You are on page 1of 149
CLASS STRUCTURE AND INCOME DETERMINATION ERIK OLIN WRIGHT New York Sydney San Francisco ‘hs ook ef eres pono by he nt fr Rosch Povey Cot © 397 he ord a en a te Une Won Sten sant ts Rea Ro ASABE NG le tal naa SEES AE Loony ero, To my parents research established atthe University of Wisconsin in 1900 by 4 grant from the Office of Economie Opportunity Its primary objective isto foster basi, maltidsciplinary research into the ‘ature and causes of poverty and means to coms It 1 cddition to Increasing the basic Knomledae fom which policies simed atthe elimination of poverty can be shaped, the Insitute strives ‘o carey analysis beyond the formulation and testing of fundamental generalizations 1 the development and assessment of reevant polley ltenatives, The Institute endeavors to bring together scholars of the highest Caliber whose primary research offots are focused on the problem af Poverty, the distribution of income, and the analysis and eveluation of social policy. offering staff members wide epportunly for interchange fides, maximum freedom for research into basic questions about poverty and social poliey, and dissemination of their findings. p The Insitute for Research on Poverty i national center for Contents is of Tables Eo Hares foe 1 THEORETICAL PERSPECTIVE 1 What is Class? ‘hodationa Vers Rebton Tore of Coe ‘Gees Define by th Tech Divisio bar Ver tony " Clases in Advanced Capitalist Societies 19 ‘Theoretical Perspectives on Income Inequality 37 (Class and income; Hypotheses 79 ‘Woking Cl adhe Pay ourenatsSeoiatnamens tapes) ao king Gs andthe Soup MongtoSugeenan o ‘Capit Clase and the Petty Bourgeois (Sal Employes) a ‘THE EMPIRICAL INVESTIGATION OF CLASS MEDIATIONS OF THE INCOME DETERMINATION PROCESS Class and Occupation ns Class and Income Spotean ee Ea chy 7 Them is Pt 8 Race and Class 9 Class and Sex 10. Conclusions Appendix A Data Sources “TP Sto nate Byes 50) “he Quay ampere! Sues 89S) Job Seniors 27 2a Appendix stat uforences Indes 1 List of Tables Base Positions Within Clie Relations Levels of Cont Within Ownership Relations Formal Catena for Conraiciory Lacations Within hase Relations {Gtr Used in High and Low Estimates for Sizes of Classes ‘hron Thooses of Income Detrmination Compares Distbuton of Occupation Within Case mm Distibution of Clases Within Occupations Distribution of Occupations Among Manager and Superson, Males Oly ‘Regression Equations for Comparisons of Class and Cccupational Situ with Ang Taal Income 8 Dependent Variable 4 Reprsslons of Taxable Income on Soluce Variables Analysis of Income Cape for Comparizons of Tote Annval Income of Warkar with Managers and Supersnare i sto Tales Regission Equations Wihin Clas Categories with Anil Taxable come ae Depandent Variable antsy etuins to Education for Worker and ManapersSupervisors ya Retumnsto Euston for Workrs and Managers by hog 15 Relums to Specie Level of Education for Workers, and Managers ue Tncome Gaps and Returns to faleation fo he Paty Sourgeose, Compared to Workers ad come Gaps a tums to veston for Small Epos, Compared to Manages al Workers 189 Industrial Categories and income Values Used in Inder Sele 17 Fucaton Coficients Among Himplosersand Other Clases ty Returns to Education for Employers mth less than, Cells Eatin we {Charette Sample nthe Hierarchy tn Organizations Study 160 ‘cones Within the Managerial Hierarchy, Adjusted! or Eaton. Agn and Tenure vs Returs (0 Elution Among Managers and Works, Canton tor Hoan Poston v8 Returns o Education Within Spwcfic Lavoy of the Mancgort earch 10 Distiution of Clas Positions Within cil Categorie 5 Regresion Equations Within Races Capote wih ‘Annual Taxable Income ae Dapenent Vara Retums to Edueation for Black and White Maes Within {las Posions 80 Occupational Distribution Among Mansgor for Black and ‘White Mee 194 Returns Euston for Backs and Whites Within the MansgerSaporvioe Category, Excluding Teachers, 15 ‘Average Income Cape Betoun Class Positons Within cil Categeries 1-194 Retuns to uc for Diet Cass Pstins Within Roca Categories ao» ‘Average Income Cape Between Races Within Class Categorie aoe Returns fo faton and Other Vaciables fr Mn und Women Within Clas Positions auaas Income Gaps and Return o Eduction end Other Vaibles {fer Class Positions Within Sox Cgoree, av Disihtion af Class Within Races Sex Categories 20 (Gaps in income Between Males ani Females Within Class Categories, 20 ito als sit 'b1 Citra for Cas Postion in he Panel Study of 112 loam the PSD Qurstonna Used to Oprtonlize lst Postion eae 133 Number of Employoos and Expected came an BA Charest ofthe Data St for Basie ‘hase Oparatonaliatione a ; List of Figures 1, ATypologyof Definitions of Css 21 Thefolatonship af Contaditry Class Locations to Basie Clases in Cpt Society 81 Formal Model of Income Dsterminito of see 312 The Base Model of tocame Determination fa Status ‘Atanpet and Human Capital Theorie 441, Antndividusl-eve! Mod of tncame Boteration for Comparing Class and Status 412 ast Model for avestgating Css Matias the Income Determination Process 443 Hypothesaed Relation of ince to Eduction 464 Rise Model of Class Meditions of Race and Sox Bets ‘on income 54 Retums to Bducation Within Clss Positions the Leva _— Relatlonship of come to Eaton for Worker. Supervaare snd Managers Fetus fo Glan fr Managers nd Workers in Difrent ‘Ayn Caos Me Ire and Eduction Within the Potty Bourgeise tn Path Diagransfor Relations of Education Indus ‘Seelr, and Occupational Statue to Annual loca Armong Employers 159 Income and Eacation fo Salt Employors with less than College Education vos lasiaton Schemsfor Managerial Positions 1 Distbutlon of Mle Respondents to Postans Within the Manger irre 1 The Uducatona! Green! i he Managerial Hiearchy vs 4 Pah Diagram fr Relationship of Education ad Hleerchiea Postion to Income (Masazere Onl) Path Diagram for Managers with Plan Ditectore Excluded Regrssion of Income on Easton for Mangers snd Workers 10 Retums to Education for Blacks and Whites in Different Chas Positions m Income Diferences Between Ros an Classes a6 Mean Income for Each Eaton Level or White and Block ‘Mola and Female Worker and Managers aa Motel of Deterunaton of Métro and Slcromedtions Withis the income Dstrtination Process zn Made! of Dotaenation of Clas Caneiousness 25 lustation of Mato of Sreturl Cmmparisns ES | Foreword Staff members ofthe Institute for Research on Poverly have con teibuted a great deal to our understanding of inequality inthe United States, For example in Public Expenditures, Taxes, and the Distrib Yion of Income: The United Stotes, 1950, 1961, 1970 (Academie Press, 1877] Morgan Reynolds and Eugene Smolensky trace the post-World War trend in inequality. Similaly, David L. Festherman and Robert M-Hauser in Opportunity and Change (Acadaraic Press, 1978) dissect, the intergenerational transmission of inequality in the U:S. and how that has changed in our time. Ek Olin Wright Cos Structure and Income Determination is e part ofthis growing body of research. But ‘heres difference: Wright the fire scholar al the astute approach this topic from a Marais perspective, As 4 consoquence, his Book is likely tobe even more controversial than others onthe ever-contoversal ‘opie of inequality Whereas most sociologists and acanomists who hove studied in equality emphasize occupation and education, respectively. Marxisi theory lesds Wright to emphasize clas. Furthermore. sine class Is defined in terms of postions within the social relations of production, ‘Wright focuses on the way in which different kinds of jobs particu leoly whether the job entatls being supervised, supervising someans else, or working as one's own bort—affet individual economic ole comes. Calling atontion to, and then demonstrating, the sigaficance Bf jobs in determining income is one ofthe important contributions ofthis work, Peshaps the book's most important contribution, however is that it attempts for the first time to test Marxist theory empirically. with ‘modern econometric techniques. As Wright nates in hie acknowledy ‘ments, faw Marist social scintists have the statistical training an skills a undertake such task, But Weight does He finds for example, that even crude measure of class explains at Inst as’ much of the variance in income as the move elaborste Duncan vecupational status scale. Similany, he finds that when class position is held constant, the commonly reported differential returns to education betwoon Blacks and whites and botween men and women virtually disappear No doubt these an! other findings lathe book wil stimulate cit ism and new research. Wright is slready pursuing the researc, be ‘use the empirical date used for this study were not ideally sulted to his purpose. Currently he is engaged in a major new data collection and analysis project funded by the National Science Foundation. The data will be gathered in four counties (he United States, Italy, Sweden. and Groat Britain), thereby making possible compartive epproach. This ‘book, therefore, may be viewed ss the opening shot in the lively in tellectual battle that its likely to stimulate Irvin Garfinkel Director, Institute for Research on Poverty Preface ‘This study bogan as an attempt to demonstrate to non-Marxist so ial sciontists that Marast categories mattered, that class was conse (quential for understanding American society. In many ways he quan tative investigation of income inequality san ideal empiical probs lem for this purpose. Quantitative studies of the causes and. conse. ‘quences of inequality heve almost totally ignored Marxist categories, ven though social Inequality probably plays 4 more central role in the Marxist perspective than in any other theoretical tradition in social science. Marxists have been suspicious of quantitative, multiveriote approaches tothe study of social reality. and the practitioners of mul variate satisties have generally dismissed Marxist theory a offering lite of interest for empirical research. The result has been that clas, ‘defined in terms of common positions within the social relations of ‘roduction, has never been systematically included in quantttive re Search on income inequality “The present research is first stop in bridging this gap between the Marxist theoretical perspective and the growing body af quantitative Studios of social inequality. As such, it will, { hope, have something to say to both Marxist and non-Mrxst social sientsts For Marsste ine resoarch represents a theoretical and empirical investigation ofthe la, ‘between socal relations of production aad social elaions of eachonse in advanced capitalist society. OF particular importance ls the anelite of various “intermediary” positions within the socal relations of ce duction. andthe relationship ofsuch positions to ncome inequality, For oneMarsists, the research demonstrates tht css position has aig, saificant and consistent Impact on income, Thus, even ifthe over ‘Marxist framework i not adopted, any thorough empltical investiga, tion of income inequality must sll include position within social rela tions of production as an independent variable inthe analysis ‘The basic theme ofthis study is that class dened ox positions within the social relations of production, plays central role in Iedioting income inequality in éapitaist society. This does tot mean that clas by itself i sufficient 1a explain all income variation. Indeed, much income inequality aceuss within class postions. Rather, the oe -aument is thal class onganizos the stucture of income tnsquelity, tothe Sense that class postion shapes ather causes of income The heat ofthe empirical investigation wil therefore be an analysis of the interactions botween css postion and various ather causes of Income, It paricu lar education, 'efore we can explore such interactions, however: s necessary to havea more precise understanding of what “class” relly means, Chep tert wil briefly discuss the range of meanings attached tothe concept of class in the social science literature, The purpose of this ehepter te less to provide a comprehensive analysis and ertique of alternative perspectives than to highlight the distinlive chorale ofthe Marat onception of clas. Chapter 2 will then attempt to develop a coherent st of eriteris for las position within advanced capitalist societies. The heart of the chapter is a ferly detailed discussion of capitalist social relations of production and how these fave been transformed in the course af apltalist development. This analysis forms the basis for a rroue efinition of classes, particularly of those social ategories tht are ten loosely described es "middle classes." Although most of thie chapter does not directly touch on the problem of income determine tion as such, it provides the general conceptual framework Tor the snalysisof Income in subsequent chapters, j | | | | Pace si ‘Once tis groundwork ld, we will turn in chaper 3 to a spe- ciiceomparisn of tho ogi of analyzing income determination within Mapiet and non Mors frameworks, The cental purpose af his chap tori to male I as clear as possible precaly howe harxsts ose the ‘problem of income eteronaion and Mow ths sate of analysis Fom buh conventional soilogie! and economies apron Thee {his chaptr wil mae the ompiria analyses that follow more acest He tornado latvelyuanalar vith log of ai hens haptr@ wil hen eth gener! nvayes af els stata in chupor 2 and the aproach 10 anaiyaing income inegualty prevented Inchaptora in oder develop sera of concrete hype boat the rationship wenn cls and tncoma deletion, he feneral toy wil bo to show how postions within he socal ee tins of production influence the was in which factors such oy educa tn are likely to fect income, The general anal wil hen be fxonded to form saris of hypoteses about te iartelaonahie tween clas and race and class and sek ho lncome determination E Chapter 4 will be followed by five empirical chapters. Chapter 5 presenta diet comparison beeen class poson an occupation Aatus ab predictors of income, The baste conclusion is ats sey Simple opertonaliation of Marxist cles ategores ca lets pow ssl avaiable in proicting income veiation ss the elaborate Dun Gan occupational satus sete. Chapter 0 explores the te less teractions wih the tncome determination proce Ii foused that the fetus to ecaton vary eanalerably beeen Classes and that these interctions cannot be considered “atfatss of he chrecterstics of the individuals occupying class positions, Chapter 7 er Toaks i de tall at the reloionship betveen specific posons within anagesl erarchios and income. Much af the genera Interpretation ofthe Tink Between elas rlatins and income inequality developed In chapter sevelves around an analysis of the logic of birch. wi the Gepilist production process The analysis in chapter allows or poral detest of ths interpretation Finally chapters and apply the general categories developed in le chapter oan apelysis of race ad sex eet on income clas ‘ally does ply afandametal etiting vole inthe stuctur of an come inequity then it would by expected that clse position would bs Important for understanding income inequely bene, tacos and sexes. One of the most sgoiflantfadings inthe sty itt the Afra rears to evetion bowen tacks and ites nde tween men and women, which have been found in virtually every Sud of race and sex effects on income, disappear almost entirely whe, lass position Is held constant, ‘This empirical investigation will not“prove” thatthe overall Marx ‘st theory of capitalist society is core. But it docs demonstrate that lass has a systematic and pervasive Impact on income Ineausliy- We trust the book will show that to ignore social relations of production ie Stratification research is thus to ignore one ofthe fundamental dirse sions of social inequality in capitalist society. Acknowledgments Acknowledgments usvally end witha caveat: Although the euthor is doeply indebied to friends and colleagues for valuable suggestions and citicisms, al ofthe errors inthe work ane his or her own tespons bitty. 1f the production of knowledge is genuinely soiel process then both the strengths and weaknesses ofa work must be understood as boing influenced by ts collective eeting. This book should be evaluated in such terms, Just as the new insights and findings cannot ‘be seen simply as tho frats of my oven reflection, so the limitations of the study should not be seen simply a9 failure of my own imagination, This wotk is tho product of an historically specific intersection of ‘cademic sociology and Marssm; its strengths, and ils weaknesses, refect that seing “More concretely, tho strengths have grown out of a period of in 'ense debate over fundamental questions of Marist theory, The varius xxiv Acdawalegmens ideas in this study have been subjcted $o many rounds of eitilsn in various study groups seminacs, and canferences, and al the end of such 4 process Its Imposaible to identify those idens that are may on id those that grew out ofthe discussions themselves, The weaknesses uf the study. on the other hal reflet the reltive underdevelopment of quantitative tesecich and sophisticalion among Marxists. This has hampered the quality ofthe sponse that [have received aa the more tampirical parts ofthe work from people commited fo ls theoretical framework. Many Messi til regard quanitative research as inti cally "undialoctica” and thus en inappropriate strategy for advancing ‘Marxist social science, and those Marxists who are more sympathotic to the endeavor generally lack the statistical skills to ge inside the emp Jcal argument. This study would undoubtedly have boon better if is empirical steaegies. nd the Tinks betaeben the thoary and the daa, hod jon subjected. to. discussions as lutense as was the theoretical Feamowork itself The origina rsoarch in this study was conducted for my doctoral Aissartation in the Sociology Department at the University of Califor hia, Berkeley. Brom the start, Arthur Stinchcombo war especially su portive of my wark, and more than anyone else has taught me how to Tink quantitative methods to theoreliesl substance. His insistence that ‘what is interesting about Marsst theory is whether ar nt its Irue” has constantly pushed me to clarify the connections between the theoretical argument and the statistics! Investigation, Barbora Heyns's ‘good-natured skepicism about the importanen of soci welations of production and exploitation in understanding incom inequality has forced me to make explicit many of the assumptions underlying the analysis Michael Reich's comments and criticisms have alsa buen ex ‘tomely valuable. More than anyone els, hes pushed met elaborete the politica implications of the resesrch and avoid geting bogued down in puroly scholesic issues. And Tom Rothenbers's endless tapacity to solve economotic problems encountered at each sage of the rescarch and to explain the solutions in ways I could wnderstend facilitated the tochnical pat of the research immensely. ‘Many ofthe coro ideas inthe study, especially those ln chapters 2 tnd 4, weve formed through my participation inthe editorial collective ff the journal Kopitlistate between 1973 and 1976, would ale ike to thank the many people who have given me writen comments on var (04S papers thal served as the bass for several of the chapters Ia th Stud Marcia Kahn Wright, Ron Aminzade, Sam Bostes, Wins Beene, Michael! Huravsoy, Roger Friedland, David Gold, Alex Hicks, Bob Jackson, Robert Kahn, Rebecca Kharkov, Andrew Levine, Ruth | | | | kent akman, jim O'Connor, Claus Offe, Nicos Poulantzas, Aage Sorenson, Ma Steuve, Al Seymanshi, Maurice Zeitlin, and Rob Mayer, "All of tho data in this study were gathered by the Institute for social Research a the University of Michigan. Without the cooperation Stabe lastitute it ould have been impossible to conduct this empirical, {vestigation of class eelations and income inaqualit. In particular, 1 fim especially grateful to Amie Tannenbaum for letting me use the [rmrhy in Organizations Study daa, and o James N. Morgan, projec, {instr of the Panel Stacy of Incame Dynamics, for his interest inthe prublom of authovty relations on tha job and his willingness to include Mhumber of questions that tapped class position on the 1975 panel of the study. 1 would also like to thank Grog Duncan, Rober Quinn. Linda Shepherd, and Graham Stalnes forthe? assistance at various stages of the Fsearch Parents are always pert of thir children’s accomplishments. But feel that lowe both of my parents an especially deop debt fr eneourag ing my intellectual development for tio decades and giving me the self-confidence necessary to make my own way and discover my own truths, In particular, I am grateful for whatever it was they did that nade me enjoy iting soi Ackonwtedmens Finally, { would like to express my gratitude toa former fellow graduate student, Lsca Perrone, During ou fis! thee years eto School, we coauthored most ofthe papers we wrote, This tock wate alirect descendant of one of those efforts (Wright and Perrone era) oe fae point, we fantasized writing aoint dissertation, each ofan sehag {ing the same document for his doctorate degroe Inthe onde henna: re wer spared the trauma of tying Io convince the unlverlty aceep soit thesis, since Luca returned to tly to auch sncoogy the University of Calabria. Nevertheless, the basic strategy Gt nelyss and the cor ideas ofthis study come equally fom haf 'n ander tobe embodied more immediately in the final ote ey ‘hs ent me he grophic usration on page xxv of the fundemontal en lusions ofthe research, Although this drawing may seem someuioy rvptie at this point, hope that by theend ofthe study la masning ra be apparent. I THEORETICAL PERSPECTIVE 1 What Is Class? Soclology’s one independent varie isa chameleon which blends Into virtually every sociological wadition. To some sociologist, class ‘elas to categories of people occupying common positions within Salus hierarchies (Warner, 1989: Parsons, 1970: Willlams, 1960). To thers, classes are defined as confit groups determined by their posi- ion within authority or power structures (Dahrendor, 1959: Lensk, 1986}. Sociologists within the Weborian tradition se classes as groups af people with common economic "life chances” (Wober. 1922: Cid ens, 1973; Parkin, 197), And Marxists have defined classes primarily in terms of common structural positions within the social ongnzation of production (Bukharin, 1721; Lenin, 1004, ‘As Stanisfaw Ossowski (163) lis emphasized, these diverse in lorpretations of clas va not simply eoflect ifering eins ao ty Cases and consequences of a particular plinomenon: they sli sae sent diferent aims about the way inaquslly shoul be conee ‘wulized in the frst place. The concept of clans is not simply «concn concept It isa essentially eanfused eon (eee Plan. L978) Th bute is over the very object of investigation shal lis caneypr oe ‘lass donotes—an not simply over the formal definition af an agrees upon ule This chapter wilt to sort out the salient theoreti serpin ‘ing of hese ince nthe dtiniion clans The pp haptr is tno udjuiete beter the competing delitions ieney int clarity rigorously tho dstinctvenes ote Nari eon ot los. The vaidatfon of such Martel concept of las mn ae thenigh & demonstration ofits capacity to foveal the. unease dynamics of soa proses (ts to explain the won and nl sah {trough nconcptaat argument The omplraliestiaon a neo, ‘quality inthe second alo hs bok wil stomp to scoops Sch #demonstistion. or the moment thes esp to spy te terncebtween the mesnny of cas thin Sorte Ico se te AN the risk of some oversimplifeation. the diverse efations of lows can be analyze tors af tho nest thew dselons (1) Whether last ty fundamentally understood in saaons oy felaional terms: (2 Welas s understood in telatinal toms, whether the pla aspect of eas relations seem load ne mart or is roction: 3) lass elalion= ae pra locate within prs {ion hether production is analyzed abovedll in ternsuthe wetnte division of labor. auth elton ur explotaton Thess thos theoretical dimensions goneat ve hse Iypes uf dtniany of las, silanated Is Figure | ea too ln GRADATIONAL VERSUS RELATIONAL ‘THBORIES OF CLASS ‘The distinction between gradational and eolational views of class isa familiar one in sociology: Using slighty different terms, Ossowski [1963) emphasized this distinction in his important study of concep Hons of elas structure, Theories of class, Ossowskl argued, could be Aided into those based on “ordering relations” (radational views of ‘lass and those based on “relations of dependence” ( Inthe rst interpretation, “the lass division is conc into gzoups differentiated according ta the degtes in which they pos sass the charactoristic which constitutes the criterion of division, as for Instance income-level{p. 145)" Inthe second interpretation, on the other hand: “social classes form a systom according to their one-sided fr mutual dependence, dependence being understood in both cases as ‘dependence based on causal relations [p. 146). ‘The hallmark of the gtadational vow is that classes ate always characterized as being “above” of “below” other classes. The very names given to different classes rflet this quantitative, spatial image: upper class, upper middle class, middle class, lower middle class, lower class, and so forth While there may be debstes about the extent to which thee divisions are purely conventional of real (th famous problem of continugus versus discontinuous gradations within systems af statfcetion), the basic conceptualization of classes remains the same: classes, in Barber's (1057) words, are “divisional units within ‘yslems of social staifleation fp. 73). Within rontemporuty sotioogy. hens Have fa tase eR ons of tila conceptions of ess: ne fis ass sting primrdy tera of inn thw aller snails tors of sad Matis Te former is walt the nut cnn pol into ft lass jour pwople consti lyscer clas: millscenne people ‘nsttote te nile ses ale pops ennaitt the pp ae spersnd Hackl [1970} essential lp His view alien Hs we ina class system, the social earch i ae pia. hi Toners in aontary vat enn Is ES = Wii sich eo taption. he shape of the lesen ones vray ential tp th shape of the income iseibubion The feet las at the Unit State fe nea rans fe hu mildly suits mong froma" psrami-shapea fo liamonulsbapert cane See usualy adopt at fst ipl, much 4 eaerpinn clase ‘unt sociologists oc those uring Tetly wil aston lino af clas donot mea he ene sretone tn un linens, Tir st como rotational cneaption ts thal hase ston reflect came positions ssithin sat hess. As Parsons (10) tcrlen elas shot bar defn a “ntact of su unt ed dual andor collective. that in thei een eatimston and thos of ath fr in the society wecup: positions of approstiutely ual status Ip 24," In somewhat sionpler languages, Willnaee (1900) as eins ‘asses In as terms In canteas to these gradations notions af class, relational excep tions define lasses by thelt sactued social relationship rather ses, Clases re not finest simply rela a ether elses, ina Social relation to other classes: White classes may lifer empirically along @ variety of quantitative dimensions, the criti for lise ate tai on qualitative difinences. Aue as in the sadatlonal perspec ve the ery names nf cases thin relainsl views of eas mile this undeniving definition, Glasses are wt labeled long «cantina From lower to upper: instead, thov have nates such ass csptlist, class, working clas: Lond, sor ruling clas. subordinato cas. f0 aradational view of lass, Ieee castes are wimp dans hag Ts of something that uppor clases five mone of income, woth education, satus-—but within reat view, the wurking lis 6 aniecent , ie locaton wHthin a soctal relation tha sane etme tins the capitalist clas. Within Weberian theory, for taneowtl irre are understood ot seller of labor power, capitals aoraeees of labor power. Theiss ts nol that workers have lest of 28 bag than captliss, bul rather hat they occupy spaciic quale See on within acacia relalonship shiek detines boi he tatifibt and the worker the soca elation of exchange onthe Labor It std bo argued thn a ns imply, hae i withing atonal vow of inequality» polio of socal elations. After allt at someone hs “mnore" socal stats than someon ts that both people agree on the relative rankings of satus po tras; and thismplies that thetwo people exist within a soca lation the pewon in the lower statue acknowledges the person in the higher tows as having greter stats, and vies verse. Even fora dimension of insquaity 8 seemingly gradations as income, ttean be ard that a pron’ income is Tower only im relation to someone elas higher Income thus theo is an tpl relational aspect of inequality within income gradations icon ofisasy stby diatn capa suine aot Clase are bused on th qualitative social relations which define such featvepostlons or on the quantitative dimansions which are gen tied by such postions. Cradational perspectives all organize thelr dele inhlons of clas sround those quanitative dimensions rltioal iow. in contrast al tempt diet to map the socal eatons then salves ‘At one lve. it appasts thatthe debate batween rolational and trsdatinol conceptions of class purely semantic one. dearer tent sbout how the word class should be used. At deeper level, however. ay in many apparently semantic debtes within soral sek tna, the dsnarcoment aver the us fa orm lets more fundamen tl disgseement over how to study the world Relational conceptions ff lass al insist in one veay or anathon that th base structures of Snagualty ns society areas structures of interests and hus th ass for collective social action. Social relations do not simply fine lasts, the slo determine classes clases at saa foros re rel Consequences of social relations “clase sucture defined in gradations terms remains fndamen. tlly «sate trxsonomy, Such dafinitions may provide a basis for SeipvelyInbeling people in tums of the dstbution of valued ‘ards, but they are nepal of designating te dynamic soil forces defind by ss quali Wie detormaiae anal transtor tha dstibution To give just one sme pile example. it hard to see how the Fro Revelution cul Bo fexplained in terms of yetational schemes al elas While it might be th ease thal mast ofthe participants in the storing of the Haat hag Salus cote af undee aan nist of the French arisociey heel tenes shove 7 ch labols do not capture the unleriving dynes a wrk Tn the revalutionary process. Thn decisive actors inthe revolution wong people defined by these position ssithin qualitative relatlons [nls Peasants, merchant capitalists, professors tty Bourgas and sana Ceotes| not by thelr Fuation on 8 simple, quantitative disension, All relational views uf class, ngortes haw they cance of ose relations, see clas structures asthe potential bass for eallective class struggle class ativty- Mars’ dstnetion between a clans and a classfortsell, DahrendorTs definition of clases ts anil groups Aetermined by authority relations, and Weber's voneoption of clase se the potential bases of communal setion al atlemypt to link the analysis of class structure to a dyesmic theory of class strut, Cradational Aefinitions of clase are whally inadequate fo this ask Knowing thata theory of elses eel ing of inequality, however. Is only the starting point, A avid variety af social roletions have laid laim to constituting the coltional bast for Classes: authority rlations. occupational relations, market relations Social relations of production, It is. therefore, necessary 10 ‘tke thooretical distinctions among relational conceptions of lass (CLASSES DEFINED BY MARKET RELATIONS ‘VERSUS PRODUCTION RELATIONS. In the theoretical debates over relational definitions of clas. to spheres of socal relations have vied for the role of cunstitling the foundation ofthe cass stractuee: market relations and prodietion eax lions. In the most general terms, market relations ace defined by the relations of exchange between the sellers and buyers various kings of ‘ommodities, Production eelations comprise the teats between ae tors within the peoduction proces tcl. ‘The classi formulation ofthe market conception of class appears ina chapter in Weber's Eeonomy- nd Society [182) entitled, “The Distbutlon of Power in th Community: Class, Status and Party. Webor sites vit ne? at et ta ate ofthe implications ofthis definition of clas tht casos cn also tt only soa i wc ee gine mat rar es copia peter soci, ey eat win ‘onl sons ewod a clssn ince th srctra bss of her ie ‘outa in mart sition ‘Weber's basse motion Ot clasor redline by captalit ex chao elton ha buon exten by eet thar na mb a chante (100, for example os argued tha lasses in Aten [iy am deominod by tres siorent nterseting markt: the SE elaor te make for cota the arte ommodiin FRE tee markets define sx clasens employers (apts) and Tints rectors and ceo a sles and conser. The me “1S hweimensons of as exchange relations) overlay. Wiley gos the more intense clas confit i ily ob "iden (1079) emphasces Webots srpoment that "marke capa ay" dtine not simpy bythe pooesion of expla or iabor power, also bythe possesion of markteelven hls fe ano Parkin MSriepp. 30-250 Ciddon dines markt capacty stall forms of ‘een atte sich inviala may bring to the bargsning en Counter {inthe markt, p. 1031" He then dacuser the specif SShucice which shape cates captalit sociey Sion mt capacy is etended expt nade sll. he the make ‘atin fle lly eich dations are er ecepaonl ‘sen, ine Sil lee en af he ways leh scapatone ae ‘Seinen ca sn a Fem nk hen ot ‘SUE a ty se thing of esas nm fara opens more han ‘neler pr itor’ very feces of ecupaton ad ae ‘ico provides aera hare mone oh ge of eats ols ‘Senso torreon ponte Saaw exaey that ann ees ‘cre alt Toran dein cla nd patios Ma 878 scoters Giddens then extends Weber's analysis hy tying to Tnk the cone pt of classes define y market capacitis to an ana of Sui relations within the production proccss sel. tn parila ela fates a numb of sctl proceses which he labels sours fps: imate strtaratin” of clas rlatonshiga, Two of thee det spect of eolations of production: the division af labor within The a texprise and the authority rlationships within the enenpese Gens irges that fn capitalist society such “proximate structuraion over with Ihe patter market expacites a has fads ta inet lass divisions defined by mre elation. Homever inspite of this nfurmlation of Wabur's conception of ‘lass, Gon sil oes th capt socal argon of economle felations aa fundamentally define by oxchange reatuns, and thus Uke Weber, he soes castes sail fundamentally tori atthe lov of the market. As a eesullelans stage ae scr primal ag imorket srvgges. Ver strssos thot the distinctive lise stole Within cepitalist octet ls "wage disputes on the labor market 11068 fp 1S1F"and Giddens emphasizes that of predominant importance in sccologca ars are the spn vert cunt which are Hnked ta ‘postions af laterest entailed by differing forms of market capacity [nu p 1951" Struggles within the production proces ih ne force auch markotbese conflicts, but the prime arena of lass cont is tls ta of production sl ‘i cantast to Weberian conceptions of elas a market relations numberof differant theoretical radians hav arg that the heat oa lass analysis must be lucatd wlhin the sphere of production. Al though. af we will sev Below, there fs le agreement among such perspectives about how production tel should be theorized, tn al ines there fa a recognition thatthe relational base of social conc, fad thas of clases, should be sought in the structure of production rather than simply in he siructure of exchange. tm one way er another, Coch ofthese pradution-levl there ngs thatthe Hosni Droddctionvefines decisively the command over socal resources ond Socal action Within such theorles, market relations may sil be of theoretical {nterest, bi tht interest seve fram the lationship of market 0 ‘rodction. Generally speaking. thls rlttonhip ie conceived in aie n EF tn rena wor 60 tapers Ia Dlplagw8 t cu individuals a sovted ino the pruton pos Te ee tran tet ie sour ft vations Py infence the wage which cates came ot sor Mollectvely. Secondly, matkets are soen as one of the impor fgantanan within which classes engage In struggles. Workers are. as er paar Unda commodity, bot Weber once exptalist attempt ia buy that commodity for as Ite ore ade ee wanes bacowe encase of pose ons cr ta lee win tensa il Ne Ae eget he pesto Bt ter tree cousin sch cance a way CLASSES DEFINED BY THE TECHNICAL DIVISION (OF LAHOR VERSUS AUTHORITY RELATIONS VERSUS EXPLOITATION In order fo put real content on the claim that classes should be defined within production rather than within the maeket tis neces: Sry fo understand what ts about the organization of production that form the basis for Uhe determination of class. Three different ways of Understanding tho stricture of production relations have dominated the analysis of clases within production: production is defined primarily as 2 system of technical dsislons of labor: production is Snalyzed above al aa system of authority relations, and, production, fnsofar as it dtormines classes, is seen fundamentally 4s a system of explttation. Let us look at each of these in turn. (lasses and the Technical Pethaps the most common of all definitions of class among sociologists is that based on categories of occupations: blue-ellar oc {upations define the working class; white-collar occupations the mi Ale class: and professional and managerial occupations the upper mie tle or upper classes (or sometimes even the "professional clas"). The theoretical status ofthis occupational typology of clases is ly not very clear. Somelimes occupations ere basally viewed vision of Labor as status categories: in that case this conception should rightfully be Seon as one variant ofthe gradationol view of lass. This is paticuarly teue when gceupation is sealed os an occupational status o prestige ‘variable and then treated asa measure of clas (or cass background). In bother situations, occupation i teated aa « proxy for market capacty, fn thus forms part ofthe definition of elasses in terms of exchanges felations (soe below). Buta last among some theorists occupational tategories are seon ae defining clases by vite oftheir lation within the technical division of labor (or technical roltions of production}. Since, itis argued, in moderainduseal society the technteal relations ‘of production determine the conditions of work. the command over resoutees, and the relative power and status of different postions i the social structure, and since occupations represent similar locations tathin the technical division of labor, ccupations should be consid ‘re the structural boss for classes. Probably the most important contemporary version ofthis concep: tualization of lass ean be found in certain theories of “postindus- tralsm.” Bell (1973), among others. has argued that in’ advanced sages of induseial development, expets of vatioussorts—scientists fengineors. certain categories of tchniians—aeo gradually emerging as ‘new dominant class. Thole position within the technleal relations of production gives thom a monopoly of scientific knovledgo, which Bll argues, enables them to contol the key institutions of the postindus- tral society. In deliberately exaggerated manne, Bell describes the class structure of postindustralsoctties as follows: ‘Touraine’s (1971) analysis of tochnocracy follows a similar logic, al though Touraine tends to be somewhat closer to Ralf Dahrendar In emphasiaing the tole of bureaucratic authority in the deftaltion of Class In any event, for bth Touraine and Bell, the role of experts and technocrats within the technical division of labor becomes the basis for defining them as a class in postindustrial society "Thane ave relativuly few sustained theoretical reflections on the logic of linking class to postions within the echnical division of labor Pethaps the most influential theoretical rationale for this conception le found inthe classic inelionalist account of stratiiation by Davis and Moore (1935) although the authors do not sjstematcally dlecus clase 43 Such, Davis and Moore attempt to understand the stractoral asi for ictributive inequality in terms ofthe functional importance” of pos. 3 na tions within the technical division of labor. The logic is that unequal rewarls are needed to induce people to fll the functionally most im- port positions, and that the functional importance of postions is Uerived from the techaveal imperatives of production systems, With inisimal extension, thls can become aa argumont that the clas struc. {ure fs ultimately based on the functional imperatives ofthe technical organization of production Inauthority definition of elas, the social content af lass eelations tefurns tothe centr of the stag. Classes aro understood as based di feetly on a system of relations of domination and suboedigation, and ‘while those foations may be shaped significantly by technical con ‘mints, tho classes themselves eannot be defined in terms ofthe tech nical division of labor. ‘Mate than any other sociologist, Dahrendorf (1054) has cham- ploned the conceptualization of elas in terms of authority eelations clases ar soclal conflict groups the determinant or differentia speci- siea| of which can be found in he participation in exclusion Irom the fexerise of authority within any imperatively coordinated association [p. 1981" Within such imperatively coordinated associations there ave sways two basic classes “command classes and obey clases. Since in the society at lange people generally belong to more than one such Association, i is likely that many people will occupy command pos tions in some associations and obey positions in cthars. Tho ovorall societal class structure, therefore, is likely to be a complex web of ‘crosscutting class cleavages based on intersecting structures of author ity zolations in different organizational settings. Lash adopts a similar postion to Dahrondert although he tends to pursue a more eclectic usage of "lass." including vane af other dimensions besides euthorty. Lonski (2966) first defines clas broadly fan aggregation of persons ina society wha stand in a similar pos. tion with respect to some form of power, privilege, or prestige Ip. 73 He then goes onto say that "if out goal isto answer the question ‘who sels what and why?’ power esses must be our chiel concern” ‘where power class is defined as an aggrogation of persons ina society ‘who sland in a similar position with respect to foroe ge some specific form of institutionalized power ‘Several general characteristics of authority definitions of elass are worth noting. Fist, authority definitions of class tond to toat all or- ganizations es conceptually equivalent. Dahrendost in particular sexs Classes es being defined by authority rlations in any imperatvely 1“ hase Stuctre ad tc er coordinated association. and provides no clterie for ordering those ‘associations into those which are contralto clas structure and those ‘which are peripheral ‘Secondly, authority definitions of class tendo soe authority itself 85 a unidimensional relation of dominationsubordination within 3 Bivon organization. No systematic theoretical distinctions are made concerning the object of authority. What mattors i having euthorty of power lite ig said about how its used, Conceptions of else in terms ftauthorty eelations thus ten to emphasize the form of class eelations fover the content of those relations, Finally, because of this formal character of the conception of class authority definitions generally do not provide a sustained account of shy socal conflict should be structured around authority selations. Inmplicitiy, one of two arguments is usualy made. Either ti assumed that human beings somehow have a intinsie drive for power for is ‘own sake, and thus the division between the powerful an Uhe power Tess iteinsically constitutes the bass for social cleavage: or ts argued that power andl authority enable the powerful to appropriate various Kinds of resources, and that asa resull the poverless will attempt to gain power for instrumental reasons. The evidence for the fist ofthese assumptions is particulary weak. People may have an intrinsic drive to control their own lives, But there is little evidence that most people hhavea basic need or drive to contel other people's lives. In any event, ompiricelly most stugsles over posser are struggles over the use of power, nol simply tho fact of power. The second assumption is thus ‘more plausible. But in order for it to provide a sound bacs for an explanation of the relationship of autharty to social conflict, itis hetossary to develop a systematic theory of the relationship between Authority and the appropriation of resources, Most discussions of a thority lack such an account. This Is precisely what the thoory of ex plaitation is meant to accomplish Clas and Expl The hallmark of Marxist discussions of cassis the emphasis on the ‘concept of exploitation. In leer chapters we sill discuss exploitation ta of hii pis ct ose Dana hay ho Uta cases canbe viewed ae simpy eamomicentogure Nove, tosis tobrand up ha gmc way a untae the stl rennin of Production (sae yur of ath rston) all aca wok - hat cn? 15 in much ateater depth: here it Is sufficlent to define it in very general lexis, Exploitation within Marsist theory denotes a relation of damn tion within which the people inthe dominant position are able to “anpropriate the surplos labor of people svthin the subordinate pos. Hom. Such labor is generally appropeated In the foe of preducts pro ‘duced by thot labor, and thus in-many instanews the expression surphis product” is used as an equivalent "surplus labor.” Surplus Tabor of surplus product n this context, refer to labor above and be- yond that which is required simply to reproduce the individuals who perform that labor. * ‘Why is tho capacity to appropriate surplus labor of such signti- cance that it can be considered the cote of the definition of ess rela ‘ons? Several reasons can be given. Fist. tho capacity of a dominant ‘lass to control the surplus makes it possible for members of that class to consume without producing (ora least to consume far in excess of anything that they produce). The contsol over the surplus product, 38 ‘ve shall se in later discussions ¢ thus one eitical bas fr the dist bution of income across clases, Secondly, the contol over the surplus product gives the drninant class substantial socal and political power beyond purely economic concerns, both bucause t provides meter resources for political activity and beeause it shapes the economic Framework within which social practices take place. Ultimately this lumplies that control over the social surplus product gives the dominant class the capacity to shape the direction of social change. socal de Yyelopment, This is most obviows in the case of material development since such development comes ditetly at ofthe use ofthe surplus (Ge. investments), But itis als tre for political and cultural develop: mont, since the use ofthe socal surplus directly and indiretly can. strains their poscble directions of development aswell ‘When lates understood in terms of relations of exploitation, the initial task of an analysis of class structure isto undorstand the soclal mechanisms by which surplus labor is appropriated. ‘The Marxist theory of modes of production i designed to accomplish this task “Modes of production ar dilferntiated fundamentally In tems ofthe central mechanisms through which dominant classes appropriate the Seplus labor of subordinate classes, For example, in classical feudal Socleties this labor is appropriated through forced labor dues: in Capitalist societies its appropriated through the difference inthe labor 16 Cm Stat ad acme Determination time embodied inthe wages of workers and the labor time embodied in the products produced by workers. (The logic of this claim will bo ‘isciseed more thoroughly In chapters 8 and 4.) ‘Once such mechanisms of exploitation are adequatly identified then the analysis of class steuctureitsof can bogin. Lenin (1914) pros ‘vides an extended definition of classes based on this concoptualization: ‘lase t age grows of ola which le um sch ar by he plo thoy ncupy nna eterna sytem sl pct ha {ny hol fo esol ganz feu coment. by they dean, Classe groupe people one of which can appa te Ibert acter swing the sift plas hey escapee ‘sem sl easy TP ee “The hear of an analysis of clas structure, then revolves around defin ing for every clase, the content ofthe “diferent placos they accupy ina {efnitesystom of social economy. ‘Within such an account of clas relations, a discussion ofboth the technical division of labor and the authority relations within produc ‘on will alsa playa role, The technical division of labor enters the story ‘as we mill see in chapter 2, one of the cetial aspects of the laces” within the system of social economy is thei capacity or in- capacity to contol the technical organization of production. To say that ‘workers do not “possess” thelr means of produetion in pat at ‘means that they donot have the capacity to shape the basic contours of the technical division of labor itself ‘Authority relations enter the account of class structures since within the capitalist mode of production the capacity to command Tabor (i.e. t0 tll workers what 16 do and be abe to imnpose sanctions If they do not do It Is an essential requirement for being able to ensure ‘that surplus Ibor is etually pexformed within production. A capitalist ‘teense deo! cles stem ones at con qty afeeat ‘nechaname of elon Mart hay cal et om» prey hao pa bya marchsa devon owt tary voted amar by ovis sped wth ifn rant moe of osu. The srr of Seen ‘Ehret prad ther av ot uh hinge asus le asta see” ein io, Sao, dino ae ‘Uren of dsr stacey saa in cea ernntion Toute canbe vowel a fon! concep of hl soe palais {Scones of prncson conta ft anal quate ius a the Y ray hive workers for eight hours, but unless the Ibor of those warkers {s controlled within the production process (Le, unless they are subor Ainated within authority eeitions). there is no way of ensuring that ‘hey will perform anything nar eight hours of sctual labor. Exploitation views ofthe cass stecture therefore incoeporate both technical and authority definitions. but subordinate them to the dynamics of control aver the surplus product. Classes. in these terms, fate most pivotally defined by the rolations of appropriation of the surplus product and secondarily defined by the eeletions of contol ‘over the technieal division of labor and relations of authority. ‘This chapter has tied to establish the distinctive charactor of Marxist definitions of class. To recapitulate: 1. The Marxist concept of class defines classes in relational rather 1 gradatonal terms. Although classes do differ along varias que Luative dimensions, the fundamental theoretical criterla for classes are tative location within social rela 2 Within the Marxist concept of class, the central axis of class relations is located within the socal organization of production rather than within the market 3. Within the analysis ofthe social organization of production. Marxist thoory roots the analysis of cass zlations in an examination of the process of exploitation rather than either the technical division of labor oF authority rolations (although both of these playa vole lathe theory as well. Classes within Marxist theory, in shot, ar defined as common posi tions withia the social relations of production, where productions analyzed above all 35 system of exploitation ‘As showld be clear from the discussion of altemative definitions of class, the Marxist definition ess on @ number of pivaal assumptions ‘in particular. that economic relations ple @ basic role tn structuring (settinglimité upon) otherelations. and tht within economic relations, the social relations of production stuctute both technical relations of production and social relations of exchange. If these assumptions are Sccoptod, then the Merxist definition of elas is very compellings If they’ are not, then this definition has no privileged claim aver other possible definitions “There is. of cours. no simple way of empirically “proving” these ‘assumptions. While its possible to establish thelr plausibility and to ilustate them by eacours to historical examples, shard to imagine «critical social or historical “experiment” which would directly vali date them toa skeptic, These assumptions thus constitute paradigmatic premises, fuse Kulhns (470) formulation, and as such thoy are not Eubject to immediate validation or refutation Instead, they should be judged on the bass of the coherence and power of the substantive theory ofelase relations which isbuilt upon thom. The cental abjective Dt this study ls to take one particular theoretical problom, income de termination. and demonstrate this coherence and power through a sys tenuate empirical investigation. ‘Before we can do this, however, tis necessary to develop more rigorously the Marxist conception of class relations. Although the det hilon above may adequatoly differentiate the Marxist concept of class ftom other definitions, tis nol yet precise enough tobe used in emp al study. How should the specific socil relations of production of Capitalist society be defined? Once defined, how ean they be oper- Stionalized for esearch purposes? What concreto criteria define the ‘various postions within the social relations of production? These and ‘ther related questions wil be discussod Inthe following chapter. 2 Classes in Advanced Capitalist Societies We mot of oars nol Be surprises fring emir patting la tetany by th pollon of the clsss i the proces of rotation Theor wert firth ca onto The previous chapter focused on the differences beloreen Marxist and other conceptions of clas. In this chapter" we will ook more ‘deeply at the Marxist conception; in particular, I shall try to elaborate rigorous understanding ofthe clas laetion of various positions in the Sct strctare commonly called "mide clas"The vague and often inconsistent discussions of auch postions by both Marxist and non Mardis these have been a sour of endles eanfuson In order o dhvelop» proper clogs map of advanced opitalist setts ii esson- tet hat e pris definition ofthese positions be developed ‘The chapter wll bon by elaborating tn some dat he general definition of clases that | presented inthe previous chap. shall {hon ave an overview ofthe substemtive analysis of capitalist clas {elotions an disuse the historia eansformatfons uf thee based tnonsons. This wll bo followed by a detailed dscusion of thre i erent eatgories of postions withi hoo class ellions tha are gon traly designated "idle clos” Finally | shal preset some very rough estimates of the distibtion of people in the class structure of the contemporary United Sates ‘onders who ae not interested inthe nuances ofthe definitions of closes ll beable to Ret a goneral understanding ofthe categories ‘Ohick underlie the theoretical and empirical investigation of Income Uetermaton inthis research by rang the Mest to sections and Skipping the rest of the chapter CLASS AS A SOCIAL RELATION What does it really man to define clases in tems of capitalist soci rslations of production? One thing is very important to clarify from the outa classes defined i this way ore not tings They arnt toners soups or ttisieal agprapations finda nore {hey soci organizations Clas relations may give ie to lass organi= ations but classes por sore Hot organizations. esses constitte ammon postions withing special kindof contradictory socal rel tionship, socal eladons of production. There ar four impertant ports ats deltion clasts eonatute common positions, those positins are relational, those rootons are contradictory, and those contradic toy tlattons ae located within production Tsalldecuss each In Cases Constitute Positions ‘To say that classes constitute “positions” implios. to use Praoworski's(1977) apt expression, that there are “empty places” inthe Social structure which are filled by individuals. The analysis of class he in Advan ‘mus be understood asthe analysis primarily of such empty places, and ‘only secondarily of the actual individuals who fill the slots. While {uestions of social mobility are important ina class analysis, there in logleal prorty to undersianding the empty places into which ind iduals ate sorted, Poulantzas (18730) has emphosized thie points "the {question of who occupies a given postion, ue, whois, or bacomes 4 bourgeas, proletarian, petty bourgeois, poor peasant, eta hove and whe he does. is subordinate to the first aspect—the reproduction of the actual positions occupied by the social classes (pp. 49-50." This ‘low is quite In keoping with Marx's awn use, In the preface tothe 13867 German edition of Copital Mare wrote: "Individuals are dealt With only in so far as they are the personifications of economic categories, embodiments of particular elaserelations and. class Positions Exist Within Relations Classes are not, however, just any “amply places” in the socil structure which can bo ordered Ina Iierarchieal fashion, As we argued in the previous chapter, classes constitute common positions within Social ations of production. and this means tha lasers aust ays bo understood in tors of their relationship to other clases, ‘The notion of “positions within relations” is a complex one, On tho one hand, the rolationship itself fs definable only in terms of the Positions which are ina relation with each other on the other end, the Positions are determined by the relations of which they are elements It Is incoreect to se classes as postions which exist independently and only then entor into relations with other classes; but it Ix also incorrect to see those relations themselves es in any sense existing prior tothe classes which thoy determine. Classes are positions within relations ‘the analysis of the positions and relations must accu simultaneously. Relations Are Contradictory Classes within Marxist thoory are more than just postions within social relations: they are positions within contredictory social rl tions. To say that arelationship is contradictory implies that there is an Intrinsic antagonism between the elements (positions) determined by that relation. “Contradiction,” in this sanee, must be distinguished from “conflict.” To say that two groups are in conflict with each other Js simply to describe them as pursuing opposing objectives it snot to ako the theoretical claim that such an opposition Is an intrinsic pa ofthe very definition ofthe two groups. In contrast, when we say that tivo classes ae in contradictory relationship to each other. such op position is viewed as necessary consequence ofthe very relationship, fehich defines the classes. For example, the bourgeoisie and the pro Tetarlat are definable only fn terms of thoi relationship to each other the existence af one class presupposos the existence ofthe otter. and they ate thus necessary conditions foreach other. Duta the same ime, the relationship which determines the bourgeosie and the proletariat isa elation of exploitation and domination: the bourgeoisie exists only bocause its ia position to dominato and explolt the proletariat. Thus, the cles inforests defined by this class relation are fundamentally op posed to each other Tis in this senge that there is an intcinsic—as Spposed te purely contingent—contradiction between classes. “This claim about the intrinsically antagonistic character of class relations is ultimately » claim about cles struggle. aot simply the nas {ure of els structure, If manifest class behavior is fundamentally de termined by clas structure (a premise ofall olational views of class), nd ifthe class relations which define that class structure are Intinsi- tally contradictory. then class struggle iself becomes an intrinsle father than 8 contingent consequence of the structure of class vel tions." While the form and intensty of class struggles may vary Instovially-indood, much of the thoory of class davoted precisely to Understanding the dynamics of such vasation—the fat of clas sug> Sle fsa constant of class societies. ‘One very important consequence follows from the proposition that rot only clos contradictions bit class stugale are intrinsic to class cities: clase structures themselves ean mover be totally static. Class Sirageles ore not simply stugles between classos; they ae struggles ver clas relations. This implies thatthe lass structures themselves sre continually transformed by the very class struggles which they ‘Setermine.* Tris for this reason that Marxists insist that an analysis of class structure ust always be historical—not inthe sense tht is necessary esegndtSean sense bing ir we gale nine puso ls nae ‘ttn in cs rocure eal wht eiting pvetoel sactare always to return to the origins of elas relations, but in the sense thet le positions must be viewed as patt of a process in which class ‘feagale constantly reshapes the “empty places” which define those positions, and class positions constantly shape the terrain on which Eras stragate ls fought. tls through the bistriel investigation ofthis Sislecticel relationship betwen class structure and class struggle tht the bnsie Logi ofthe clas structure sll can be revealed Within Production ‘The final element in the Marxist definition of classes, as wos steessed In chapter 1, is tat the contradictory social relations which fletermine classes are located within the social organization of produc tion itself. “Production.” inthis context, must not be understcod na fowly as the production of phys commodities, but includes the Drodclion of services as well Given these four elements of the definition of class, the theoretical starting point of clase snalysis isto decode the historical transforma: tions of the social relations of production in order to uncover the class positions which these relations determine. This is no simple task, for those sola relations ore often hidden by the outseard Frm of capitalist Institutions Many theorists have mistaken form for substance and have thus completely mystified the nature of clas rlations in capitalist, fociely. Dahtendorf, for instane, confuses formal legal ttle to property—the outward appearance of class relations—with substantive foletions of production, and thus iiss thatthe formal separation of Tesel ownership from actual control in the modern corporation implies the demise of capitalist rolalions of production altogether. What we Imus ois go below the level of outward appearances inorder to dis= Cover the substantive processes that define class relations in capitalist scaly OUTLINE OF THE ARGUMENT Before plunging into the historical examination ofthe processes underlying clas relations tight be helpful to anticipate the conclu Sion ofthe analysts shall argue that capitalist social relations of prod: tuction can be broken down into thies interdependent dimensions oF processes 4. Social relations of eantral over money capital, i. contral over the flow of investments and the accumulation process, oe alleraativaly, sontzol over how much Is produced end what is produced, 2, Social relations of contro! over physieal capital, Le. contol ver tho use ofthe physical means of production, or control over how things are produce 5. Social rulations of authority, io. control aver supervision and iscipine within the labor peacess. © ‘Tn first of those soften referred to a8 “teal economic ownership: the second and tied are often grouped together under the rubric" posses- “The term control within each of these dimensions of socil rela- tions needs some explanation. As shall use the term, conteol does not primarily reer to an aspect ofthe relationship of people to things, but Father an aspect of the social relations among people. tn everyday len fuage. conto! implies a eapaeity to make some kind of decision, and thus capacity to dispose of rome Kind of esource. A social relation of control thus implies that this capacity is an atibute of a reat Individuals per se, in these terms, do nol “have” contol over money capital physical eapital, or labor: that contol is lodged in the social Felation ito sehich the individual enters. To say that “capitalists con trol the means of production, for example, isto tay thatthe social relationship between capital und labor stmultaneously confers on the apitalist position the capacity to dispose of the means of production tnd deprives the working-class position of that capacity In asnse, the focal relation between capital an labor defines a relationship between, these positions and things and thus between the Incumbents ofthese positions individuals} and things This distinction between “positions” and “individuals” cannot bo coveremphasizd, It becomes clearest when the actu decision-making process is lodged in a collectivity af postions, so that even in be- bhavioral terms individuals qua individuals are nat “making” deisions But even when a single capitalist makes all of the decisions about investments, use of physical capital, deployment of lebor, ete, the This leaking down cls relatos bn developed i dient ayy 2 rl apr ei ral Bl Poe 797 {G577) Rebelo te teres tent apach am rsa bared ‘ty father Mat ppronces to lar came und in Wright 078 _ hae in Advanced Capt Sects 2s control involved in such decisions must be understood as. an aspect of the social relation botwoon capital and labor and not simply a charac- teristic of the capitalist es an individual human being It is by virtue of being in a particular postion within this social roation and not by virtuo of being an “individual human being” that capitalists have this control. An individual who leaves a capitalist position within tho so- tial relations of production and becomes a worker lses the capacity 10 ‘ispose of the means of production (i, no longer has contcol over the ‘moans of production) ‘There is cler logical hierarchy among the three dimenstons of control, Coneol over investments sols Limite on the range of possible decisions over the use ofthe physical means of production, and contol ‘ver the physical means of production sels limits on control of actual Tabor within the labor process In effect its impossible to make sigalt icant decisions about investment hich do not have real effect on the contol of phystcal capital and labor, f only in the sense that future ‘options aver the allocation of phiysleal capital and labor become con ‘lalned. Its possible, on the other hand, to have control aver labor within the labor process whieh dors nat have significant direct effects ‘on averll investments, It is thus reasonable to Consider a rentier a= 8 member of the capitalist clas, wharess foreman should not b €luded in thet class. The former controls investments (even ifin a rather passive way: the llter only controls labor within the labor process “The fundamental class antagonism between workers an cap lalists ean be viewed as a polarization of each ofthese threo under Iying processes: capitalists contol the authority structure as a whole, dec how the physical means of production ate tobe used, and con trol the accumulation process. Workers, in contrast, ar exchided from control over authority relations, the physical means of production, and the Investment process, Whon tho capitalist system is analyand at the highest level of bstraction—tho level of the pure capitalist mode of production—these faze the only clas positions defined by capltalist relations of produce tion.* When we move ta the next lose level of abstraction —what is fenerally called the level of the "social formation’ —other class pos tions appear. First eal capitalist societies always contain subordinate telations of production other than those of the capitalist mode sat. In ok stato, nia, ator se tot a hc th na anal tevtalma interme of temas funomenat contacts ond dtr (ch Ia stn emo ems covert cottons and Scrminios particule, simple commodity production (ke, production organized {or the marke by independent, sel-employed producars who employ no workers) has always existed svithin capitalist societies, Within swe ple commodity production the potty bourgeoisie is defined as having, fscanomie ovinership and possession of the means of production, but having no control over labor power (sincano labor power i employed). Table 2.1 illustrates the relationship of the petly bourgeoisie tothe working class and the capitalist class in terms of the theo underlying Procosses of clas relations. ‘Secondly, the three processes that comprise capitalist socal rela tions of production do not always perfectly coincide. This fact Is the key to our understanding the class position ofthe social categories that fare Inbeled "middle class” (or more exactly “nev middle classes” to ‘istinguish them from the teditional potty bourgeoisie). Tho new mid dle clastes can be defined at social eatogories that occupy contradic tory locations within class rolations. Of courso, ina sono all class positions are contradictory ln that class eeations are intrinsically an- {agonistic social relations. The poiat is that corain “empty placss” in the clas stucture constitute doubly contradictory locations they rep- resent positions which are torn between the basic contradictory class tolations of capitalist society (That I, they represent postions which deviate from the “pure” patterns Mlusrated in Table 2.1] Rather than le such a cumbersome expression as “contradictory positions within the basic contradictory class relations of capitalist society." I shall for conventence simply refer to these positions as “contradictory cass lo- ‘ite notion f “anti rom tat witch preva epost ance a nce or model the ras of det of meine ie sifted inorder to dng ‘vith the hear seta imply eit thon opener. Astracton rts Taemerely sary, omnia conventions vad emt gations te >on, Swen iach) Doe ast 0) “setae ‘Throo clusters of such contradictory class locations are especialy Important (oe Figute 2.1. p. 42): 1 Managers and supervisors occupy @ contradictory location be- tween the bourgeoisie and the proletariat 2. Semioutonomous employees who retaa relatively high levels of control over thelr immediate labor process occupy «contradictory loc tion hetween the working class and the petty bourgeoisie, ‘3. Small employers occupy a contradictory lation between the bourgeoisie and the petty bourgeoisio* This conception ofthe middle cass will constitute our general cane sion. Now let us tun to the historical analysis which suggest that tha isa sonsible way to analyze the social word ‘THE PROCESSES OF CLASS RELATIONS [A number of critica historical transformations of eapitalist produc lion eam help us to unravel the various pracesses which underli the terwaen ha yty bourse or al empayars oma soe he eo ied nde pr hy ce class relations of advoncod capitalism. Since our basic purpose I not to Tavestignte the trnsformations of cage structures as such, and since toch of these transformations hasbeen thoroughly studied elsewhere, 1 ‘hall only briefly review them hor. “The Worker's Loss of Control aver the Labor Process “Tho saga ofthe progressive dispossesion ofthe dict producers in he course of epitalist development her boos tld many’ times. The pot hot need stressing hee is tha he los of eantal er the ibor Process is not an allornothing. phenomenon, but has. occured Erdually over a ong period of time end esate im varying degrees oven tbany. Inthe caries capitalist prodetion, the diet producers gon ty mind constable ton over the Ibo pro, en fapeetlly in cottage industries, they even ovned all or pact of their {mediate means of production. Ths, although economic conditions Ceuta acted oss powerful constraint on workers in cttage indus Ives nevertheless thoy often retained roltvely high levels af coneo ver the poco of ther labor he Tongth of tho working day, and other Styne he labor procass Such a situation mage it more dificult for Capitalists fo raise the rate of explototion, and this In urn acted 35a Serio constreint onthe accumulation proces fn erly capitalism (oe ‘Wright 178», pp. 170-71) Iho hr history of clots strate baton capitalists end work ers, expecially inthe nsteenh entry can be seen as true over the terms of the contol ofthe labor process. As Marlin (1574) has Sng, one of tho lor impulses for the eration of Fetries wos the {sie to undermino worker Conta Ata minim, factory owners had much greter conto over the lant of he worklng day ed generally ‘ver other aspects ofthe labor process a wal than capitalist i ho ting-ou system PimiGace workers vere gathered within factories, the assault on thelr remaining contol ofthe labor proces continued athe form of techn {al innovations which fragmented the prodvtion process and whch froeesivly “desl the labor force (see Broverman. 1974) Gopal could force workers to work n the fatory for tn hours by the clock, but a long asthe workor maintained rea autonomy it the Tabor process it was difiattfor the enptaliet toe sure of getting anywhere near ten hours of actual Inbor from the worker. The close "persion of the labor process Is mach easier when tasks are simple tnd routlnized and theis pace is determined by machinery rather thea the worker. Thus, capitalist look for innovations which tend to reduce Skill Tove and reduce the autonomy of workers on the jb. The cul. nation ofthis process was the mass-production assembly Hine regulated by the principles of Taylorism, im which the worker lost almost all autonomy and baceme vitually a human component of machinery i al Many Marxist accounts of transformations of the labor process leave the analysis at this point, concluding tht such trnsformatoas Ihave produced a monotonic trajectory of prletaieniztion. Such an account is incomplete: there are atleast three importent countertn ences to thie general process a) succossful resistance uf workers (0 less of control over work: (b) changes in technology which generate sew skills: c) changing conditions of accumulation and social organ Zation which may encourage a Yelaxation of stret contro within pro duction To say that control over the labor process is a dimension of class Felations is simultaneously to say that it isa dimeasion of clas strug fs, While the balance the interests ofthe bourgeoisie may dominate ‘elthin this steuggle, nevertheless there are innumerable cases where Fesistance has been successful Burawroy (1978) argues this point bail. Henly in ertique of Braverman. Taylorism itself, Hurewoy argues, 5 ‘good example this t did not universally increase the actual cape Tey af the capitalist class to control the labor of workers, fr thei collec tive resistance to Teylorism In many cases rediced the control of capi- tal, Burawoy suggests that it ns only through the technological changes mechanization) which accelerated nthe wake ofthe failure of ‘Taylorism that such esistance was weakened, Noble (2978), a care: {ul case study of tho mechine industy, argues that numeriea-contrl technologies were introduced consciously to reduan highly skilled ‘machinists’ contzol ofthe labor procos, but that workers wore able to subvert the elfactiveness ofthe technology to such an extent thatthe et effect was visually no degradation of work, The lbor process is thus an arona of struggle, not just of domination, and workers have often effectively prevented capital from undermining their control ‘within it The inttoduetion of neve technologies may eso genarate new sills nd new categories of bs in which the worker may have preter im mediate conttal over the labor process, Whole new industries have emerged in the past ew decades eg, computers} and al least some of ths jobs within those incusris requte considerable skill and invalve ‘ansiderable degrees of autonomy. The question is, then, shell thls “counteracting tendency” of the expansion of skilled positions ie systemically stronger or weaker than the tendency toward degradation. Lite systematic data are aval tle fo assess this issue. Anetdotaly, te cler thal within the sectors ff new technologies ther ae real peessutes to rodues the ski levels of the nowly created positions. When computers were fist boing de ‘eloped, the actual operators of computer hardware tended fo be n= incers. Gradually over the last twenty years this Job has boon “de- Skilled until at presen, computer operators are generally technicians With only one or two yeate of port-high-school training ‘Ansedotl evidence, however, can alorays be countered with op: posing anecdatal evidence, What is needed is systomatic data which ex {thine the proces of proletrianzation overtime. Two recent studies tne by Brawoing snd Singelmann (1976) and another by Wright and Singelmann (3078), provide some very provisional findings on these issues, Browning and Singelmaan decomposed the changes in the oc tupationel structure between 1960 and 1970 into three components: (a) hotcupational shift effect due to the changing occupational stucture ‘within industries, (on industry-sift effet due to the changing dis {eutlon ofthe poptlation aerase industries, and (e) an interaction- tect due to the interaction of changes in the occupational and indus iy structures. This procedure makes K possible to see whether the ‘eral grovith of professional occupation (as defined by conventional Census categories), for example, is due primally to an upgrading of obs within'» given industy (occupationshift effets) or to a more rapid expansion of those industries which employ eelatively more pro fessionals (industry-shift fects), ‘The rosulls were striking. The ‘occupationalsbift effect indicates an actual dectine in the proportion OF the labor force in pofesslonal occupations. Thetis Fall Industrial Sectors had graven atthe same rae, then there would hve been propor tionately fewer professionals in 1970 than 1960. All ofthe nt geowth of professlonals in the sociely as # whole, therefore, was due to the Industey shill effect (the intaraction-shiN effect was nogligible). and most of this industry-shift effect was due to the rapid expansion of social services during the petiod "The study by Wright and Singelmann has extended this analysis, ina very crude way, to examine changes in tho class structure. Ifa tough distintion is made between relatively autonomous and rela- tively nonautonomaus employees (on the basis of data from the 1908 Michigan Survey of Working Conditions). then it can be shown that within industil sectors there was a substantial decrease in autono- lanes in Advanced Capital Societies a ows nosan woo 1980 on 197, Aan lf ent nein autonomy inthe society as whole wae deta he selaiely moe Fpl expansion of hove Industial sectors with a flava Tos gro letarianized labor force, i Thase data suggest that the procs of degredation of labor, ofthe tedtn of worn autonomy a msonateononey wha Copitlsm that may to greater or less! extent be countarocod thanges in the industial strate. I the mos ecent prio ths hy Aw ll involved the rpid expansion of soil servic Ith fal ers ofthe sate ints future expansion of such serves hen his Counteracting tendency would self be expected to doing ‘The third counteracting tendency to an intensification of pro length gna nent dns of wre “human feltions” approaches tothe problem of worker productivity which tnve replace, a Test prt the principles of sc dselpling an tude scientific management asthe ideology of abor contol, One part fac approche therchent a ls snd the enlargoment ofthe sphere of declslon-making under the cone the worker. = : ae Such shifts inthe ideology of control have often been taken as indicating ansformation inthe rel relations of contol wth duction. Without suggesting that heve change a of to conseyuenee, Iotusnotethat generals theenlargedautonomy embodied in jobenti sent schemes fs confined within very narrow lite and suboroinated torthe imperatives of increasing workor productivity (eno Zimbullst, 10975. That i, contol Is rloxed—and generally peripheral contol that—only when tis more than compensated for by increased produ tion. Thus, in roporeto the Confernce Boat Rosh (197) wren "ie caret ph nob design ion aig internal mation om he ‘ployee shat be pets take wth mse ton aeons ‘eh pr ha dad nee a ome Greater worker control of the labor process, or what is often called ‘worker participation,” is one Important form of this redesigning of 2 lo Srna Iam Determination jubs to increase productivity. In a second Conference Roard report on worker pattieipation in management, Roach (1978) writes improving produciigr moan [snc ‘Sagem ns bah wind se odo eltvetnin The occasional trends toward increased seorer participation do not contadit the importance of contol ofthe labor process as «die tenon of clas rations ater they ovals underyng age Cg Tales to evtoct as much aca or out of th worker during the Iwork day as posble (this would baraly be denied by any capitals {Conte ever the labor process is basic means of accomplishing this. Unilercetinhisorcl conditlons or example, shen # large propor Mon ofthe dst wok ee a hey pretreat Seourgoosrtisans,peaants, et) ith ile experience of fctry s- Tinie und wthout proper work batt, tt and despotic contol of the labor process maybe the most effective structure of conto om the Eapitalo point of view. Under contemporary condition, a pata fnvation of diect contol nay sceomplish the sme end “The las of contol over the abo process by wokers is thus nota ous process of prolearaniation. Workers resst imple, homoge thelr ovn degradation, 2 times sucessfully technological change ands relatively autosomous job postions, even if that autonomy is Fimultaneowsly being eroded: end changing social conditions of ac: taumulation make possible less rigidly authoritarian forms of contol {within the labor process. Nevertheless for our prosent purposes al of these countertendeneies sll demonstrate our central point: the social relations of control aver the labor process constitute a basic dimension of class relations 'Na development In capitalist social rel often as “proof” that Marx's image of clas structure is outmoded than > an the so-called “separation of ownership and contro in the modern eoeporation. OF cours. no one can deny the considerable growth of ‘managerial hierarchies in the modern corporation and the yeneral de cline of the traditional family-owned firm in favor of the jlntstock company falthough, as Zeitlin [3474] frcaflly argues, there ae con siderable data to indicate thatthe proponents ofthe “managerial revo lution” have grossly exaggorated those changes). The issue is not whether professional managers playa bigger role in running corpora tions today than a hundred years ago.but how such postions shod be structurally interprted in terms ofa theory of elas relations ‘The apparent separation of ownership snd control in the large ‘corporation hides a whole sores of structural transformations and dil ferontiations. Two such transformations ate of particular importance here: the functional differentiation between economic ownership and possession, and the dissociation between legal and economic owner- Ship. the minateenth century, ll thre of these ware embodied in the entrepreneurial capitalist. AS capital became more concentrated and Contalizd, those thre dimensions of ownership tendo! ty become et least partially differentiated, Before wo proceed further, these terms nas to be defined some what more precisely. Legal ownership i simple enough, It constitutes the various forms of logal itl to property inthe means of production ‘The usual form of such ownership in advanced capitalism is stock ownership. Possession ise bit more complicated. It designates de Vrooy's words (1973). “the ability to put the meons of production to work. I thus pertains to the management of eapitalitfeeores[p.. [As we will see below, possession, in turn, ean be divided into authority "lations (control over labor power inthe labor procass| and control aver the actual physical means of production. Finally. economic ownership ls the most complax of all thre dimensions. Betelbelm (1975) defines itas “the power to assign the objects on which it bears especially the ‘means of production) to specifi uses and to dispose of the products obtained through these means of production (p. 8” Less abstractly this means contol over the flow of resources into production [i.e fnvestment and accumulation) “Marx was one of the fist writers to recognize the dual quality of the captalis as oth towne ond the manager of epi x Copal 1 wales Po hase Structure and Ime Dtrminton ‘itl property cpt aude th ration poe i ylding ines ‘inet tnd cela on poco ocr wich in prai a enter rte thought p78) Inoue the ener of apt rs othe enn of enon tronertp while the employer of eapita afer to that of possesion. Inthe cuts of capitalist development, this distinction between functions of capital bog to correspond to distinction between ac teal positions: Agni, Mars wets: Stuck emp in genral—deelped wih hv ca ys a resign oceans ro management 3 anti om the ‘Somkiptept iawn rerone Tht mane ha ‘ris, ptr al the el fnetin: ertains to the uncon capalst Sweeny tneonry eae tthe pat apse po fiom tam the production proces 987 a 87-8) “This pata separation of econotnle ownership from possession a consoquenes of the concentation en conrallationafexpital growing tut ofthe accumulation process (Concentration reas to the increas Ing absolute magaitude of epital unite: centration tothe Ine ing relative magnitude? icreming concentration and eenalation Ive encourage! the diferentition of econo oneahip abd pos Sesion orto reso fin, nd mow sya he sno of al ‘Swnership and production increse, becomes leas and las practic forthe nine Individuals fo be equally tnvalved fn both functions Competitive pressure wil end to push capital to ire professional tnanagers to deal with specific aspects of production and eventually to fel coordinate production as a whole, Secondly. thro has boon pened tendency inthe development of monopoly capitals for the SGncenttton and centalizaton of eeoromle owberahip to develop tnote rapidly than the concontaton and cantrlzation of possession. {oncenteaton and contalztion of possession ree to the absolute fn elative growth inthe sale of prodocton under unified manager “thf sro nso iin Mat ay hal et es aceon min te ens bow onset lr othe ‘fei tn op atin Ni ha mic oe shod tate ces, ta azaton et er Toth sive nt of wee at th hepa the i Catoush ths (Tod soe crue ent reat s a'p. hanes in Advanced opti Soi 3% ‘mont rather than simply under unified ownership] Poulantzas (1875)

You might also like