweden than in the United ‘States, The class dummy variables lone explain 13 percent of the variance in the Swedish equation but only 6 per cent in the US {equition, When the various elas experience variables are added (Cass Seucture and Class Consciousness 275 srr, grey a in non) (rau nce) FS 3B feu 25) Py SER BB! E88} Ci ery276 TABLE 77 oma) See Sy aS omar BS ee sacsean eee coe Se, fa ihe fe ESS Saar enresrey Stereo Class Seucture and Class Consciousness 277 to this equation (variables 14 to 16), the R?increases ony to 8 per cent in the United States, but 17 percent in Sweden. The mage itudes and significance levels of the regression coefficient forthe las dummy variables in equations I and 2 in table 7.7 and the ‘lass experiance variables in equation 3 are also consistently grea: terin Sweden, In particular, except for unemployment experience, the cass experience variabes are at best marginally significant ia the United States equations but are quite significant inthe Swedish equations. (On these coefficients, the ciference between the US land Swedish equations are. generally statistically significant). ‘Gearly, class postion and class biography are more salient deter: ‘inant of consciousness in Sweden than inthe United States ‘Second, in both Sweden and the United States all of the class: consequences variables have significant effets on consciousness ‘Asin th case of the class dummy variables and the class experi= fence variables, the magnitudes of the raw regression coeffictents fare greater in Sweden than in the United Stats for these variables, but the differences are not statistically significant except forthe union membership variable, Immediate clas experience, me sured both by curent location and biography, thus appeas to be ‘more salient determinant of consciousness in Sweden than inthe United. States, whereas the consequences of clas—income, hhome-ownership,eto-—appest to be equally salient in both coun “Third, in neither the United States nor in Sweden does gender, net of the other variables in the equation have any effect Al on class consciousness, as a measured in this study. On the ‘ther hand, the effects of age difler dramatically between the (Wo countries: in equation 3, age is the second best predictor of the consciousness scale inthe US, while in Sweden it has no predictive power whatsoever” There are several possible explanations for {hs Age could constitute a lifecycle variable, and it s posible that besause of the way labour markets and social security are ‘organized in che two countries there are more antagonism 31003 age lines inthe US than in Sweden. More plausibly, age is cohort ‘ariable. The relative historical continuity in Sweden in class po ies from the 1930s to the 1980s could explain the absence of any ng cohort effects on clas consciousness, whereas in the United ates the relative discontinuity represented both by the pre-war and post-war eras, and later, bythe experiences ofthe 19605 could explain the much stronger age effects. Finally, even though we have observed dramatic diferences between Sweden and the United States if We poo the two samples278 int single equation (oot shown) in which couniey appears asa {Ramya tonal sy no means the exe reir of omlgusnes, inthis pooled equation, working-class conscious tes depends more upon whethe or atone isa worker ora union tmombcr tan whether or not one a Swede oe and Amerian. Conclusions “The results in this chapter can be summsrzed in three over- aching conclusions. Fist, the data are systematically consistent ‘in he proposed reconceptslzation of cass in terms of relations Ur expotation, Clas ates sre polarized i he ways predicted by the eaplotation centred concep, and in general they vary loss he dimensions of the ae ypology mate ia the expected ‘Second, the data support the thesis that the underlying structure ‘of elas relations shapes the ovral patter of eas consciousness. Ae ted capa st, Sede and the Unie State ae tnany respects pola cases smong advanced capitalist counties {Eom of Css formation, sate expansion, income inequality, wel- fare state programmes snd soon. Yet in spite ofthese dramatic fifecencn, te baste patter Hnking class structure to cass con- Sciousess is very similar inthe two countries they ae both poar- {ind along the three dimension of exploitation and the values on the eonstousnes ele basicaly vary monotonically a one moves long these dimensions Tirally, wile the overall patterning of consciousness struct auydetrnined By case relations, te evel of working-class con Sctosness a given soity andthe nature of the clas colons tharare built ufonthowe cs relatos are shaped bythe orga tetional and poi practices that characterize the story of lass Siroggle, For al of thr reforms and their ffs a bing 8 Sable ciass compromise in Swedish socety, the Swedish Socal Demosratc Panty and the assogated Smedsh labour movement have adopted statis which enforce certain aspects of working Clas consousness Tater than absorbing i into asd bourges {Meologial hegemony “These strategies have affected each of the three class eonscousness dncased earlier perceptions of theonesofconsequenees and preferences (or understandings of interest), To. mach greater extent than inthe United Sates the (Class Structure and Class Consciousness 279 discourse of pois in Smedon often explicit involves ‘clas The ‘ery name gen in the mass medi to The Conservative paris Sweet "ourgsol partir hence acre asin defining the tran of poles But more inporant then the use of words, the Soa Destoratic Party hasbeen an aren it hich issues" of power and property have teen debated and become part ofthe agenda of politics in Sweden. The sect of these debates hasbeen to enphisie the existence of atenanes {othe existing dsuibutions of power and property. Propssssoch 4 the Meier plana programme cirrenly under consieration {0 gradually erode pavatecapitas ownership ofthe prinpal ‘means of production through the se of union controled invest tment fundslstat hs well The Meidner plan has ben Widely Aebated asa proposal to tansform power relations inthe sodety asa whole, Even hough the moe radical versions ofthe propos have not reesved wide suport the very Yat ofthe ebute ll ‘pens up the tern of alternatives. "The strategies of partes and union ia Sweden have also had the eet of shaping the el and peceed tress of ans fategores of wage-zamers. State-welfar polices pursued bythe Sock Democrat Party hve petra fad arin ania character to them, dsiibuting benefits of diferent sors o most Satogris of wage earners, ths reducing the tendency for wage famers in contradictory exploiting cass locations to see ther intrest as plurea with those n exploited postions: Above al perhaps, the effectiveness of the Swedish Inbour movement i Massively unionangwhite-oflar employees and even substan! seamen of managerial employes, has heightened the degree of Deroeived community of interests among wage earner in ferent fas postions. This does not imply thet the objective basis of onli of interests among wage ener in aifefen eases has Aisappeare, but simpy that their common interest 28 capita. ‘ell exited wage-carershaveasumed proater weight lave to ther fferenia interest with respect to organiaton ander ental explotaton ‘In contast to the Swedish case, political aris and unions in the United States have engaged in pracces whi, witingly oF bnwitingly, have undeemined working-class consciousness: The Democratic Party has systematically placed potted discourse froma language of class. While there are exception of couse, he eneral tendency has been to organize soil confit n noua ‘ays andro emphasize the exremely limited range of alternatives280 for dealing with problems of power and property. State welfare policies have tended to heighten rather than reduce class-based Ghisions among wage earners. And the ineffectiveness of the fabour movement to unionize even a majority of manual industria ‘workers let alone white collar employees, hs meant that the per ceived i ‘camers have tended tobe large tl Sisrevs capital Ax a result a8 the thetore of te 1984 Presiden tial campaign demonstrated, the labour movement is regarded as ‘special interest group in the United States, rather than a repre- sentative ofthe general economic interests of wage-carers. The net result ofthese differences inthe polities statepis and ideologies of parties and unions in the two countries i hat class has considerably greater ideological salience in Sweden than inthe United States: class location and class experiences have a bigger impact on clas consciousness classes are more polarized ideolog tally; and the working class coalition built upon that more polar- {ned ideological train is much bigger. Notes 1, Geo Lb, ior and Clee Contos, Cie Mas 1971, (origina ei), 192, 9.51, Eek Met aunt a 1 0 eae» ihn ‘ojertve telogy Mo ips at there ei ‘SELSSSSS SESS Rial stad coon ot "Eth sone seein wich ne cou etna fos ‘comioe estan pops ofs calles, whencomcouses ued desbe ‘STi lhcmte ant ot pl the fore abet ape the iol iene paie een eens ae ee tas clecine nie feat fo SP hc iowa conetalanon of comaowsnon and ois oeouness crea soy Class Structure and Class Consciousness 281 ia sped here, on Eset, Nara, Fcon cand Sey ian a ea, 2 Romer ely nd she ely of Power Landon ‘toe Ty Emory vss. ct ce are et es SSeS roiecnmamaremione ns 7 Tie ean een eae Sarita octentr rates aeare yea Heyes art ogo in ee et Repco Storey seat eb fot se wa ag dt rom | ‘reedom, since inequalities are a iportant impediment ty D. W. Litton, Gl and Cla Contcioumen in Advanced Cpl, ‘Eitan Sc 1 fo ermpl Masson rue atthe con betsn eri ispsie nit ey ss 122, se een MeL ened tt of ei Te iho te ated a rerondets inceding corey e-mplne ‘wheter tly bad ben secede ps Th woud have eld282 Tin qe ery ee cant neem spite uc a nna he a ne gle a hn Fetes = Wand women woud be edna to haves epimers) viedo mates egrets 55 oli Und Sa etme ahh re roi ha {Suporte rented cove a mon apr waa ‘Spit np ethan yemloee On perce tener, on Succi times aber camer Hovanioaecsepevinrenmenete ‘STH Te ets ac ed on he eoig aaron fo ae SPS, Sacl worting secon» ele 10,11, 18 low exit) anh nto ana ot pte ed ote este mes “SE nugget ein? cit ferme ese ajith mart anny he va tg cot ‘Te sonnet gusto mp nthe poser wed Yo ast he mean LSE ee here oe ay ‘Metcrd eauned the uion membership dummy vate w 210 a te Pi tbe mcrveng varie atte nthe repent ered of Neha frbersmbigty wth he ail, fered ton the tte 1 Inet cannes ce enh een een nf Utena tun he oma uses Only abou $8 pr cn the v3 Shade eter Conclusion “his book Began by arguing that contemporary Marxist lass analysis has been siempting to bridge the up between the Sac poled sical map of clases andthe cone conjunc {url aalis of lass formation and das struggle Tn thi study our tnain preoccupation has been to pprosch this problem by 3s: fematcally rethinking the structural categories themselves ia a vay suitable for incorporation into middle-level theories and Empirical esearch. While we have explored many diverse prob tems, three overarching consisionssaem petcualy important the fist concern the ably ofthe proposed resoneeptuazation ot class sroture; the second involves the salient features of om temporary capitalist lass stature sing this econeeptalation, andthe third is aboat the role of plies in sass analysis ‘The Rxploltation Centred Concept of Cass ‘My earlier work on class structure euffered, I have argued from the tendency to displace the concept of exploitation from the centre of cass analysis. This weakened the sense in which class telations were intrinsically relatons of objectively opposed inter= fests, and posed a series of specific conceptual difficulties, ‘Tse dificulties, ombined with my empirical esearch on class ‘scucture and my encounter with the theoretical work of Jol Roemer, have precipitated the reconceptualization of clase re tions in terms ofthe multcimensional view of exploitation elabo- Fated in chapter three. Classes in capitalist society, now argue, should be seen as rooted inthe complex intersection of tee forms of exploitation: exploitation based on the ownership of capital assets, the control of organization assets and. the posses- sion of skill or redentil assets. While T have some reservations 283282 Tin qe ery ee cant neem spite uc a nna he a ne gle a hn Fetes = Wand women woud be edna to haves epimers) viedo mates egrets 55 oli Und Sa etme ahh re roi ha {Suporte rented cove a mon apr waa ‘Spit np ethan yemloee On perce tener, on Succi times aber camer Hovanioaecsepevinrenmenete ‘STH Te ets ac ed on he eoig aaron fo ae SPS, Sacl worting secon» ele 10,11, 18 low exit) anh nto ana ot pte ed ote este mes “SE nugget ein? cit ferme ese ajith mart anny he va tg cot ‘Te sonnet gusto mp nthe poser wed Yo ast he mean LSE ee here oe ay ‘Metcrd eauned the uion membership dummy vate w 210 a te Pi tbe mcrveng varie atte nthe repent ered of Neha frbersmbigty wth he ail, fered ton the tte 1 Inet cannes ce enh een een nf Utena tun he oma uses Only abou $8 pr cn the v3 Shade eter Conclusion “his book Began by arguing that contemporary Marxist lass analysis has been siempting to bridge the up between the Sac poled sical map of clases andthe cone conjunc {url aalis of lass formation and das struggle Tn thi study our tnain preoccupation has been to pprosch this problem by 3s: fematcally rethinking the structural categories themselves ia a vay suitable for incorporation into middle-level theories and Empirical esearch. While we have explored many diverse prob tems, three overarching consisionssaem petcualy important the fist concern the ably ofthe proposed resoneeptuazation ot class sroture; the second involves the salient features of om temporary capitalist lass stature sing this econeeptalation, andthe third is aboat the role of plies in sass analysis ‘The Rxploltation Centred Concept of Cass ‘My earlier work on class structure euffered, I have argued from the tendency to displace the concept of exploitation from the centre of cass analysis. This weakened the sense in which class telations were intrinsically relatons of objectively opposed inter= fests, and posed a series of specific conceptual difficulties, ‘Tse dificulties, ombined with my empirical esearch on class ‘scucture and my encounter with the theoretical work of Jol Roemer, have precipitated the reconceptualization of clase re tions in terms ofthe multcimensional view of exploitation elabo- Fated in chapter three. Classes in capitalist society, now argue, should be seen as rooted inthe complex intersection of tee forms of exploitation: exploitation based on the ownership of capital assets, the control of organization assets and. the posses- sion of skill or redentil assets. While T have some reservations 283284 about the clas character of the third of these eatgores, this tcooceptton never sve many of heels Kad encountered with my previous approach fo class srutre “Te empirical invesogations we have explored add considerable cre ths econeptutznto, Fn, chaper Eve when Setormaly compared the explotaton-cntred concept £0 660 ‘alsin anual labour aston ofthe working cass and the froduatte-labour deinion the exploitation-ented concept Fired considerably beter, While the results were not without some bigs and are tho subject o alternative interpretations, 0 Senedl where the iterative definitions dsagreed about the class Sfpaicater postions, the data supported thc class placement Seeing to the loge and citeria the expotation-esed con “Recon, when we examine th lina eens se ture and income ineqalty n chapter st he fesuls were almest, iSely predicted bythe explovtation centred concept. Tis was eoinlon prediton since t involved specifying the way Income Sola srs the tee dinenton tthe dass race mt YaThe pater followed these expectations very clovly income ‘Berend essentially monotonially a we moved along al ofthe mensions of exploitation taken singly or togeter. inulin chapter seven, the iavesigation ofthe elationship tetnoen dap stu an clas comssousnes has added urher {Site creiiity ofthe resoneeptoalation. The pattems of vari of consciousness across postions inthe class structure matrix {Sore closely tothe terete! expectations. The results seem ote tclaelyrobuse and at least on the bas ofthe variables we ‘ave Considered: do not appear tobe arias of cerain possible {ources of apurousess. Huthermorc, the same basic pater ‘Sheed ine counts whith are dramatically ferent in thelr neal politcal complexion. ‘Tike together, these dive empiri ul lend conser able support to the. new conceptalization of clas sructure. Benak ests of is sort however, ean never provide define Trajedgement, Alternate explanations ofthe observed patterns retin ve andthe cnet have awn a et Shiy open co both theorteal and methodological question. But ani Snore compelling val conception of las enters the fray of oretee and empieal adjudication, there are competing ‘eatons to adopt some variant ofthe approach proposed her. Conclusion 285 ‘The Class Structure of Contemporary Capitalism Using this new conceptualization of clas structure, we have sys- tematially explored the contours of the American and Swedish class structures. Leaving aside all of the details of that analysis, there are two broad generalizations that we can make. First, in both countries, in spite of the technical and socil changes of contemporary capitalism, the working class remains by Tar the largest class the labour fore. Even i we adopt a narrow specification ofthe working cas, which excludes various holders fof ‘marginal’ exploitation assets, around forty per cont of the Tabour force je in this clase It these marginal categories are aadded—and there are good reasons to do so, particularly inthe fase of the ‘semicredentaled employes’ category—then the ‘working class becomes a clear majorty ia both countries, Second, and equally important, while the working class isthe largest clas, a substantial proportion ofthe labour force occupies exploitative locations within the clase structure. Even if again we txciide all possessor of marginal exploitation assets fron) hi Sesignation, somewhere around one quarter of the abou force it Sweden and the United States are exploiters. Looked at in terms ilies rather than individuals, an even higher proportion of les have atleast one person in an exploiting class within them, probably around forty per eet ofall households. Ths's not to say {har such individuals and famlies re ner exploiters. The central “argument in the reconeoptualization of the "middle class is that such postions are simultaneously exploiters and exploited. This is preciely what defines the complexity of their cass interests and puts them iato what I have called “contradictory locations within Exploitation relations’. My guess is that most of these individuals nd families are stil more captalistically exploited than they are exploiters through other mechanisms. Nevertheless, this Joes not ‘obliterate the fat that they are exploiters and that, asa result they hhave material interests which are fundamentally different from those of workers, aes Structure and Poiies Cass structure is of pervasive importanee in contemporary social life. The control over society's productive asets determines the Tundamentel material interests Of actors and heavily shapes the286 capacities of both individuals and collectives to pursue their inoreste The fat that a ssbxtantia portion ofthe population may te relatively comforiable materially doesnot negate the fact that thelr expats and interests remain bound up with property tions andthe associated proceses of exploitation ‘Nevertheless, n spite Of this importance, the effects of cass sirutre ave mediated by pic. Class relations may define the terrain upon whic interests ate formed and collective capacities, forged, but the outcome of that process of ess formation cannot bettoad of the clas sirvcture tet in the empirical investigations we have discussed, political fac- tors have enfered in two ental ways. Fist of al, nthe statu fomparsons of slaser in Sweden and the United States, the dit- ferences in ther clas strctures seem largely ttbutable to plit- tal processes, The size of the state sll has significant impact on the elas dateButions of the two unre, contibuting to the {eater nmber of aon-manageril experts in Sweden than i he Einted States, and explaining almost entirely the smaleraurber of sll employers and pety bourgeois in Sweden. More subd, politcal dynamics are probably impested inthe much higher Feels of supervision of he Amerian than of te Swedish work force, and of the much closer sociation between expertise and futhonty inthe United States than in Sweden. While the broadest ontous ofthe two counties’ ass structures are shaped By the level of economic development and the funamentalycaptalist, Character of bot societies, the variations in their dass structures tre certainly sgnifeantlyafeced by politcal processes “The second crucial way that polis have entered our empiial investigation i inthe process of consciousness formation, and By Extension, sass formation. Altiough the some base lnkage be- fteon clas strueture and clas consciousness exist in Both oun. {hen the iological consequences of this nk ae cotingent om their political and historical differences. The higher depree of polreation in Sweden andthe moch brosde ideologies basis for Tiporking class colton ste the results ofthis politeal mediation the consciousness formation process Political Implications ‘The preoccupation throughout this book has been on conceptual ‘problems in the analysis of classes, and the theoretical snd empit- El implications of proposed solution to those problems. Except Conclusion 287 in passing, relaively litle attention has been gven tothe impli tons of the analysis for socialist polities Three such impictons ‘Seem paticulany important: the centrality of acl Gemmcrary the poi agenda for socialism; the meses of conceving te process of cess formation in contemporary capitalism ab prob: fem of clas alliances; and the importance of erating the pot ‘medtations which wil make such alliances possible: Let ue brief Took at each of thse in tare es * So ong as Marist believed that socials was the oly possible future to capitalism, to be miitanly ant-capitast was equivalent {0 being prosocials. Destroying capitalism was both necessary 8nd sufcen Tor creating the conditions for socilsn, Once ap talism is viewed as having multe futures, once itis admitted that posteapitaist societies ate posible with new forms of eas sue tates, new mechanisms of exploitation and domination, then this Smple equation of anti-spam with socialism breaks down, It thm becomes ncesary to think through igorously what means to siugale ponitvely Tor socialism rather than simply agaist capitalism "The reconcepwalizaton of cats proposed in his book suggests that the hear ofthe postive struggle for socialism frail Gemo- acy. Sova, a8 has been defined inthis Book, is. socket ‘within which contol over capital ass and organizational assets sre no longer sinticant sources of exploitation Fr thik to occur, Private ownership of capital ssets and herarchical-uthortan onal ver organization assis must be eliminated, Taken together, thismplies that socialism means radial democratic cor {wo over the physical and orgeniztinalresoutes used in produc. tion, “This of course, nots novel conclsion, The increasing avare- ness of the importance of democracy has been one of the hal Imatk of recent poles debates onthe left. Indeed, it would ot te going too far to say that, at leat inthe American context the problem of democracy has tended to. place the problem of Socialism from the entre stage offi politica courte Instead of displacing socialism by democracy asthe cote politcal agenda ‘ofthe lf the arguments inthis book suggest tha the struggle for Socialism and the struggle for democracy are two sides of a single process Without «reistabrtion of organization assets through a Semocratiation of the process of control and co-ordination of production, orgsniatiom asst exploitation would continue and {pon tht exploitation « new srustre of as relations would be288 built. Democracy is not simply question of how the pos TRatiotos ofthe state arc organized; ano bears direc on ow {las relations thomscives are constituted Tithe importance of radial democracy as an objective of srug- se one ofthe basi pola impiations ofthis ta, the prob {Emauc harater of the proces of clas formation needed to accomplish that goals anther It were true that the cass struc fare of contemporary capitalism Was basically polarized between & thasive working clas andthe bourgeos, then the problem of {os formation sould be much ple thant Basil the as ‘Soul be one of forging cole organizations of individuals al {twhom share the same fundamental lass interests, Bu as have STguds the clase structures of actually existing capi are not ‘Sibpe polarized sractures A substantial propordon of the pops ini, atleast inthe advanced capitalist countries, oupy coo teedciory locations, wituin explosion relations, Toetons Which they are simultaneously exploited and exploiters Tes dit. Teutto imagine a sonaro im which socialism would become area Posty a these wt toe the cxpeation a i Ein segment of the peopl in such contadctoryloatons. Yet at Tous in terms of thir material interest the incumbents ofthese ontauctory locations ar either ditety threatened by socialism, frat lest have relatively ambiguous material interest in a soi ‘transformation ia siasnaies “This poses dep lemma for socialists: socialism s achievable only mith te co-operation of segments ofthe population for whom Sbcllsm docs aot pose clear material advantages.* How ean thi lemme ea wiki Thee etal ide Spproaches that are implicit nsoclist arguments, The fist is 0 ‘ebay deny the problem. Socialism, Goel ora ‘iminate the massive waste in expla (excessive ita Spending, sdverising, conspicuous corporate consumption, ete) ‘Ear the vast majrty of the population will be beter off in & SScialist society. Im terms of the sna inthis Book, eal prod: ity ofusfulconsumption wouldenpandsomich ha many people contradictory locations within expatation relations would act Teller off and only a yey few would be worse offi ist and organization exploitation were climinated In effec, {gangument implies that most ofthe labousime erated by he eduction of capraint waste could be Tedtected towards wef Tatra consumption, ths signcanly raising the average sta {rd of ving. This would mean that even i consumption Fevels Conclusion 289 were substantially equalized in a socialist society this might not Jmply'a reduction of the standards of living of most people in contradictory locations "This kind of argument often meets with a fair amount of cepic- fam. A ratically democratic socialism will have to devote & great deal of socaly necessary labour-time’ to democratic participation in onder for the democratic inttations of prodition to funtion tffectivey, Much of the reduction of waste from capitalise, there- fore, wll be needed simply to make time available for democratic participation, rather than to produce for personal consumption Furthermore, i would be reasonable to expect in setlist society that quite diferent kinds of efficiency criteria would be instituted in production. For example, under democratic conditions workers ‘may opt fora slower pace of work which could reduce (otal social Productivity. 1s therefore very difficult to know in advance what ‘ul happen to overall socal productivity ina socialist society, and ‘hus what will be the fate ofthe material interests of people in contradictory locations in capitalism, "The second solution to the general dilemma face by socialists in ‘ying 0 gain the collaboration of people in contradictory loca- tions is to emphasiae a range of interests other than individual Consumption. Arguments for sacilism in terms of the quai of life, the expansion of rea freedom, the eduction of violence and so on, provide a bass for building clas coalitions for socialist ‘objectives ” Such goals donot eliminate the contradictory material Interests which members of such a coalition would bring (0 tocialist struggle, but they have the potential of neutralizing theit fects “The process of clas formation through which 2 viable, cohe- sive socials coalition is forged isnot simply a question of socialists figuting out what Kinds of goals wil have the greatest appeal 0 contradictory locations within exploitation relations, As OUT “npirical investigations have emphasized, the entice process of| ‘ass formation is heavily mediated by politics and ideology. This, then, isthe third general political implication of the analysis in ‘order to create the conditions under which a democrat socialist ‘ass coalition is posible, these mediations themes have 0 be transformed, “This is not a new idea in Marxism, Lenin's classi eal fr “smash: ing’ the capitalist state was based on the view that this state apparatus was organized in such a way that it prevented the Work. ing class from becoming the ‘ruling class. Only by destroying this290 apparatus and replacing it by a qualitatively distinet kind of| Shparatus would socialism be possible ‘Even if we reject Lenin's rather monolithic view of the struc ture ofthe capitalist sate ad see greater possibilities for political action within its apparatuses, the basi intuition behind Lenin ‘esis remains sound. The political and ideological context within Ivhich strugales for socialism occur significantly shapes the poten- tial for diferent kinds of class formations. This means that iis Important for socialists to identify those features of capitalist polit- {cal and ideological insttations which play particularly important ‘les in defining this terrain of struggle: and thus most pervasively curtail or enhance the long-term possiblities of creating radically ‘Semocratic socialist coalitions. To take just a few examples: the ‘ferences in labour law between the United States and Sweden txplain, in part why the levels of unionization are so dramatically Siferen inthe two countries, and this in turm has significant impli ‘tions forthe clase coalitions between workers and contradictory locations, The differences in electoral insitutions between coun tres can make it extremely difficult fr radical partes to gain any political presence (asin the United Sigates) or relatively cases (asin West Germany). The extent to which soel welfare programmes ae primarily organized around meanstests in which recipients fre sharply distinguished from non-recpient, or a universal programs, in which everyone receives benefits (but different people pay diferent amount af taxes) may havea large impact on the level of support for such programmes in particular, and the broader politcal coalitions that are formed around such suppor. In each of these eases, poiial reforms have the potential t0 enlarge the social space for socialist trugales. This isthe core of| ‘what was called "nonreformint elorms' in the 1970s: reforms Ivithin the existing society which transform the conditions of sub- sequent struggle and potentially expand the very horizon of histor ical possibilites ‘Class structures may determine the limits of possible class for ‘mations and cass struggles, but within those imi a wide range of| Giferent Kinds of siagele can occur. Such struggles may largely reproduce the existing class structure, o may st the stage for new {orms of post capitalist exploitation, or open the possiblities for Socialism, Whether ar not the left wl beable to forge the cond tions in ‘capitalism which make democratic socialism posible ‘depends, in par, on its ability to identify the kinds of institutional feforms of existing society that enhance the potential for class {ormations engaged in strugale for sucha future Conclusion 291 Notes sagan conte Jha Coban Soa Roget Ot Baroy pon EA en Bow Bs Cd ed Tavs Weton, Bente he cl tea al yen re th i RS Sem Ser We cn ay rg “ ‘these red hae rah aa goals abo Tantnnto Sos Poicr Sasa, va. ao 2 198T Cvs Oe and Ung te ft tnge heen oes he cee arb ese {fades people experince n Jing whether oro! ospor sven sea, ‘See the "Twp Lope of Coecive Acne’ Poa! Pow and oval They, NL Sey es Zt, em 9. my las, Cis andthe Ste, chapters our sade nents seeAppendix I Practical Strategies for Transforming Concepts “The process of concept formation ie always simultancously a pro- ess of concept transformation, There are always conceptual av ‘materials which o info the praduction of any new concept. The task of this appendix isto lay ut some ofthe ways in which such a transformation of existing concepts occurs. To do this we will ist Took briefly at the circamstances swithin which an impulse for Inching the attempt at producing new concepts slike) to our This wil be followed by #cscustion of different forms oF concept transformation, different practical ways in which conceptual raw Inaterals are worked on to produce new concepts. This iscusion E not meant to be a comprehensive methodological analysis of lltemative approaches to producing and transforming cone; bur rather an exposition of «variety of practical stratepies tht f have found useful in diferent contexts Oceasions for Concept Formation Many, perhaps mos, theoreeal innovations hinge on the ntoe {cto of new concepts o the reconsuuction of old ones. Three ircumtances typically imate such changes: encounters with piel problems, dicoveres of coneptual inconsistencies and ‘eing wt the remienion of calc? conceptual Wansfrme- “The mest common motivation for prodcing new concepts is undoubiedlyctatsfacton withthe sity ofexstng concept © deal with espinal problems, The accumulation ofempieal cases Seneca hein oneal map ot Suggess hat the map int propety raw, that ne concepts are teed, Two suc ekamples have Deen dpcssed ia this book: the {meee of locations wii he social elatons of production of apa sotenes which do not easly tint ether the expat its or the working class andthe emergence of post-aptal Soaedes which donot ely ft ito. the capitallm-soialsm “iehotomy, The ft ofthese provided the sls fo the into- {futon ofthe eoacept contetictory locas within lass rel- 292 Appendis 1 293 ons, the second for the concep sate mode of production’. In both cases the pre-existing concepts within Marat theory sensed ‘nable to dea effectively with these stvctaral changes ‘Nom, may turn out that these apparent counterexamples to the exsing concep frames can on cone ispecton be sccommodated: What needed may simpy a afistion of rising definitions or a drawing out of het more subtle plea tion ater than a substantive transformation of those definitions This possi isetely posed inthe debate ove he clas charac: ter of "actualy exiting socism’ (he USSR, Eastern Europe, Guna, Cabs, ec). Instead of treating these cases as inconsistent ‘it the eapiaism-socialism conceptal dichotomy, they ca, fr {romple, be regard ascii societies whoo const insite onal forme have been influenced by the continuing existence of powerfol capitalist societies. This implies specie causal argu tent sbost the effect of capitalism on soi inatiations, Dut leaves ntact a particular definition of socialism as pubic owner ‘hip of the mean of production 'A Second stimulus forthe transformation of concepts comes {com the dacovry of theoretical inconsistencies within the ray of existing concepts. Theories are not just olecions of eonoepts ‘hich ar inked through various Kins of propoitins, The con- ‘pls themselves are imerdependeat in various ways, In partial, Some concepts canbe vied an sb-speces of mre general con epts. It may turn out, then, thatthe extera which Gane te feneal concept may be incompatible withthe specication of Particular subsategoy within ‘A good example ofthis problem is pote inthe recent debates over the concep ofthe “Asie mode of production particulary ' elaborated i the controversial book by Barty Hindcts and Paul ‘Hirt, Precepts Mode of Producion.* Ther esentialarBu- ment is thatthe concept of an Asie mode of production Hepiimate because teannot be propery subsimed under the fencral conpt of mode of production, The geet oncept Species that fo count at mae of production there mst bea ‘pectic form of corespondence between the relations and ores at production. Such corespondence, they argue, can be estab fled forthe capita and the feudal modes of production, ut ot forthe hypothesized Asiatic mode of production, No concept ofa mode of production can be derived from the talent ‘ouple, no ariatd combination of eats foros of proton et294 ‘be deduced, and no gstematle conition of existence forthe moe of| ppropriation ofthe surplsprodac, oxen, canbe cnsed Seen Sie ateere emma mee tg Mesa Peretti ant ed eee ah es me ees a Sab eee emg sae ore ee ein need pr of cnr comer mg Ch as ee Sica Seta tes ete So maces rey Pio sed sya lied Snes fie ely Sey al eet SEES om crate oon Son bo meg ame se ete WE oes ae ns oa len eta a ee Siar may ce ale STORE eter athe cis in Forms of Concept Formation ‘Once the need is recognized, a variety of strategies can be en ployed for transforming concepts. In practice of couse, the process ‘may be quite haphazard and unsystematc, and without mach sll consciousness. Four general strategies, however, seem to underlie ‘many successful productions of new concepts: drawing new lines of Appendix 1 295 demarcation; respecting existing lines of demarcation, re Searegatng categories ander ore general eer; and decoding {he conceptual dimensionality ofa descriptive taxonomy [New Demareations. One of the basic ways in which an existing foncept may prove unsatisfactory is that It incorrectly subsumes ‘ite heterogeneous cates under a single heading. The task of Shncept formation, then, isto specify # new line of demarcation ‘within the conceptual fed "A. good example of this is the problem of post-capialst societies, Traditionally most Marnists have argued that socialism, 4&8 the transitional form of proition to. communism (or the “Tower stage" of communism), was the ony possible form of post- capitals society. The imple capitalism-socialism dichotomy was ‘Sten as an adequate conceptual map of real posites. Under Such a conceptual framework, societies sich asthe Soviet Union ‘were necesaly treated as either a variety of socialism or variety ‘of capitalism (ne. state-capitalist society). As T argued in chapter three, an alternative isto intradice anew line of demarcation: the “istintion between the socialist mode of production, the capitalist, ‘mode of production and what might be ealled the state mode of production’. What was previously subsumed under ether capital Em or socialism is then treated a «distinct mode of production i its own right A Similar operation occurs i the transformation of the concept ofthe working class as wage labourers into a varity of alternative ‘ncepts, Poulantaws’s concept of the new pelly bourgeoisie, for texample, represents anew line of demarcation within the category ‘age labour" He argues that mental labourers and unproductive labourers, although they are wage-earners, are in an entirely di- ferent elas from manual, productive wage-earers. What was pre- ‘iously# single conceptual category i tus spit into 10. Respecifications of Lines of Demarcation. It may happen thatthe problem with a concept i not that it needs to be ‘numberof distinc concept, but that the eiteria which boundaries need modiffeation, There may be redundant rier, insufficient criteria or simply incorrect eniteria. “This kindof dispute over concepts has played an important role in the long-standing debate over the proper definition of capital: fim within discussions ofthe transition from feudalism to capital-295 ism. Theres no dispute among theorists over the descriptions of the end poins ofthe proces mature instal captain i seen 8s system of production wth wageabour and private ownetship Ot the means of production; clssieal feudalism f seen ah apc {ural production within which rorplss Is appropriated through xtta-economic’coetcon. ‘The dhagresment centres on the appropriate eters fr specifying the onset of capital, and ts {or detning the theoretealy pertinent minimum conditions for capitals tbe capitalism: suffeent to have economic at. iy oriented towards profit maximization and atcumlation on ‘uarke for an economic stem to be caps, orn aso neces Sary that there be a fee market in labour power. that explo tation operates through the ring of fee mage labour? Ina sar manner, my debate with Pouantas over the dfii- tion ofthe working ess canbe interpreted asa pute over the appropriate lines of demarcation of he concept.’ Poulnteas cot Sered all unproductive wage Tabouters to’ be non-workes fepoedtht he productive “anproitve oar con wa an inappropriate eiterio for specying the boundary of the workin Glue Polat sso concrete metal ual abou te finction tobe a citron for the boundary ofthe working Gass Here my csapreement with him war gh diferent This istne. tion does derive from a structural feature of production relations vthich i appropriate for defining the working css Tngued, but the formulation aterm of mental labour wa iaorret tis not Oy virtue of Being & manual Inbourer pore that wage abouret butside of the working cass, but by virtue of having pervasive fontrol over one's om labour proses, or what Tealed "seme stony Whe rue that ch aunomy hres af ‘uch mental abour, Poslantst mi pecied the presse natore ‘tthe cls erterion, My tanslormation of Poulntae's concep of the working class in tis instance was to rexpeiy this line of ‘demarcation fn terms of real elations of autonony ad eon, Reaagregating Categories third way of ransfoming concepts to Sabeume thom nt new way unde 9 more encompassg coe pt concept which denies move fundamental Dounsry tevin for the concepts agprgated within Wrest the fist Strategy dscased above inveved splint single concept onthe rounds oft ema heterogeneity n thease dnt cones Se eagrepated onthe grounds oftheir eal homeonene). ‘An-example of sich cnceptal aggregation i the elahoraion Appendix 1 297 tod retnement ofthe sone ofthe apa wae in ent Marust theory A range of sonerete forms of the state can found in cataist societies: iberal bourgeois democracies, fascist, ‘Gettorhip itary uta, social democraie welfare states and Soom The cena tess of defender the concep ofthe eaptl- Issa” such as Poulantzas and Therbom is that all of these ‘Sverse forms ofthe stat can be ubsomed under the more general Concept ofthe capitalist state" This concept does no, of couse, Splythat cere ave no theoretically significant ferences among thee diverse sub-types of the capitalist state, but simply that there te certain deep srutural properties which they all holdin com ‘mon and whieh jos Sdenlyng them all with ingle encompes- Sing concept. This aggregation process has the effect of transform lebih corpo ach ofthe mee forme of he ate ing iggregated, fr they are no longer defined solely in tems of formal politcal instutional characterises, but in terms of thet class fara aswell Of cour goes thot sng that hs aim tay be incorrect Bach ofthese types ofthe slate may be simply ‘Sitesi capa soc’ rather than subtypes of the “eps tate, Thej'may have no distinctive or common class character ‘The debate ove the concept of the capitalist tte i thus adebate ver the leptimacy ofthis particular conceptual aggregation as & process of concept formation.” Decoding the Dimensonaliy of Taxonomies. The final general Statey‘of concept formation fs perhaps the most complen. I Ives anforming the tixonone wd ees si theories Int conceptual typologes. A taxonomy is a Isto wis which are ciferentiated on the basis of immediately ent empl citer; (pology, on the other band, is & ‘htorcly contracted se of entopoics flerentated on the Sasso teres speed dimensions” Sometines happen thats theory may develop a iuitive YPoIy Telbgnicing the underlying dimensionality of the categories. In Sich eases, conoopt formation consists of making explic the implict, underteorzed loge of the typology already in we ‘etme give an example of thi strategy from work onthe theory othe state, One ofthe problems facing anyone doing research on ‘heat show to ans state polis: One approach s simply 0 {ake as given the spending calories defined bureaucratcally ‘itn sate budgets This would constitutes desriptive taxonomy298 of sate spending, with tems broken down by state agency and as ine we cha ft of state budgetary items i obviously unsatsat ftom a theoreti point of view. The task of concep formation then, so wansform ti is nto a coneptualysrctured ype 6." One sch typology reongaizes site pies along two (1) Whether the intervention is primarily atthe evel of eieula- lion or production: (2) Whether the intervention is commodified or decommodifed. Gireuiation interventions involve the allocation and redistribution of resoures that have already been produced Most welfare spend ing would fall under this category. Production level iterenos, gate ote hand vote the ate dey iy decon fo po: ace certain Wewalues, rather than simply allocating ext Tesourees. Miltary spending is clasical example of Production level intervention. The distinction between cone ‘modified and decommodiied interventions soncerns the extent hich the intervention works tough the market, reinforcing the ommodified character of soci production, ot, onthe omar, operates outside of the market, pote even ating against Ne logic of market relations. A national health service tn hich the slate directly organizes the provision of heath cae isa relatvely Since & Soar, vol. 37:3 a $1973 and 1974 onto, Ted "Objective Ieee an the Socogy of Power's Sociology, ‘ol 15, No.2, May, 1981 ae o Bestau, Dabil, Deine personel ef sacar de ls, Pats, Pesce “Universitaire Se Frane, 1977 ‘Bowles, Sem, David Gordon and Thomas Weiskop, Beyond the Wase- Tan, New York, Anchor, 1988, Brea Hay Labora Monopoly Capa, New York, Moxy ‘Brenner Jahan and Maria Rams, ‘Rethinking Women's Oppression, "New Left Revie, Nomber 189, March- Apel, 1984 Brenner, Rober, "Te Origins of Captalit Development Ceigue of 332 Bibliography 333 Neo-Smihign Marssn’, New Lef Review, Number 104 Juy-Avgsst, 15th. 25-93. Brownlag H. and J Singelmann, The Emergence of Service Socey, ‘Nason! Technic Information Service, Springfield, Mo. 1975 arawoy., Michal, Manufacuring Consent, Chae, University of ‘hep Pres, 1979. Durawoy. Mishel The Polis of Production, London, New Left Books, 1988, cameron, J, "The Expansion ofthe Public Ezonomy’, American Poca “cence Review, 1978, v.72 carat, ©, The BeonomicIdenifeation of Social Classes, London, Roatieye & Kepan Paul, 1977. cohen, OAs Karl Mort's Theoy of History: A deme, Princeton, Princeton Univeray Press. 1978 cohen, G's Labor Thzory of Valu and the Concept of Exploits Thon’, Piiophy and Public Afar, 8, 1979 can’ GA Reply to Elster, Theory and Sater, vl. 11:3, July 198. Gen Jess Cla and Ci Sosy, Ament, Universiy of Mas Tachi Frew, 18 cohen foshua and Joel Rogers, On Democacy, New York, Peegun "Books, 1983, Cal, Ri The Credentl Socey: A Historical Sociology of Eduction Sand Sisson, Orlando, Forde, Academic Press 1979 ‘consol Win, On atresia Pais, Pls & Soc, 2:4, 197, “357 ‘cRempon, Rosemary and John Gubbay, Beonomy and Clas: Sacre ‘Rew York, st Maris Pres, 1978, Dando Raph. Cless and Clas Conf n Indust Soca, Plo ‘Ato, Stanford Unversity Press, 1989 Doty Marc, Studies nthe Development of Cpt, Cambridge and ‘New York, Cambridge Univer Press 1963. Edvarde Risard, Conte Terrain, New York, Baie Books, 1979 Ehrenreich, Baars an lta Brerich, “The Brofesona Manager ‘Gse, Radel America, 112,197) ger, on, Mars, Puctonism ad Game Theor’, Theory end Soe Mey vol1ytS3oad2, ly 1982, ser, Jon, Sour Grape, Cbg, Cambridge Universy Pres, 1983. Elser Jon, Mating Sse of Mars; Cambae, Cambridge Universi) Press, 1985 sping Anderson, Oda, Roget Fedland and Erik Olin Wright ‘Modes "Gr Caw Stage and the Cepitist State, Kepllina, Number inte sping Anderson, Gs, Polis Agsint Marks, Princeton, aceon ‘rverty Pres, 1988 rec, France, The Internal Suuctr ofthe Proletariat « Marist "Knut Soci Revolution, No, 26, 1978334 Geass, Raymond, The Idea of Cre! Dheory: Habermas and the Frank- “rt Schoo, Cambridge, Cambridge University Pres 18 Giddens, Anibony, The Clas Sruture of the Advanced Societies, New "York, Harper and Row, 1973, Giddens, Anthony, ‘Postrpto The Cass Struc of the Advanced ‘Soci, second edo, New York, Harper and Row, 199. Giadens, Anthony. A. Contemporary Cotgue of Histol Material, ‘Berkeley, Univer of California Pres, 1982, Gough, la, The Poiia! Economy ofthe Welfare Sue, London, 1979. Goulden, Avia, The Fue of Pletal andthe Rist 0d New Cas, ‘New York, The Seabury res, 1072 Groves, Ror Mand Robert. Kab, Surveys by Telephone, Osando, ‘Academe Pros 1979 Groves, Robert, M,‘An Empirical Comparion of wo telephone sample hig Journal of Marketing Research, 15,1979, pp. 622-3 ‘Hartman, Held, "The Fal a5 the Locs of Gender, Cas, and Pies! Stugue the Example of Housework Snr, v0, no. 198). Hindess, Bary and Pal Q. Hist, Pre-Capuaior Modes of Production, Tondon, Rowe & Kepan Paul 1978 alowayJobn and Sl Pict (ed) Sta and Capi, Austin, Univer- Sty of Texas Pres, 1978 Holmwood, J. ML and A Siewart, “The role of contadtion ia Modern “Theories of Soil Srtieation, Sociology, No.7. Ma, 1983 Insite for Labor Eduction and Research, Wha’ Wrong with he U.S. ‘Eeonom?, Bost, South End Pres, 1982, Jackman, Mary and Robert Jackman, Clas Awareness in the United ‘Statr, Berkeley, Univer of Calfornin Prose, 1983 itn, E‘Colaponents os ference betneen two ate Jounal of the American Steal Azociaton, vl 80,1985, 1168-1174 Konrad, Goorge and Ivan Sle, Iulecualy on the Road 10 Clans Landeckr, W'S, Class Boundaries, American Soclogcal Review, ‘ol 25, 1960, pp. 868-877. ‘Ls Gear Power and Pree: teary foal aration New Levine Anarew and Erk On Weight "Rationality nd Class Strogle ‘New Left Review, Number 123, 1980, pp. 47-68, Levine, Andrew, Arguing for Soca, Londo Paul, 1984 Lipset Seymour, M.,Pltzal Man, Garden Cis be, Routledge de Kepen Lipset, Seymour, M, “Soil Staifcation: Soil Cla’, Irrational “Exeylopedia of the Soil Sceneny D. eS (ed), New York, rowel Colier and McMilin, 1968, voi 5: 296-516, Livingstone, Davi, las and Clas Consiuanen in Advanced Cap- “alm, anplished mana, Onto Instat for Sader m Ea ‘ton Foronto, 1985 Bibliography 335 Loren, Charles, Clase inthe Unied Star, Davis, Calc, Cardinal Pobisers, 1977 Lakes, Georg, Mizory ond Class Consciousness, Cambridge, Mas. ‘MLL. Pres, 1971 orginal edton, 1922 Lakes, Steven, Power radia sie London, MeMilan, 1974 ‘Mann, Michael, Concouress and Aton cong tke Wester Working ‘Cass, London, MeMiian, 1973 Marsal, Gordon, "Some remarks onthe stody of working clas con scones Pair Soi, 1253, 1983, ‘Mary, Ker, Capa, vl 1, New York, Ineratona Publishers 1967. ‘Necoats Martin ‘Broetarst and Mig Clr in Mars’, Sue onthe efi No.7 1967. Noble, Davi, 'Sovil Choice in Machine Desge, Poles and Soci, 324, 1978 [Novy Robert, Anorchy, Site and Uropia, New York, Basic Books, "974, ‘ote Clas, ‘Strctral Problems ofthe Capita Stat: clas ulead the ‘pote sytem’, nom Beye ed), Geman Poca Sud, Rost Sage, 1978 Ofte, Clas and Helmut Wiesenthal, Two Lage of Collective Acton’ in Mawrse Zeit (ed), Pole! Power and Social Theory, ol 1, Crean Which, Connection, AT Pes, 1980, Cllmann, Bere, "Toward Cast Coniciouses in the Working Css Poles & Soci, Fal, 1972 ‘Oreown,Stanau, Clas Sracure nthe Seal Consciousness, London, ‘Routledge & Kepan Pl, 1963. Pater Grand Ar Mile, Industral and Occupational Trends bn En ‘ployment, Universty of PennnyWania, Wharton Schl, Indust ‘Rescrch Department, Philadelphia 154. Parkin, ranks Mors ond Clar Theory a Bourgeois eigue, New "York, Cohumbia Univers Pres, 1979 Police & Socey, Specs Ine on John Roomer's Theory of Class and "Expotiation, sol. 113, 1982 Poulantns, Nios, Poli! Power end Socal Claret, London, New Left ‘Books 1973 Poslantans, Nios, Clases in Contemporary Ceptlim, London, New Left Books, 1995, Prieworsi, Adem, From Proletariat into Cans the proses of cass for ‘maton from Kael Kasse The Clet Straggle to Recent Debate Pots &Sacey, 7:4, 197, Praeworsy Adam, The Material Bass of Content Economic snd Pol ‘i ina Hegemonie Stem in Maurice Zeitin (ed), Pola! Power 4nd Soca! Theory, vo JAK Pres, 1979 rneworal Adam, Mistral itrerts, class ompromise and the Trans tion to Socials. Potcer & Society 10°2 1980 Preewortk Adom, “Socal Demoerscy st an Hiiorial Phenomenon, [Now Lef Review, Nomber 122, 1980336 Ravi, John, A Theory of Juice, Cambridge, Mas, Harvard Univesity Press, 1971 Roemer, John, Genera Theory of Exploitation and Clas, Cambridge, Mass, Haracd University Pres 1982 ‘Roem, John, “Should Maris be Inerested in Exploitation", Univer- iy of Calflora, Davis, Deparment of Econonic Working Pape Nimes 221, 1983, ‘Shale, Michael, "The Soil Democratic Mode! and Beyond: wo genera tions of comparative esearch onthe welletesste’- Comparative Soi Research vl 8, 1988 Shanin, Theodor, The Late Marx and he Russian Road, New York, sity Reve Pret, Tse ingeimana, Joachim, From Agrculure io Series, Bevtly His, Sage Miskcnices 197F e i Sian, Carmen, Workers Convo and Socklit Democracy, London, New ett Booka/Vero, 1982 Skoepol, Theda, Stes and Socal Revolutions, New York, Cambridge ‘Unwerty Press, 1979 ‘Skoepot, Theda *Poltcal Response to Capalist Cr: ¢o- Marxist “Theories of the State and the Case of he New Deal’, Polis & Soe, Fate Leads and Promising ‘Str in Corrnt Theres and Resear in Peter Enann Thode Show poland Dietch Rucschemeyer (es), Bringing the State Batk In, Now York, Cambridge Univerty Press, 1988. Steedman, la, Mere fo Sa, London, New Left BookyVerso, 1977 Stedman, Ian eal, The Vale Controversy, London: New Lett Books, "oat Stephens, Jn, The Transton to Socialism, London, MeMilan, 1979. Stewart, AK: Prandy and RM. Blackbsn, Soll Sidenon ond ‘eeupation, London, Macln,1980 ‘smeery, Pal The Debate onthe Teanslion: a rig’ n Rodney Hi ton (ed), The Tansaon fom Feudal to Caplin, London, New ek Books, 1978 Szeon, rm and Rober Manchin, ‘Sci Paley and State Socialis’ in (G Expng-Anderon, {- Ralawater and M, Rei els), Stapation ad Renewal in Soc Policy, White Pits, N.Y Sharpe Pubsbers, 1085 Seeleay, Wan and Wilt Mari, New Clas Theones and Beyond, “published manuseipt, Department of Soslogy, University of Wis comin, Madison, 1988, ‘hero, Goran, What Does the Ruling Class Do When i Rules? Lon: don, New Lett Books, 1978 ‘Therbora, Goran, The Power of Ideology and the Lielogy of Power, ‘London, New Left Books, 1982 ‘Therborn, Goran "The Prospects of Labor and the Transformation of| ‘Advanced Capialsn’ Now Left Review, Number 145, 1984 Bibliography 337 ‘Tompson, EP The Making of the Bnlh Working Clas, Haronds orth, Peni, 1968 Walter, Pat Been Capital and Labo, Boston, South End Pres 1979 Wats, Pe Modem Worl Sn, New a, Adee Wiles, Per The Dinrbuon of Income Eas and Wet, Ante, Nort Hand 1975 Wolpe, Harold), The Arclation of Modes of Production, London, ‘oaiege fe Keri Pal 980 ‘Wight, tk Ot, Claw Boundaries in Advaned Capital Sot’, bw Left Ree, Noser 98, 1976 Wight, Erk Oi, Clas Sractre and Income Inequality, PRD dsr tin, Department of Soclogy, Besley, Unersiy of Calor, 9% Weigh sk Ob, Cas Ci and he Ste, London, New Lat Boks ‘Wight, Eek Ola, “The Value Controversy and Sail Reser New Teh teen, Number 16,1979 repted in an tedman a (os), The Value Conrovery, Landon, New Lett Book, 198 ‘Weg, Eric Oli Cla Sc and Income Detrmination, New York, ‘Academie Pres 1979 ‘Wight, Erik Oli’ Vans of Marist Conception of Cas Strut Polis & Soy, 93,1980. Wight Erk Oi, "Clas and Ocspto’ Theo and Soci, 1980, Wasnt Ent Ob, Reconeraton inl iceman a (el) Phe Value Contovery, London, New Left Book, 1981 Wight Elk Ol, “he Star of the Poi i the Concept of Cas ‘Sucre’, Poe @ Soe, 11:3: 1982 Wag En Of, Cpt’ Fuure, Socal Review, 0. 68,198, Wag, Erk Ot Giddens Crique of Marin’ New Left Review, ‘Nlimber 15,143. Wilt Esk Ot “The Fall and Rie of he Petty Bourgoin ited menace Cl Stace tnd Cat Comcnness Projet ‘Working Paper Wadon, Waco, 1985, ‘Witt, Bek Ol, The Comparaive Sud 0f Cos Structure and Cs Consiounes Pant Une Cosbook, Ante bor, Michigan ne. Univer Consort for Pot and Socal Resch 1985 ‘Wahi Ek Obs and Laca Pcrne,‘Marnst lar heres and Income Inegeaty Amencen Sonnogal Review, si Fobra, on Woaht Erik On and Joachim Singslnan, ‘Proletaanizton in the ‘ean Ca Stren Mra nga, eed ye arswoy aod Theda SkocpoSuplement othe America lurnal of Sociology ol 8198 fIndex Abaryon, 10-1, 38,107 ‘audio of detonate, ropa sate, 137-161 tape of destin, 22-25 ine tenting ‘Nicer tien 170 nef proato,19,12, eran 146-147, Fase oon), 176-08 (see sb, das coments sig pee reste) Bourges, 7 8, 34,37 39,4546 14de1s0 195,307, 26, 263, 21 (Se sho, aa eaopes) Bourgas eons, 82 epi 6, 836 om spittin, 69,75 338 ode puttin ve. soa ae ean tte, see. eee yates Sse esgere a cn, 7,107,168 Se ot expo, 36,75 Se a ‘and class conacioumnes, 246, ‘Erethatons, 154197, 197-209 Senn raion, i 3 see cm ety, 185-185, 19, coer ea ca Sree, ‘Sogn at Cas Concho, 14 eam ten 30 ‘esp Forno, 30-3, Indes. 339 (Seti, sation of cicada) ‘Sn on win ce ‘Evel of concept, «2-51 ‘epiamien cance clan ‘soca 8 pecems in omen, I-87 (Sees clamnncn e 3 tg of he GoshePropra, 72 Sis aie by agony and eden, gato eee ht mi of Lvs oder, 217202 Spann 76-77 Sender, 158 he sc ets nc40 ‘andes aire, 3-37, tine tery spt, 67-71 ‘cea 3,6 timate oan SOC ‘cion ‘eds expat, 100 rove righ 6-7, 80-81 Note ocean {Sab onameiionee si set) any 128-150, 225-232 (ee a, end, 7,3, $9, 8569.7, 75 T7249, 108 10,213, Finn enn, 3, $2 Te tso, 90 aa tReet, 8, 117-18, ital eaecry i, 2.35.58, Soro gh tox tictig ist Housenines 158 Se sho, eae ) (Se so, dss consomnens Inco 35 104 TET Te8 Ls 66, Inet (Se so, citer) Laur pow, 77 104 {ako ey of a, 99, 100,20, Lesion 118-422, 27 anger 8 45, 469.7. sot te IS Sine! tofu ewpyme, 206-207 cio eat: 173-18 Mare, 107,296." Maen 8 3368107, we earch ile den 13,802, 19,136, semen Fito ele of 8 Sidenote intrpenteaa 911-112 siete, 207 Occapcion, 18 ‘nd wtory, ‘ogists ‘ones ‘Reson concept of, 80-82, Se persion, 181-152, So (See, epic) fea 8031 Pent, 7, 61,70 105,18 (See ‘eed Wel) Pay borgai781149, 53,59, ‘1, 37,217, 25.180, 285 Poulan sconces 39-40 isnt dopa, Poi 22-225, 281.252, 264, at 38h Prfrencs 6,9, 287, 288-249 Procuejunpoductve labor (See at Faloa cs os). Fear a as, wing ace, 96,101 134, 200-201, 327 Settenptyment wens 17 i te. 58,54 Index 341 Soa Emp, 2.195. 198 (Sets tomas) Sec tne SSS many, 107, 0p Toned sai? son 3085 17 287-299, 38 exgitnn 70 onan) soviet Oi, cbt Sie 1 9, 122-128, 1936, “20 Spear, 233,29, pet tate, 122, 297 ‘b'day 202-203, Sse sus ut socet,35 7 Beton nh ta 0 epocaion, 8, 4 102, Sst proces, aed men 358-299 ‘Ses tcmgn, 211-213 Sones: ee Tat San erie fro 10661 sero 108 Seay eo, Super st, 208-208, 223 Fest 1 {esinologal teins, 31, $9, Ther of Surplus Va, 6 “hind wo ets 118, 126 Gee topes) Teli 10 2, "eo cat sonaounes, o8-204 2758032 cer ns, wae 8 agimesgttcme wap gh wiaamemerates teehee mecca eS ie Name Index sia sgn gpa an GRR ‘Woon (Seat, gender) hv of, 196,196,285 sen 8 gg Andon. m2, ‘Rei, ot, ‘edand Ste ‘derson 320 Gea 2816 ‘Secs 62,106,130 Bae, A.M 1328 ‘Soups te fic in ‘Gols 4.9, 0, 5,89, 1058, ‘Becker Fe Gordon’ Dy 133, 1340, 2918 Beton 8 ‘Grove. tbe ewe Be Sit, 9, 3026 Beverm. Hao, R636 Broo 1386 Halby, C1 Bee 30 rs 5 Hindi 130, 298-204 sole Comeoe 3, 2388 He P13, 293-298, 019 Saroeas 6. sie Holiowe i, 35 SEQLSIA. SA tom on. Haron My an open ae, 30 tie or Labour Easeion and erg ‘Koarad, G, 41, 600, 1039 ister, 59m, 2818 Teni, V. in 14, 269.90, 2910 3aaa ens, G. 298, 300, 3028 igstoee, D-W., 2810 ber, 38 1320 ake Sian Mana. hi 281n Mandal’. 28t0 ‘Maras Ky 8,10, 12-13, 16,160, "eeie 96,107, 13,118 am de 8 Noble igs Noss. 880 Off, C5 10, 2530 apa Bete ‘mate 38 oat 8 Sm 29,3028 NES, in 36 13914, 1 REA a Preeworiki, A, 178, 18,120, 123, Raman, ML 1350 Reem 3 9 605 6478.77 Rogers. 153, 2948 sinck ‘Singsinana, 3, 202, 238n, 318, 3230, Stendman, i, 9, 3010 Blackburn, 620, 1930 "eng “Thompron, E. P., 24, 38-890 Sen 2800 300 Ween Hs ethos 218 ‘Rr On, Sn nn 1890, 23Re,239n, 2949, 3026