Class and Status: The Conceptual Distinction and its ... · of class and status as two qualitatively different forms of social stratification retained currency through the 1970s,

INTRODUCTION

Weberrsquos distinction between class and sta-tus (Weber [1922] 1968926ndash40) is com-

monplace in materials in introductory coursesand texts dealing with social stratificationSurprisingly however contemporary researchmakes little use of the distinction Moreover theconcepts of class and status are often applied inways that at least from a Weberian standpointappear unclear if not confusing

In the United States much of the refinementof Weberrsquos approach appears to already havebeen lost by the 1950s and 1960s as leadingauthors in effect reinterpreted class in terms of

status Thus one finds definitions of socialclasses on such lines as ldquostrata of society com-posed of individuals who accept each other asstatus equalsrdquo (Lipset and Bendix 1959275)or ldquoaggregate[s] of persons within a societypossessing approximately the same statusrdquo(Shils [1962] 1975249) Subsequently theessentially one-dimensional view of stratifica-tion implicit in such definitions was confirmedthrough the widespread acceptance of the notionof ldquosocioeconomicrdquo status which while littleexplicated allowed stratification to be treat-edmdashin a way highly convenient to quantitativeresearchersmdashin terms of a single continuousmeasure such as the Duncan Socio-EconomicIndex (Duncan 1961)

In Europe and especially in Britain the ideaof class and status as two qualitatively differentforms of social stratification retained currencythrough the 1970s following its effectivedeployment in Lockwoodrsquos influential TheBlackcoated Worker (1958) and also in variouscommunity studies Interest in status howeverthen rather rapidly declined In part this can beseen as a response to actual social changemdashthatis to the rather evident decay over the postwardecades of many features of the traditional sta-tus order The revival of academic Marxism anda consequent preoccupation on the part ofMarxist and non-Marxist sociologists alikewith issues of class was also important Morerecently the theoretical efforts of Bourdieu (see

Class and Status The ConceptualDistinction and its Empirical Relevance

Tak Wing Chan John H GoldthorpeUniversity of Oxford University of Oxford

In this article we return to Max Weberrsquos distinction between class and status as related

but different forms of social stratification We argue that this distinction is not only

conceptually cogent but empirically important as well Indeed class and status do have

distinct explanatory power when it comes to studying varying areas of social life

Consistent with Weberrsquos assertions we show that economic security and prospects are

stratified more by class than by status while the opposite is true for outcomes in the

domain of cultural consumption Within politics class rather than status predicts

Conservative versus Labour Party voting in British general elections and also LeftndashRight

political attitudes But it is status rather than class that predicts

LibertarianndashAuthoritarian attitudes

AMERICAN SOCIOLOGICAL REVIEW 2007 VOL 72 (August512ndash532)

Direct correspondence to Tak Wing ChanDepartment of Sociology University of OxfordManor Road Oxford OX1 3UQ UK (twchansociologyoxacuk) We thank Geoff Evans DavidLockwood Don Treiman five anonymous reviewersand the editors for helpful advice and comments onearlier drafts We are also grateful to Arts CouncilEngland for access to the detailed occupational codesof the Arts in England data set The views expressedin this article are entirely our own and not necessar-ily those of the Arts Council Our research is sup-ported by an ESRCAHRC research grant under theirCultures of Consumption Programme Phase II awardnumber RES-154-25-2006 Earlier versions of thisarticle were presented at the ISA RC28 meeting atUCLA August 2005 and at a seminar in Oxford

Delivered by Ingenta to University of Aberdeen

Tue 22 Jan 2008 150742

esp 1984) have attracted much attention par-ticularly his attempt to rethink and indeed over-come Weberrsquos opposition between class andstatus (1984xii) that is by treating status as thesymbolic aspect of class structure that is itselfdeemed to be not reducible to economic rela-tions alone (cf Weininger 2005)

In a previous article we sought to reassert theconceptual value of the distinction betweenclass and status and to argue on empiricalgrounds that in present-day British society atleast a status order is still discernible (Chan andGoldthorpe 2004) In the present article wehave two further complementary aims We seekto show again on empirical grounds how in dif-ferent areas of social life the stratification ofoutcomes whether seen as life-chances or aslife-choices may predominantly occur on thebasis of either class or status In this way wethen hope to clarify and reinforce the case fortreating class and status as different forms ofstratification that exert their effects throughquite distinct social processes or mechanismsBoth require recognition if a full understandingof the structuring of social inequality in con-temporary societies is to be obtained

CLASS AND STATUS

Taking a broadly Weberian position we regarda class structure as one formed by the social rela-tions of economic life or more specifically byrelations in labor markets and production unitsThus a primary level of differentiation of classpositions is that which sets apart employersself-employed workers and employees In mod-ern societies however further differentiation

must be recognized among employees in termsof their relations with employers as these areregulated by the (implicit as well as explicit)terms of their employment contracts In recentyears a fairly wide consensus has emerged atleast among sociologists engaging in compar-ative empirical research (eg Blossfeld Millsand Bernadi 2006 Breen 2004 Shavit Arumand Gamoran 2006 Shavit and Muumlller 1998)to treat class operationally on these lines on thebasis of the EGP or CASMIN class schema(Breen 2005 Erikson and Goldthorpe 1992Erikson Goldthorpe and Portocarero 1979)1

Table 1 shows the versions of the schemathat we use in this article The following pointsshould be noted First in the formation of theschema employment status (employer self-employed employee etc) and occupation aretaken as proxies for employment relations it issupposed that individuals with similar employ-ment status and occupation are likely to be sub-ject to similar forms of employment regulationand thus to have similar class positions Thetheoretical basis for this approach is set out byGoldthorpe (2007 vol II chap 5) drawing onthe analysis of problems of work-monitoring and

CLASS AND STATUSmdashndash513

Table 1 Versions of the Class Schema

National StatisticsOriginal Version Socioeconomic Classification

I Professionals and managers higher grade 1 Higher managers and professionalsII Professionals and managers lower grade 2 Lower managers and professionalsIIIa Routine nonmanual employees higher grade 3 Intermediate employeesIIIb Routine nonmanual employees lower gradeIVac Small employers and proprietors (including farmers) 4 Small employers and own-account workersIVb Self-employed workersV Technicians and supervisors of manual workers 5 Lower supervisors and techniciansVI Skilled manual workers 6 Semi-routine workersVII Nonskilled manual workers 7 Routine workers

Note NS occupational unit groups in combination with employment status codes are allocated to classes follow-ing algorithms available for the original version at httpwwwcfacukCAMSISDataBritain91html and forNSndashSEC at httpwwwstatisticsgovukmethods_qualityns_secdownloadsNS-SEC_User_2005pdf

1 An important alternative class schema is thatproposed from a Marxist standpoint by Wright It isbeyond the scope of this article to compare the twoclass schemata Suffice to note that Wright (199737)himself recognizes that although the two schematahave clearly differing theoretical origins ldquoas a prac-tical set of operational categories the [Wright] classstructure matrix || does not dramatically differ fromthe class typology used by Goldthorperdquo


Tue 22 Jan 2008 150742

of human asset specificity as found in the lit-erature of transaction-cost and personnel eco-nomics To test the validity of the schema as thusconstructed a good deal of empirical researchhas been undertakenmdashand with largely encour-aging resultsmdashboth with regards to the schemain its original version (eg BirkelundGoodman and Rose 1996 Evans 1992 Evansand Mills 1998 2000) and the more recent ver-sion represented by the National StatisticsSocio-Economic Classification (NSndashSEC)mdashaclassification that has since 2000 replaced theRegistrar Generalrsquos Social Classes in UK offi-cial statistics (Office for National Statistics2005 Rose and Pevalin 2003 Rose Pevalin andOrsquoReilly 2005)

Second while we regard the class structurethat the schema represents as being preemi-nently a structure of inequality we do not envis-age classes as always falling into a simplehierarchical ordering (cf Dahrendorf195974ndash77 Giddens 1973106) Individualsin different classes may be advantaged and dis-advantaged in differing and perhaps not entire-ly commensurable respects as a result of theemployment relations in which they areinvolved Thus while members of Classes I andII the professional and managerial salariat canbe seen as generally advantaged relative tomembers of other classes and members ofClasses VI and VII the wage-earning workingclass as being generally disadvantaged anysingle ranking of the intermediate classes ismore problematic Their employment relationsclearly differmdashcompare for example those ofa bank clerk (Class IIIa) a self-employed elec-trician (IVb) and a factory foreman (V)mdashbutin ways that give balances of advantage anddisadvantage (eg in regard to job securityearnings stability and prospects and promotionopportunities) that are not readily orderedConsequently our analyses treat the classes ofthe schema as is the usual practice as unorderedcategorical variables

Third the classes of the schema are notintended to capture ldquorealrdquo sociocultural group-ings in the sense of collectivities recognized byand subjectively meaningful to their membersand with well-defined social boundaries as cre-ated say by processes of selection socializationor closure In other words and still followingWeber we do not treat classes as ldquocommunitiesrdquo(ldquolsquoKlassenrsquo sind keine Gemeinschaftenrdquo) but

simply as existing insofar as ldquoa number of peo-ple have in common a specific causal compo-nent of their life-chancesrdquo (Weber [1922]1968930) that derives from their relations with-in labor markets and production unitsmdashwhichwe wish to add also importantly condition var-ious life-choices that they are typically requiredto make From this standpoint class effects arebrought about in two main ways On the onehand they result from events that impact onthose holding different class positions with sig-nificantly differing frequenciesmdashquite inde-pendently of their class awareness (egmembers of Class VII do not have to think ofthemselves as being working class to have a farhigher risk of unemployment than members ofClass I) And on the other hand such effectsresult simply from members of different class-es pursuing their particular interests and goalsin ways shaped by the particular patterns ofconstraint and opportunity by which their classsituations are characterizedmdashindependent ofany influence of class-specific values and normswhich may or may not exist (cf the distinctionbetween Massenhandeln and Gemeinschaft-shandeln made by Weber [1922] 1968930)2

Further following Weber ([1922]1968932ndash39) we regard a status order as astructure of relations of perceived and in somedegree accepted social superiority equalityand inferiority among individuals This doesnot reflect personal qualities but rather thedegree of ldquosocial honorrdquo attached to certain oftheir positional or perhaps purely ascribedattributes (eg birth or ethnicity)3 The socialhierarchy thus created is expressed in differen-tial association especially in more intimatekinds of sociabilitymdashWeber speaks of com-mensality and connubiummdashand in lifestyles ofdiffering distinction that are seen as appropri-ate to different status levels In status orders in

514mdashndashAMERICAN SOCIOLOGICAL REVIEW

2 Our understanding of class is therefore very dif-ferent from that of authors who argue for the ldquodeathof classrdquo (eg Kingston 2000 Pakulski and Waters1996)

3 This understanding of status should be recognizedas clearly different from that found in the current lit-erature on status generating and confirming interac-tional processes especially in microsocial contextswhere the emphasis clearly does fall on personalqualities and their perceived worth (eg Jasso 2001)


Tue 22 Jan 2008 150742

their most developed formsmdashsuch as werefound in early modern Europemdashthe demarca-tion of patterns of association and lifestyles isclear-cut and often institutionally grounded asfor example through sumptuary legislation Inmodern societies however the development ofideas of citizenship implying a fundamentalequality of legal and political rights (cfLockwood 1992173ndash78 Marshall 1950)means that the status order takes on an increas-ingly conventional character or in other wordsis for the most part maintained only informal-ly Moreover the egalitarian ideology of citi-zenship results in a greater reluctance on the partof those treated as social inferiors to respondwith deference and in claims to superioritybeing less often made at least in an explicit andpublic way (for Britain see McKibbin 1998Runciman 1997) Thus the hierarchy of statusrelations becomes less one of well-defined sta-tus groups than one of relatively loose social net-works and its expression is more implicit orcovert

Although there are good grounds for sup-posing that in present-day societies the strati-fying force of status has weakened it would berash to suppose that status can now be simplydisregarded Most obviously issues of statusare still widely recognized among the popula-tion at large When the topic of class is raisedin everyday conversations or in the media orwhen members of the public are asked aboutclass in interviews with sociologists it is infact status rather than class following the dis-tinction made above that is chieflymdashand quitereadilymdashdiscussed For example phrases suchas ldquoclass distinctionsrdquo ldquoclass barriersrdquo or ldquoclassconsciousnessrdquo are commonly used in waysthat make it apparent that they in fact refer todistinctions of status and to status exclusivenessand sensitivity4

We have already provided systematic evi-dence of the persistence of a status order inBritain (Chan and Goldthorpe 2004) Drawingdirectly on the work of Laumann (1966 1973)we take occupation to be one of the most salient

positional characteristics to which status attach-es in modern societies And we assume closefriendship implies a relation of basic equalitybetween individualsmdashthat is one into whichstatus differences are unlikely to intrude We usenational survey data to investigate the occupa-tional structure of close friendship and we takedissimilarity indices for the occupational dis-tributions of friends by occupational groupingas input to a multidimensional scaling analysisFrom this a leading dimension emerges onwhich occupational groupings are orderedaccording to the degree of similarity of theirfriendship patterns and which can we believebe most plausibly interpreted as representing sta-tus That is to say starting from the structuringof a relationship implying social equality astructure of inequality can be inferred5

Table 2 shows the 31 occupational group-ings used in the scaling exercise in the order inwhich they appear and with the scores that theytake on the dimension that we see as capturingstatus This hierarchy of occupations has clearcontinuities with that described for earlier peri-ods in historical and pioneering sociologicalresearch (Chan and Goldthorpe 2004) In gen-eral occupations that require working with sym-bols and perhaps people and especiallyprofessional occupations confer the higheststatus while those that require working direct-ly with material things confer the lowest statusAt a more detailed level managers employed ina more blue-collar milieu in industry or tradetend to rank lower than managers or indeedeven routine administrative employees whowork in an entirely white-collar milieu whileoccupations that require working with both peo-ple and thingsmdashsuch as many occupations in thenow expanding services sectormdashhave typical-ly intermediate rankings We report various


5 As noted above in addition to occupation sta-tus is also likely to be based on ascribed character-istics such as ethnicity Thus Laumann (1973) treatsethnoreligious affiliation as a further characteristicof primary importance for status in modern Americansociety However he shows that the status-conferringeffects of occupation are largely replicated withineach ethnoreligious grouping that he distinguishesand that no interaction effects occur We expect a sim-ilar situation to be found in the British case althoughwe do not so far have data available that would allowthe matter to be empirically investigated

4 Indeed ethnographic work (eg Deverson andLindsay 1975) even if of a rather unsystematic kindhas indicated that in the right context individualsmay still be quite ready to speak in ways that obvi-ously imply status superiority and derogation


Tue 22 Jan 2008 150742


Table 2 31 Status Groups Ranked in Descending Order of Status Score and RepresentativeOccupations Within Each Category

Rank

01

02

03

04

05

06

07

08

0910

1112

13

14

15

1617

18

19

20

21

2223

2425

26

27

Title

higher professionals

associate professionals in business

specialist managers

teachers and other professionals ineducation

general managers and administrators

associate professionals in industry

scientists engineers and technologists

filing and record clerks

managers and officials nec administrative officers and assistants

numerical clerks and cashiers associate professionals in health and

welfare secretaries and receptionists

other clerical workers

buyers and sales representatives

childcare workersmanagers and proprietors in services

plant depot and site managers

sales workers

health workers

personal service workers

protective service personnelroutine workers in services

catering workersstore and dispatch clerks

skilled and related manual workers nec

transport operatives

Code

HP

APB

SM

TPE

GMA

API

SET

FRC

OMOAOA

NCCAPH

SEC

OCW

BSR

CCWMPS

PDM

SW

HW

PSW

PSPRWS

CWSDC

SMO

TO

Representative Occupationsa

chartered accountants clergy medicalpractitioners solicitors

journalists investment analysts insurancebrokers designers

company treasurers financial managerscomputer systems managers personnelmanagers

college lecturers education officers andinspectors school teachers

bank and building society managers generalmanagers in industry national and localgovernment officers

computer analysts and programmers quantitysurveyors vocational and industrial trainers

civil and structural engineers clinicalbiochemists industrial chemists planningengineers software engineers

conveyancing clerks computer clerks libraryassistants

security managers cleaning managersclerical officers in national and local

governmentaccounts assistants bank clerkscommunity workers nurses occupational

therapists youth workerspersonal assistants receptionists secretaries

word processor operatorsgeneral assistants commercial and clerical

assistantsbuyers and purchasing officers technical sales

representatives wholesale representativeseducational assistants nursery nursescatering managers hoteliers publicans shop-

keepers and managersclerks of works farm managers maintenance

managers transport managers works man-agers

cash desk and check-out operators sales andshop assistants window dressers

ambulance staff dental nurses nursing auxil-iaries

caretakers and housekeepers hairdressers andbeauticians travel attendants undertakers

fire service and police officers security guardscar park attendants cleaners counter-hands

couriers and messengers hotel porterspostal workers

bar staff chefs cooks waiters and waitressesdispatch and production control clerks store-

keepers

gardeners and groundsmen printers textileworkers woodworkers

bus and coach drivers truck and van driverstaxi drivers

Score

05643

05337

05107

05017

04114

03116

03115

02559

0235502274

0223802228

01539

01443

01193

01097ndash0453

ndash0625

ndash1151

ndash2121

ndash2261

ndash2288ndash2974

ndash3261ndash3353

ndash4072

ndash4114

(Continued on next page)


Tue 22 Jan 2008 150742

other instances where occupations located ineconomically advantaged classes are of rela-tively low status and vice versa (Chan andGoldthorpe 2004391ndash92) These examples areimportant insomuch as they demonstrate thatalthough both the class schemata we use and ourstatus scale are occupationally-based constructs(with in the case of class other information onemployment status also being required) quitedistinct even if moderately correlated aspectsof occupations are in fact captured on the onehand the employment relations typicallyinvolved and on the other the degree of socialhonor typically conferred

Several further points are of relevance forour present purposes First while status asdetermined by an analysis of the occupationalstructure of close friendships is correlated withincome and education the correlation is quitemodest especially for income (Chan andGoldthorpe 2004) The status scale is clearly tap-ping something other than socioeconomic sta-tus insofar as this is determined by acombination of income and education6 It isalso distinct frommdashsupposedmdashscales of occu-pational prestige insofar as these simply reflectjudgments of job rewards and requirements (cfGoldthorpe and Hope 1974)

Second while a status gradient can be seenas running across classes (as represented by theschema of Table 1) from the higher profession-al and managerial salariat down to the non-

skilled working class there is still a good dealof variation in status homogeneity within class-es Some classes notably Class II the lowersalariat and Classes IVac and IVb smallemployers and the self-employed show rela-tively high internal stratification by status whileothers notably Class I the higher salariat andClass VI skilled manual workers are far morestatus homogenous (see Chan and Goldthorpe2004 fig 6)7

Third although we shall for conveniencerefer to the occupational categories of our scaleas status groups we would again as in the caseof classes not wish to imply that they are ldquorealrdquosociocultural entities or at all events not onesof a clearly bounded kind Rather as we alreadyindicated we see the social organization of sta-tus in modern societies as taking the form of rel-atively loose social networksmdashnetworks ofrelations often extensive in space among indi-viduals who come together as equals in moreintimate forms of sociability and who tend tovalue and seek to pursue broadly similarlifestyles with perhaps only quite limited nor-mative reinforcement from the expectations of


28

29

30

31

Note The four broad status bands that are for some purposes used in the analyses are indicated by dashed linesa Representative occupations refer to those that account for relatively large numbers of individuals within eachcategory and at the same time give some idea of its range

Table 2 (continued)

Rank Ttile Code Representative Occupationsa Score

skilled and related manual workers inconstruction and maintenance

skilled and related manual workers inmetal trades

plant and machine operatives

general laborers

SMC

SMM

PMO

GL

bricklayers electricians painters and decora-tors plasterers roofers telephone repairmen

fitters setters setter-operators sheet metalworkers turners welders

assemblers canners fillers and packers foodprocessors moulders and extruders routineinspectors and testers

agricultural workers factory laborers goodsporters refuse collectors

ndash5014

ndash5121

ndash5589

ndash5979

7 In the context of a comparative research projecton social stratification and cultural consumptioncolleagues from six other countries (Chile FranceHungary Israel the Netherlands and the UnitedStates) have followed the methodology describedabove in constructing status scales for their soci-eties although using for the most part data on mar-riage rather than friendship Preliminary analysesindicate a high degree of cross-national commonal-ity in the ordering of occupations on these scales

6 When the status scale is regressed on income andeducation the coefficient for income turns out to beinsignificant (Chan and Goldthorpe 2004 table 3)


Tue 22 Jan 2008 150742

significant others8 We thus treat status effectsin those areas of social life where they are appar-ent as resulting primarily from individualsrsquoefforts to maintain and possibly to enhancetheir status via various lifestyle choices andcommitments

This final point is of particular importance indifferentiating our position from that recentlytaken up by Grusky and his colleagues inresponse to claims of the decay or death of classin contemporary societies (eg Kingston 2000Pakulski and Waters 1996) If such claims areto be adequately countered it is argued classanalysis and it would seem the study of socialstratification more generally will need to beratcheted down to and regrounded at the levelof occupations (see eg Grusky 2005 Gruskyand Soslashrensen 1998 Grusky and Weeden 2001Weeden and Grusky 2005) In this way ldquorealrdquosociocultural entities may be more readily iden-tified arising from processes of occupationalselection socialization and regulation and itwill in turn be possible to show the full extentto which and range of mechanisms throughwhich experience at ldquothe point of productionrdquoimpacts on social attitudes and behavior Weourselves take occupation (along with employ-ment status) as a proxy for class position andagain as a prime characteristic to which statusattaches However we are concerned only withquite specificmdashand quite differentmdashaspects ofoccupations in thus seeking to make ourWeberian approach operational Indeed we donot consider it necessary to conceive of occu-pations classes or status groups as being ldquorealrdquoentities in Gruskyrsquos sense to show that individ-ualsrsquo class and status positions can and dohave very real consequences for their sociallives This becomes apparent in the analysesthat follow9

RESULTS

CLASS AND ECONOMIC LIFE-CHANCES

Given our understanding of class in terms ofemployment relations we expect individualsrsquoclass positions rather than their status to be themajor influence in determining their econom-ic life-chances Goldthorpe and McKnight(2006) present empirical results for Britainshowing strong connections between class andrisks of unemployment short-term variability inearnings and long-term earnings prospectsand they also spell out the underlying mecha-nisms operating through different forms ofemployment contracts It is not easy to envisagesimilar mechanisms that might operate in thecase of status Nonetheless we seek to test ourexpectations more strictly at least in regard tounemployment and earnings prospects by bring-ing status as well as class directly into the analy-sis10

With regard to unemployment risks we usethe same data set as Goldthorpe and McKnightthe British Household Panel Survey (BHPS) Wetake individuals ages 21 to 54 in 1991 who wereinterviewed in all years between 1991 and 2002(N = 2860) Over this 12-year period 826respondents (289 percent) reported at least onespell of unemployment and 299 (105 percent)reported a cumulative unemployment durationof 12 months or more To avoid cases of mere-ly transitional or even planned unemploymentbetween jobs we concentrate on these latterindividuals who experienced what could be


detailed occupational classification with which theywork are to be ordered and according to what crite-ria When attitudinal and behavioral differences arerevealed across occupational groups (cf Weeden andGrusky 2005) these are likely to result only in partfrom their stratificationmdashthat is from structuredsocial inequalities in some form or othermdashand in partalso from purely ldquohorizontalrdquo differentiation (soci-ologists and accountants say may have similar classand status positions but still very different occupa-tional subcultures) It seems essential to specify somebasis on which these two kinds of effect are to be dis-tinguished (see Goldthorpe 2007 vol II chap 5)

10 We cannot include analyses of variability inearnings since no data set is available that containsinformation on this matter and also occupationaldata of a kind that would allow us to implement ourstatus scale

8 While Weber ([1922] 1968932) sees statusgroups (Staumlnde) as tending unlike classes to formas communities (Gemeinschaften) he still adds thatthese are often ldquoof an amorphous kindrdquo especially inldquomodern democraciesrdquo and he speaks also of statusldquocirclesrdquo (Kreise)

9 While we see the research program envisaged andembarked on by Grusky and his colleagues as hav-ing the potential to revitalize the sociology of occu-pations its relevance to research in the field of socialstratification must be difficult to assess until theproblem is resolved of how the categories of the


Tue 22 Jan 2008 150742

regarded as long-term (or recurrent) unem-ployment

Table 3 reports results from binomial logis-tic regression analyses in which experience (ornot) of long-term unemployment is the depend-ent variable In Model 1 class is included as anexplanatory variable along with a number ofsociodemographic variables that for presentpurposes we treat as controls11 Class effectsshow up quite consistently and on essentially thesame pattern as found by Goldthorpe andMcKnight (2006 fig 2) The most striking fea-ture is the much greater risk of long-term unem-ployment run by members of Class IVbself-employed workers and of Classes VI andVII the working class than by members ofClass I and Class II the salariat At the extreme

an unskilled worker in Class VII is four times(e1398) more likely to have been long-termunemployed than a higher-level professional ormanager in Class I As noted by Goldthorpeand McKnight the form of regulation ofemploymentmdashthe service relationshipmdashthat istypically enjoyed by members of the salariat ismore likely than the basic labor contract toimply an expectation of continuity of employ-ment or at least of employability and furthershould job loss occur usually involves a muchlonger period of notice during which alternativeemployment can be sought (see also Gallie et al1998)

Model 2 repeats the analysis with status asmeasured by the scale presented in Table 2included Two points of main importanceemerge First while class effects are in mostcases lowered and those for Class VI and IIIbbecome marginally insignificant at the 5 percentlevel their pattern is little changed In particu-lar the far more serious risk of long-term unem-


Table 3 Determinants of Long-Term Unemployment (logistic regression N = 2860)

Model 1 Model 2

0 SE 0 SE

Age 34ndash43a 145 (150) 152 (150)Age 44ndash51 154 (175) 155 (176)Femaleb ndash710 (145) ndash674 (148)Cohabitc 636 (221) 633 (221)Single 668 (215) 670 (215)WidDivSep 1097 (193) 1092 (193)Childrend 102 (150) 095 (150)Class IIe 339 (265) 267 (273)Class IIIa 763 (279) 645 (298)Class IIIb 859 (335) 664 (375)Class IVac ndash073 (557) ndash311 (595)Class IVb 1193 (303) 897 (398)Class V 622 (309) 317 (407)Class VI 1145 (291) 800 (416)Class VII 1398 (242) 1062 (379)Status ndash413 (359)Constant ndash2986 (265) ndash2836 (295)

Pseudo R2 065 066Log-Likelihood ndash89538 ndash89472

a Age 21ndash33 as reference categoryb Male as reference categoryc Married as reference categoryd Childless as reference categorye Class I as reference category p lt 05 p lt 01 (two-tailed tests)

11 Descriptive statistics for all variables includedare available from the authors on request


Tue 22 Jan 2008 150742

ployment for unskilled workers remains almostthree times greater (e1062) than for those inClass I Second although introducing statusreduces class effects somewhat the effect ofstatus itself is far from significant In account-ing for this more detailed analyses show thatthere is a particularly weak association betweenstatus and the risk of long-term unemploymentacross the higher ranges of our status scalewhere the risk tends to be well below average12

Turning now to economic prospects wefocus as do Goldthorpe and McKnight on age-earnings curves We cannot however followthem in using the data set of the New EarningsSurvey because of difficulties in applying ourstatus scale We therefore continue with theBHPS data set despite problems that arise withrelatively small numbers Goldthorpe andMcKnight (2006 f igs 4 and 6) show thatmarked differences exist in the economicprospects of members of different classes asindexed by age-earnings curves Most notablywhile the earnings of those in the salariat tendto rise with age up to around age 50 reflectingit may be supposed the operation of incremen-tal salary scales and promotion ladders forthose in the working class earnings tend moreor less to level out in their thirties13

Figure 1 Panel A presents age-earningscurves based on BHPS data for men andwomen who were employed full-time in 2002and were in Class I Class II or a combined blue-

collar Class V+VI+VII14 These curves bring outessentially the same class differences as thoseobserved by Goldthorpe and McKnight Theother panels of Figure 1 present curves for broadstatus bands within these three classes using thefour major divisions that we proposed within ourstatus scale (see Table 2)15 It can be seen (PanelsB and C) that within both Class I and Class IIthe higher and lower salariat the curves for sta-tus band 1 lie somewhat above those for statusband 2 but are still very similar in shape (allow-ing for some fluctuation in band 2 probably dueto small numbers) and further (Panel D) thatwithin Class V+VI+VII both status bands 3 and4 show the same much flatter curves with thecurve of status band 4 actually lying above thatof status group 3mdashreflecting the fact that themanual occupations that predominate in band 4yield generally higher earnings than the personalservice or people-processing occupations thatpredominate in band 3

In sum thinking in terms of status as well asclass does not appear to add a great deal to ourunderstanding of differences in age-earningscurves To check this impression more formal-ly Table 4 shows results from analyses basedon the same data used in Figure 1 in which weregress earnings on age and age-squared

It is evident from the first panel of Table 4that as would be expected the coefficients forboth age terms are significantly larger forClasses I and II than for Class V+VI+VII Butthe second panel shows that while the coeffi-cients for status band 1 are larger than those forstatus bands 2 3 and 4 there is far less differ-


12 Since class and status are correlated if class isdropped from Model 2 status becomes statisticallysignificant and negative suggesting that low statusrespondents face higher risks of long-term unem-ployment We regard this though as a misspecifiedmodel A graph plotting the risk of long-term unem-ployment by status group is available from the authorson request In another model not reported here weadded educational qualifications in the regressionThis again leads to marginal reductions in classeffects but the effects of different levels of qualifi-cation are not themselves statistically significant

13 Neither Goldthorpe and McKnightrsquos analyses norour own claim to trace the actual life-course earningsof individuals but rather to show how age-specificearnings differ by class Some distortion is possibledue both to cohort effects and to selection effects (seeGoldthorpe and McKnight 2006) but not we believeof a kind sufficient to disturb the main results of theanalyses

14 We combine Classes V VI and VII and pool menand women together since we wish to consider earn-ings within combinations of age class and broad sta-tus band and for such analyses the sample size ofeven the BHPS is relatively small We do not considerage-earnings curves for Class III that of routine non-manual workers because it could in this case be mis-leading to treat men and women together AsGoldthorpe and McKnight (2006) show age-earningscurves show marked differences by gender and thisproblem could only be addressed by making a furtherIIIaIIIb division

15 For purposes of comparison with classes we usethe four broad status bands defined within our scale(see Table 2) However in all other analyses statusis treated as a continuous variable using the statusscore indicated in that table


Tue 22 Jan 2008 150742


Figure 1 Median Annual Earnings by Age for Men and Women in Full-Time Employment byClass (Panel A) and by Broad Status Band Within class (Panels B to D)

Table 4 Parameter Estimates and Standard Errors of Age and Age-Squared in OLS RegressionModels Predicting the Logarithm of Annual Earnings (Men and Women Combined)

Age Age Squared

0 SE 0 SE N0

Class I 116 011 ndash127 013 954Class II 111 011 ndash124 013 1255Class V+VI+VII 055 007 ndash063 009 1600

Status Band 1 145 01 ndash163 013 1317Status Band 2 056 009 ndash058 011 1540Status Band 3 088 011 ndash101 014 786Status Band 4 051 009 ndash057 011 1155

Class I amp Status Band 1 137 012 ndash152 015 735Class I amp Status Band 2 054 023 ndash053 026 218

Class II amp Status Band 1 147 017 ndash168 022 577Class II amp Status Band 2 070 013 ndash074 017 585Class II amp Status Band 3 251 058 ndash299 076 76

Class V+VI+VII amp Status Band 3 063 013 ndash073 015 497Class V+VI+VII amp Status Band 4 045 009 ndash052 011 1090

Note The regression models also control for the logarithm of hours worked and gender

inco

me

inco

me

ageage

inco

me

inco

me

ageage


Tue 22 Jan 2008 150742

entiation among the latter16 Similarly theremaining panels of the table show thatalthough within both Class I and Class II ageeffects on earnings are clearly stronger for thosein status band 1 than for those in lower statusbands in Class II the coefficients for statusbands 2 and 3 go in the ldquowrongrdquo directionmdashamajor factor being the inclusion in band 3 ofprotective service personnelmdashand in ClassV+VI+VII the coefficients for status bands 3and 4 are not significantly different

Insofar then as risk of long-term unem-ployment and age-earnings curves serve well asindicators of economic life-chancesmdashas goodindicators of security and prospects respec-tivelymdashwe can say that our expectation thatclass rather than status will mainly differentiatesuch life-chances is in general confirmed Weturn next to a quite different topic that of cul-tural consumption considered as an aspect oflifestyle In this case and following from ourearlier discussion our expectation is the exactreverse of the above we expect that stratifica-tion here will be on the basis of status rather thanof class

STATUS AND CULTURAL CONSUMPTION

For Weber lifestyle is the most typical waythrough which members of different statusgroups even within the purely conventionaland relatively loose status orders of modernsocieties seek to define their boundariesmdashthatis to establish cues or markers of inclusion andexclusion Furthermore a number of morerecent authors emphasize cultural taste and con-sumption as an aspect of lifestyle that is of par-ticular importance as a means of the symboliccommunication of ldquodistinctionrdquo and thus ofexpressing a form of hierarchy that is set apartfrom that of mere economic advantage (egBourdieu 1984 DiMaggio 1987 Peterson1997)

In a series of published and forthcomingpapers (Chan and Goldthorpe 2005 2007b2007c) we examine the social stratification ofcultural consumption in three different

domainsmdashmusic theater dance and cinemaand the visual artsmdashusing data from the Arts inEngland Survey of 2001 (Skelton et al 2002)17

For each of these domains we first applied latentclass analysis to raw data on the frequency ofdifferent kinds of consumption for men andwomen ages 20 to 64 to establish patterns ofconsumption and in turn types of consumer Wethen used multinomial logistic regression analy-ses to examine the determinants of individualsrsquoapproximation to one type or another

An initial finding from these analyses is thatcultural consumption in England is not stratifiedalong elite-to-mass lines in particular we can-not identify an elite that is distinctive in con-suming ldquohighrdquo cultural forms while at the sametime rejecting ldquolowerrdquo or more popular formsThere is in other words no evidence of a closehomology between cultural and social stratifi-cation We do find support albeit with somequalifications for the alternative hypothesisthat the main axis of cultural stratification inmodern societies at least is one that separatescultural ldquoomnivoresrdquo from cultural ldquounivoresrdquo(see Peterson and Kern 1996 Peterson andSimkus 1992) The former have relatively highlevels of consumption of all genres within aparticular domain but the latter are restricted intheir consumption to popular genres only18 Forour present purposes therefore the questionof chief importance is that of the basis on whichomnivorendashunivore stratification occurs

Table 5 shows selected results from our arti-cles previously referred to specifically from ourlogistic regressions in which the dependent vari-able is type of cultural consumer We presentthose results that relate to the major contrasts ofinterest that is the effects of covariates on the


17 Chan and Goldthorpe (2007a) which examinesthe social stratification of newspaper readership isalso relevant

18 For example in the case of music omnivoreshave similarly high levels of consumption of popand rock as do univores but unlike the latter also con-sume classical music opera and jazz In the case oftheater dance and cinema omnivores have rela-tively high attendance at plays musicals pantomimesballet and other dance performances as well as thecinema while univores are essentially restricted to thelatter With the visual arts however as noted in thetext below the most extreme contrast is that betweenomnivores and nonconsumers

16 For example the 95 percent confidence inter-val of the linear age term for status band 2 is 038 ndash074 (ie 056 plusmn 196 009) which overlaps withthat for status band 3 066 ndash 110


Tue 22 Jan 2008 150742

log odds of being an omnivore rather than eithera univore or as with the visual arts a type thatcan only be described as a nonconsumer orinactive19

Note first from Table 5 that the range ofsociodemographic variables that we introducechiefly as controls have significant effects inonly rather patchy albeit fairly plausible waysTurning then to the stratification variables onwhich our interest centers one can see that theeffects of class on the chances of an individualbeing an omnivore rather than a univore or aninactive are largely insignificant across each of


Table 5 Determinants of Types of Consumer in the Domains of Music Theater Dance andCinema and the Visual Arts (multinomial logit model N = 3819)

Music Theater Dance and Cinema Visual Arts

Omnivore Omnivore Omnivoreversus Univore versus Univore versus Inactive

0 SE 0 SE 0 SE

Femalea 156 (137) 615 (092) 223 (192)Marriedb ndash321 (176) 148 (112) ndash200 (239)Separated ndash065 (214) 188 (139) 180 (295)Age 066 (006) 005 (004) 026 (009)Child (0ndash4)c ndash391 (214) ndash562 (113) ndash639 (285)Child (5ndash10) ndash340 (188) 070 (100) 260 (232)Child (11ndash15) ndash397 (191) 088 (105) 039 (252)The Northd ndash470 (193) ndash231 (124) ndash089 (253)Midlands ndash198 (184) ndash207 (123) ndash880 (279)South East 060 (198) 083 (135) ndash150 (270)South West ndash224 (238) ndash189 (153) ndash174 (321)Income 012 (007) 026 (005) 006 (009)CSEOtherse 1006 (276) 169 (152) 1220 (499)O-Levels 1109 (242) 668 (128) 1072 (462)A-Levels 1523 (265) 1130 (145) 1849 (471)Subdegree 1851 (266) 1027 (160) 2219 (469)Degree 2367 (256) 1223 (151) 3260 (450)Class IIf ndash135 (172) 078 (126) 613 (241)Class III ndash329 (247) ndash161 (160) ndash396 (376)Class IV 299 (291) ndash205 (203) 699 (411)Class V ndash253 (382) ndash134 (218) 073 (554)Class VI ndash107 (317) ndash199 (195) ndash480 (514)Class VII ndash109 (387) ndash507 (230) ndash325 (646)Status 1047 (287) 631 (179) 1229 (402)Constant ndash5906 (472) ndash2118 (292) ndash5461 (688)

a Male as reference categoryb Single as reference categoryc Childless as reference categoryd London as reference categorye No qualification as reference categoryf Class I as reference category p lt 05 p lt 01 (two-tailed tests)

19 In the case of the visual arts we also need tomodify the omnivorendashunivore distinction to allowfor a type of consumer we label as a ldquopaucivorerdquo whoconsumes across a modest range of genres and in thecase of music to allow for a type of omnivorendashlis-tener who consumes most genres via various mediabut not in live form Supplementary tables for Tables5 to 7 are provided in the Online Supplement on the

ASR Web site httpwww2asanetorgjournalsasr2007toc058html


Tue 22 Jan 2008 150742

the three cultural domains The effects of statusin contrast are significant for each and quitestrong20 It is true that we here treat class on thebasis of the seven-class NS-SEC rather thanthe original nine-class version of the schema thatwe use in analyzing the risk of unemploymentHowever while we measure class in a somewhatless detailed way than previously we think itunlikely that this in itself could be the source ofthe clear predominance of status over class thatwe find21

In addition Table 5 shows that while incomehas a significant effect in only one domain (the-ater dance and cinema) level of educationalqualifications is generally significant in itseffects even if not always in an entirely mono-tonic way But the question then arises of howfar given that class income and status are alsoincluded in our analyses education is to beregarded as itself operating as a stratificationvariable It seems reasonable to suppose thatlevel of qualifications is to some extent at leastpicking up individual psychological attributessuch as information processing capacity thatcould exert a quite independent influence on thelikelihood of being a cultural omnivore (cfBerlyne 1974 Moles 1971 see also Chan andGoldthorpe 2007a)

Regardless of what view may be taken on thislast issue Table 5 still provides clear support forour expectation that differences in lifestyle willbe associated withmdashand can indeed be taken asexpressions ofmdashstratification by status ratherthan by class22 By way of illustration we can

turn to some of the descriptive detail of ourresults For example while members of ClassesI and II the professional and managerial salari-at are more likely to be cultural omnivores thanare members of other classes the importance ofstatus stratification within these classes is muchin evidence We have previously observed thatin the higher ranges of the status scale profes-sionals generally rank above managersCorrespondingly we find that the groups thatmost regularly show the highest proportions ofomnivores are higher professionals teachersand other professionals in education and spe-cialist managers for example finance IT andpersonnel managers who often have profes-sional qualifications and operate in professionalroles23 Conversely other types of managers inmanufacturing transport construction or serv-ices whose status rankings are similar to oreven below those of some groups of routinenonmanual workers in Class III have only a sim-ilar or if anything a lower probability of beingomnivores

So far then we have sought to bring out thecontrast between the stratification of econom-ic life-chances and the stratification of cultur-al consumptionmdashthe former primarily reflectingindividualsrsquo positions within the class struc-ture understood in terms of employment rela-tions the latter their position within the statusorder We do not wish to suggest however thatit is possible for all areas of social life to be sim-ply divided into those in which either class orstatus is the dominant stratifying influenceOften the situation may be more complex Wenow seek to illustrate this point by focusingspecifically on individualsrsquo political commit-ments and value orientations

CLASS STATUS AND POLITICS

A relationship between class and political partysupport has long been recognized Indeed sev-eral authors view the development of electoralpolitics in modern societies as ldquothe democrat-


dropped from the model Details available from theauthors on request

23 As seen in Table 2 these groups rank 1 3 and4 in the order of our status scale The second-rank-ing group associate professionals in business alsotends to show relatively high proportions of omni-vores

20 For example in the case of theater dance andcinema the probability of a hypothetical womanwho is 40 years old childless lives in London andearns pound25000 being an omnivore rather than a uni-vore is 23 to 26 percentage points higher (dependingon her educational attainment) if she is at the toprather than at the bottom of status hierarchy Fordetails see Chan and Goldthorpe (2005206)

21 NS-SEC could in fact be regarded as instanti-ating the conceptual approach of the schema in amore reliable way than previously Moreover as wenote in the articles previously cited even if we sim-plify our measure of status to the four broad statusbands that we introduced above in our analyses ofage-wage curves this still does not remove the clos-er association of cultural consumption with statusthan with class as treated by the seven NS-SECclasses

22 This remains the case even if education is


Tue 22 Jan 2008 150742

ic translationrdquo of the class struggle (Korpi 1983Lipset 1960) Of late though it has been wide-ly argued that class politics are in decline andsome authors claim a growing importance forwhat has been variously termed identitylifestyle or status politics (eg Beck 1992Giddens 1994 Hechter 2004) In Britain theassociation between class and vote did in factweaken at the General Election of 1997 (Evans1999) and has since remained at a lower levelthan previously This is not to say however thatclass is no longer a major influence on votingnor that it has become overshadowed by otherinfluences such as status (Brooks Nieuwbeertaand Manza 2006)

To investigate this matter further we turn tothe data set of the British Election Surveys and

to obtain an adequate basis for the kind of mul-tivariate analysis we wish to undertake poolthe data for the 1997 and 2001 elections InTable 6 we take party choice between theConservatives Labour and all other parties asthe dependent variable in a multinomial regres-sion that includes similar explanatory and con-trol variables as we use in our previous analyses

Under Model 1 it is clear that at all eventsin the case of the major contrastmdashvotingConservative rather than Labourmdashclass is themost important explanatory variable (using theoriginal version of the class schema) and on apattern that is familiar from all earlier researchThe higher salariat of Class I the small employ-ers of Class IVac and the self-employed work-ers of Class IVb are the most likely to vote


Table 6 Determinants of Party Choice at the 1997 and 2001 British General Elections (multi-nomial logit model N = 3407)

Model 1 Model 2

Conservatives Others Conservatives Others versus Labour versus Labour versus Labour versus Labour

0 SE 0 SE 0 SE 0 SE

2001a ndash220 (101) 259 (095) ndash220 (101) 258 (095)Age 030 (003) 009 (003) 030 (003) 009 (003)Femaleb 250 (101) 016 (098) 230 (103) 008 (099)CSEc 126 (152) 281 (151) 126 (152) 281 (151)O-Levels 468 (151) 490 (157) 465 (151) 489 (157)A-Levels 288 (179) 797 (175) 280 (179) 794 (175)Subdegree 381 (168) 1154 (166) 363 (169) 1146 (167)Degree ndash313 (179) 977 (169) ndash347 (183) 961 (172)Income 032 (004) 006 (004) 032 (004) 006 (004)Class IId ndash159 (151) ndash036 (152) ndash129 (154) ndash023 (155)Class IIIa ndash313 (182) 108 (183) ndash259 (190) 132 (192)Class IIIb ndash600 (216) ndash001 (211) ndash499 (240) 045 (236)Class IVac 655 (265) 235 (310) 760 (287) 283 (329)Class IVb 248 (232) 079 (260) 395 (278) 147 (305)Class V ndash1046 (231) ndash043 (206) ndash879 (289) 034 (272)Class VI ndash1126 (222) ndash466 (215) ndash944 (291) ndash382 (289)Class VII ndash1013 (182) ndash222 (178) ndash838 (257) ndash141 (257)Status 243 (252) 111 (256)Constant ndash2651 (289) ndash1828 (281) ndash2709 (296) ndash1857 (289)


Note Others = all other partiesa 1997 as reference categoryb Male as reference categoryc No qualification as reference categoryd Class I as reference category p lt 05 p lt 01 (two-tailed tests)


Tue 22 Jan 2008 150742

Conservative rather than Labour and unskilledworkers in Class VII are the least likelymdashinfact only about a third (endash1013) as likely asmembers of Class I Further the probability ofsupporting the Conservatives rather than Labourrises with income

Level of educational qualifications also hassignificant effects on voting and in the contrastbetween voting ldquootherrdquo rather than Labour edu-cation could be regarded as more influentialthan class It should be noted however that inneither contrast are the effects of educationmonotonic Thus in the case of theConservativendashLabour contrast those with O-levels or subdegree qualifications are most like-ly to vote Conservative For this reason therole of education specifically as an indicator ofstratification again seems somewhat problem-atic

Introducing status into the analysis (Model 2)we see that in most cases class effects are some-what reduced (though the effects of incomeremain unchanged) However while in the con-trast between other parties and Labour no classeffect is now signif icantmdashonly educationeffects in much the same way as beforemdashin theConservativendashLabour contrast class effects aswell as those of education remain very muchon their previous pattern Unskilled workers inClass VII are still less than half (endash838) as like-ly to vote Conservative rather than Labour as arethe higher salariat of Class I Furthermore theeffect of status itself is clearly not significant ineither contrast On this evidence then we cansay that class has certainly not disappeared asa basis of the stratification of political parti-sanship in Britain And at least so far as the cen-tral division within electoral politics isconcerned class remains of obvious importanceeven when the effects of status are also takeninto account24

From the general theoretical position that wehave adopted the results in Table 6 should be inno way surprising Individuals holding differentclass positions as we wish to understand themin terms of employment relations can quiterationally see themselves as having differentinterests (eg on economic inequality and the

redistribution of income and wealth levels ofpublic spending on social welfare and relationsbetween employers and employees) These inter-ests are likely to be better represented andupheld by different partiesmdashfor those in thesalariat and the petty bourgeoisie by theConservatives and for those in other classesparticularly the working class by Labour25

However in addition to these more or lessstandard leftndashright issues it may be supposedthat various other issues also have resonanceamong the electorate and perhaps to an increas-ing extent Issues relating to social order and tothe limits of freedom and authoritymdashissuesconcerning tradition and respect compliancewith the law and its enforcement and censor-shipmdashattract much attention today Likert-typescales with good reliability are now availablethat allow for the measurement of individualsrsquovalue positions on the leftndashright dimension andalso on what might be called thelibertarianndashauthoritarian dimension (EvansHeath and Lalljee 1996 Heath Jowell andCurtice 1994) We would then ask if in theplacing of individuals on the left-right scaleclass maintains its importance relative to statusin the same way as with voting and if sowhether a similar situation obtains in regard tothe libertarian-authoritarian scale

To address this question we draw on the dataset of the British Social Attitudes Survey of2002 that allows respondents to be given scoreson both the leftndashright and libertarianndashauthori-tarian scales (see the Appendix for the surveyitems used in the construction of these scalesand also Park and Surridge [2003]) Table 7shows the results of regressing individualsrsquoscores on these two scales on a similar range ofexplanatory and control variables as those weuse in regard to voting

In the case of the leftndashright scale the resultsin Table 7 are in most respects similar to thechances of voting Conservative rather thanLabour Class effects are generally significantoften strong and on essentially the same patternas beforemdashdue allowance being made for thefact that we have reverted to the seven NS-SECclasses and class effects are as it were rein-forced by income effects The effects of educa-


25 On the rationality of class voting see Evans(1993) and Weakliem and Heath (1994)

24 If education is dropped from the model classeffects become slightly stronger while the statusparameter remains nonsignificant for both contrasts


Tue 22 Jan 2008 150742

tion are also on the same pattern as for votingthat is those with intermediate-level qualifica-tions tend to be more right-wing than eitherthose with lower or higher qualifications Againtoo the effect of status fails to reach signifi-cancemdashalthough only marginally so and withthe sign of the coefficient indicating a tenden-cy for higher status to be associated with a moreright-wing orientation

In the case of the libertarianndashauthoritarianscale however an entirely different patternemerges All class effects and likewise those ofincome are now far from significant while theeffect of status is both significant and quitestrong The higher a personrsquos status the morelikely the person is to express libertarian ratherthan authoritarian values Specifically otherthings being equal an increase of one standarddeviation in status is associated with a changeof ndash51 (ndash1381 366) on thelibertarianndashauthoritarian scale Education also

shows some positive libertarian effects althoughonly for those with A-level qualifications orhigher and by far the strongest effect occurswith graduates26


Table 7 Determinants of Political Attitudes on LeftndashRight and LibertarianndashAuthoritarian Scales(OLS regression)

LeftndashRight LibertarianndashAuthoritarian

0 SE 0 SE

Age 005 (005) 030 (005)Femalea 605 (166) 136 (169)10ndash23kb 257 (214) 297 (217)23ndash44k 958 (233) 209 (237)gt 44k 2153 (277) ndash111 (282)CSEc 472 (276) 193 (281)O-Levels 1039 (261) ndash261 (263)A-Levels 1090 (294) ndash726 (299)Subdegree 1089 (289) ndash822 (292)Degree 153 (321) ndash3223 (325)Class IId ndash873 (282) 020 (288)Class III ndash1233 (359) ndash004 (367)Class IV 021 (429) ndash138 (438)Class V ndash1553 (434) 082 (443)Class VI ndash1551 (406) ndash340 (416)Class VII ndash1732 (453) ndash130 (462)Status 684 (377) ndash1381 (385)Constant 11711 (529) 21363 (537)

N 2421 2441R2 13 17

a Male as reference categoryb Income lt 10k as reference categoryc No qualification as reference categoryd Class I as reference category p lt 05 p lt 01 (two-tailed tests)

26 Park and Surridge (2003) in their analysis of thesame data set find stronger and more consistenteffects of education in regard to libertarianndashauthor-itarian values This suggests that neglecting status asthey do may lead to an overestimation of the impor-tance of education If education is dropped from themodel the status parameter still fails to reach statis-tical significance at the conventional 5 percent levelfor the leftndashright scale and while some patchy classeffects appear for the libertarianndashauthoritarian scalethe estimate for status almost doubles in magnitudeDetails are available from the authors on requestPark and Surridge also include a measure of religiousadherence and find that this too has significanteffects We repeated our own analysis with religionincluded and obtained similar results to Park and


Tue 22 Jan 2008 150742

A more detailed examination of our datareveals that the far more important effect ofstatus on libertarianndashauthoritarian than onleftndashright value orientations is a phenomenonchiefly of the higher ranges of the status orderFigure 2 shows that at lower status levelsmdashthatis in status bands 3 and 4 (marked in Figure 2by lsquo+rsquo and lsquorsquo respectively) which comprisemainly routine service and manual occupationalgroupsmdashleft-wing and authoritarian values tendto go together rather closely (cf Lipset 1960)But as status increases authoritarian valuestend to give way to more libertarian ones andleftndashright differences become apparent Thusamong groups in status band 2 (marked by lsquoorsquoin Figure 2) who tend to hold middling posi-tions on the libertarianndashauthoritarian scale onefinds both groups that are quite right-wingsuch as plant depot and site managers andmanagers and proprietors in services and thosewho are more centrist such as buyers and salesrepresentatives and several lower administra-tive and clerical groups A similar divisionoccurs in status band 1 (marked by lsquorsquo in Figure2) among groups who are alike in having themost libertarian values Thus general managersand administrators specialist managers andhigher professionals are clearly more right-wingthan are teachers and other professionals ineducation scientists engineers and technolo-gists and the three groups of associate profes-sionals

This indicates the importance of the class-sta-tus distinction in allowing a new perspective onthe social stratification of political partisanshipand of value orientations27 Class politics areclearly not dead Class can still be regarded asthe main basis of social cleavage so far asleftndashright issues are concerned that is issues

that turn on divergent interests arising out ofinequalities in economic conditions and life-chances But in regard to libertarianndashauthori-tarian issues it is status not class that appearsas the major stratifying force Adherence to lib-ertarian values we suggest tends to be a fea-ture of a high-status lifestyle and generalWeltanschauungmdashin just the same way as isrelatively high and omnivorous (rather than elit-ist) cultural consumption In this regard weagree with Hechter (2004404) that status pol-itics are likely to bring together ldquoindividualswho have a common interest in consuming cul-turally specific goods and who are attributedwith a specific degree of social honour on thisaccountrdquo28 It is still important however to rec-ognize that status politics do not replace classpolitics but rather coexist with them in a com-plex interrelationship29


Surridge but without the pattern of our other resultsbeing appreciably affected

27 If the class dummies are dropped from the mod-els in Tables 3 and 6 or from the regression predict-ing leftndashright values in Table 7 the status parameterbecomes statistically significant Likewise if thestatus parameter is dropped from Table 5 or from themodel predicting libertarianndashauthoritarian scale inTable 7 then some class parameters become signif-icant This is not surprising because class and statusare correlated However we would regard these asmisspecified models

28 This line of argument could be subjected to amore stringent empirical test if we had a data setthat included information on both individualsrsquoculturalconsumption and their sociopolitical values as wellas class and status But so far we have not been ableto find such a data set for Britain For the UnitedStates DiMaggio (1996) reports that art-museumvisitors tend to be more liberal tolerant and open tothe values of other lifestyles and cultures than non-visitorsmdashbut no more likely to have voted for Clintonthan for Bush in the 1992 presidential election andno less likely to be supportive of capitalist econom-ic institutions This points to the relevance of distin-guishing as we have suggested between orientationson the leftndashright dimension differentiated by classrather than by status and orientations on the liber-tarianndashauthoritarian dimension differentiated by sta-tus rather than by class Unfortunately the data setused by DiMaggio contains information only on edu-cation and income

29 The various associations reported in this articlebetween cultural consumption politics and attitudeson the one hand and status or class on the other mightarise partly as a result of self-selection That is peo-ple with certain unobserved dispositions mightchoose to take up particular occupations and there-by acquire a specific status (and class) At the sametime such dispositions might be associated with par-ticular lifestyles or attitudes Having said that classand status might also shape attitudes and other lifechoices (Kohn 1977) Both causal and self-selectionprocesses could be at work


Tue 22 Jan 2008 150742

In the case of politics we have sought toshow that both class and status are involved inthe shaping of partisanship and value orienta-tions but in clearly differing ways Class isdominant with regard to leftndashright issues thatinvolve primarily material interests Status how-ever prevails when it comes tolibertarianndashauthoritarian issues that involve touse Weberrsquos phrase ideal interests We believethere may be many other situations of this kindthat rather than calling for analyses primarilyin terms of class or status will require a care-ful separating out of the influence that eachexerts30

As an illustration a rather fierce debate hasrecently broken out over social inequalities inhealth between one the one hand epidemiolo-gists and medical sociologists favoring materi-al or political economy explanations of suchinequalities and on the other those favoringpsychosocial or cultural-behavioral explana-tionsmdashto follow the terminology used inBartleyrsquos (2004) valuable review of the debateThere are obvious parallels here between class-based and status-based explanations and itseems essential that to advance the debate thestratification of different health outcomes shouldbe investigated on the basis of reliable measuresof both class and status rather than of just oneor the other or of ad hoc measures that confusethe two

By extending the substantive range of analy-ses of the kind that we have presented in this arti-cle a more comprehensive mapping should bepossible of the relative importance of stratifi-cation by class and by status across differentareas of social life In turn a wider basis maybe provided for a more detailed investigationthan has so far proved possible of the range ofcausal processes or mechanisms through whichthe effects of class and status on individual livesare actually brought about


Figure 2 Mean Position of the 31 StatusGroups on LeftndashRight andLibertarianndashAuthoritarian Scales

Note Groups in status bands 1 2 3 and 4 are markedby the following symbols respectively + Table2 provides a key to the status groups abbreviations Thedashed lines refer to the mean sample values of the twoscales

CONCLUSIONS

In this article we have aimed to show that recenttendencies to disregard or to seek to elide theWeberian distinction between class and statusare unfortunate Indeed where the distinction isnot recognized and empirically implementedresearch into the stratification of British socie-tymdashand we believe of other modern soci-etiesmdashmay well go astray This point seemsespecially relevant with regard to studies of theimpact of social stratification on life-chancesand life-choices across different areas of sociallife As shown in our analyses individualsrsquoposi-tions within the class structure understood interms of employment relations have a prevail-ing influence on economic life-chances as indi-cated by risks of long-term unemployment andearnings prospects while it is their positionswithin the status order that are of key importancein regard to at least one major aspect of lifestylelevel and pattern of cultural consumption Ifthe study of these outcomes had been under-taken on the basis of some unidimensional con-ception of stratification a less clear appreciationof their structuring would in all probability haveresulted

30 Because in the foregoing analyses we focus onstatus as attaching to occupation it is quite possiblethat status effects on both cultural consumption andpolitics are underestimated For example these effectsmight be somewhat greater if status as attaching toethnicity were also taken into account

liber

taria

nndashau

thor

itaria

n sc

ale

leftndashright scale


Tue 22 Jan 2008 150742

APPENDIX

THE LEFTndashRIGHT AND

LIBERTARIANndashAUTHORITARIAN SCALES

The two scales used in the section ldquoClass Statusand Politicsrdquo are constructed additively fromthe following survey items Each item has fiveresponse categories ranging from 1 (agreestrongly) to 5 (disagree strongly) The values ofthe leftndashright scale range from 5 to 25 withhigher values denoting more right-wing viewsand the values of the libertarianndashauthoritarianscale range from 6 to 30 with higher valuesdenoting more authoritarian views Cronbachrsquosalpha for the two scales is quite high at 82and 74 respectively

LeftndashRight scale

bull Government should redistribute income from thebetter-off to those who are less well off

bull Big business benefits owners at the expense ofworkers

bull Ordinary working people do not get their fair shareof the nationrsquos wealth

bull There is one law for the rich and one for the poorbull Management will always try to get the better ofemployees if it gets the chance

LibertarianndashAuthoritarian scale

bull Young people today donrsquot have enough respect fortraditional British values

bull People who break the law should be given stiffer sen-tences

bull For some crimes the death penalty is the mostappropriate sentence

bull Schools should teach children to obey authoritybull The law should always be obeyed even if a partic-ular law is wrong

bull Censorship of films and magazines is necessary touphold moral standards

Tak Wing Chan is a university lecturer in Sociologyat the University of Oxford where he is also a fellowand tutor of New College His research interests aresocial stratification and mobility and the life courseespecially family formation and dissolution

John H Goldthorpe is an Emeritus Fellow of NuffieldCollege Oxford and a Fellow of the British AcademyHis main area of research is social stratificationHis two current research projects are concerned withproblems of education-based meritocracy and withsocial status and cultural consumption He is alsointerested in the closer integration of sociologicalresearch and theory The second two-volume editionof his book On Sociology is published by StanfordUniversity Press

REFERENCES

Bartley Mel 2004 Health Inequality AnIntroduction to Theories Concepts and MethodsCambridge UK Polity

Beck Ulrich 1992 Risk Society Towards a NewModernity London UK Sage

Berlyne Daniel ed 1974 Studies in the NewExperimental Aesthetics Steps Toward anObjective Psychology of Aesthetic AppreciationWashington DC Hemisphere

Birkelund Gunn E Leo A Goodman and DavidRose 1996 ldquoThe Latent Structure of JobCharacteristics of Men and Womenrdquo AmericanJournal of Sociology 10280ndash113

Blossfeld Hans-Peter Melinda Mills and FabrizioBernardi eds 2006 Globalization Uncertaintyand Menrsquos Careers An International ComparisonCheltenham UK Elgar

Bourdieu Pierre 1984 Distinction A Social Critiqueof the Judgement of Taste London UK Routledgeamp Kegan Paul

Breen Richard ed 2004 Social Mobility in EuropeOxford UK Oxford University Press

mdashmdashmdash 2005 ldquoFoundations of Class Analysis in theWeberian Traditionrdquo Pp 31ndash50 in Approaches toClass Analysis edited by E O Wright CambridgeUK Cambridge University Press

Brooks Clem Paul Nieuwbeerta and Jeff Manza2006 ldquoCleavage-Based Voting Behavior in Cross-National Perspective Evidence from Six PostwarDemocraciesrdquo Social Science Research3588ndash128

Chan Tak Wing and John H Goldthorpe 2004 ldquoIsThere a Status Order in Contemporary BritishSociety Evidence from the Occupational Structureof Friendshiprdquo European Sociological Review20383ndash401

mdashmdashmdash 2005 ldquoThe Social Stratification of TheatreDance and Cinema Attendancerdquo Cultural Trends14193ndash212

mdashmdashmdash 2007a ldquoSocial Status and NewspaperReadershiprdquo American Journal of Sociology1121095ndash1134

mdashmdashmdash 2007b ldquoSocial Stratification and CulturalConsumption Music in Englandrdquo EuropeanSociological Review 231ndash19

mdashmdashmdash 2007c ldquoSocial Stratification and CulturalConsumption The Visual Arts in EnglandrdquoPoetics In press

Dahrendorf Ralf 1959 Class and Class Conflict inIndustrial Societies London UK Routledge ampKegan Paul

Deverson Jane and Katharine Lindsay 1975 Voicesfrom the Middle Classes A Study of Families inTwo London Suburbs London UK Hutchinson

DiMaggio Paul 1987 ldquoClassif ication in ArtrdquoAmerican Sociological Review 52440ndash55

mdashmdashmdash 1996 ldquoAre Art-Museum Visitors Different



Tue 22 Jan 2008 150742

from Other People The Relationship BetweenAttendance and Social and Political Attitudes in theUnited Statesrdquo Poetics 24161ndash80

Duncan Otis Dudley 1961 ldquoA Socioeconomic Indexfor all Occupationsrdquo Pp 109ndash38 in Occupationsand Social Status edited by A Reiss New YorkFree Press

Erikson Robert and John H Goldthorpe 1992 TheConstant Flux A Study of Class Mobility inIndustrial Societies Oxford UK Clarendon Press

Erikson Robert John H Goldthorpe and LuciennePortocarero 1979 ldquoIntergenerational ClassMobility in Three Western European CountriesrdquoBritish Journal of Sociology 30415ndash41

Evans Geoff 1992 ldquoTesting the Validity of theGoldthorpe Class Schemardquo European SociologicalReview 8211ndash32

mdashmdashmdash 1993 ldquoClass Prospects and the Life-CycleExplaining the Association Between Class Positionand Political Preferencesrdquo Acta Sociologica36263ndash76

mdashmdashmdash ed 1999 The End of Class Politics ClassVoting in Comparative Context Oxford UKOxford University Press

Evans Geoff Anthony Heath and Manusur Lalljee1996 ldquoMeasuring LeftndashRight andLibertarianndashAuthoritarian Values in the BritishElectoraterdquo British Journal of Sociology4793ndash112

Evans Geoff and Colin Mills 1998 ldquoIdentifyingClass Structure A Latent Class Analysis of theCriterion-Related and Construct Validity of theGoldthorpe Class Schemardquo European SociologicalReview 1487ndash106

mdashmdashmdash 2000 ldquoIn Search of the Wage-LabourerService Contract New Evidence on theValidity of the Goldthorpe Class Schemardquo BritishJournal of Sociology 51641ndash61

Gallie Duncan Michael White Yuan Cheng andMark Tomlinson 1998 Restructuring theEmployment Relationship Oxford UK OxfordUniversity Press

Giddens Anthony 1973 The Class Structure of theAdvanced Societies London UK Hutchinson

mdashmdashmdash 1994 Beyond Left and Right The Futureof Radical Politics Cambridge UK Polity

Goldthorpe John H 2007 On Sociology Vol 2 2ded Stanford CA Stanford University Press

Goldthorpe John H and Keith Hope 1974 TheSocial Grading of Occupations A New Approachand Scale Oxford UK Clarendon Press

Goldthorpe John H and Abigail McKnight 2006ldquoThe Economic Basis of Social Classrdquo Pp 109ndash36in Mobility and Inequality Frontiers of Researchfrom Sociology and Economics edited by SMorgan D B Grusky and G S Fields StanfordCA Stanford University Press

Grusky David B 2005 ldquoFoundations of a Neo-Durkheimian Class Analysisrdquo Pp 51ndash81 in

Approaches to Class Analysis edited by E OWright Cambridge UK Cambridge UniversityPress

Grusky David B and Jesper B Soslashrensen 1998ldquoCan Class Analysis be Salvagedrdquo AmericanJournal of Sociology 1031187ndash1234

Grusky David B and Kim A Weeden 2001ldquoDecomposition Without Death A ResearchAgenda for the New Class Analysisrdquo ActaSociologica 44203ndash18

Heath Anthony Roger Jowell and John Curticeeds 1994 Labourrsquos Last Chance The 1992Election and Beyond Aldershot UK Dartmouth

Hechter Michael 2004 ldquoFrom Class to CulturerdquoAmerican Journal of Sociology 110400ndash45

Jasso Guillermina 2001 ldquoStudying Status AnIntegrated Frameworkrdquo American SociologicalReview 6696ndash124

Kingston Paul W 2000 The Classless SocietyStanford CA Stanford University Press

Kohn Melvin L 1977 Class and Conformity AStudy in Values 2d ed Chicago IL The Universityof Chicago Press

Korpi Walter 1983 The Democratic Class StruggleLondon UK Routledge amp Kegan Paul

Laumann Edward O 1966 Prestige and Associationin an Urban Community Indianapolis IN Bobbs-Merrill

mdashmdashmdash 1973 Bonds of Pluralism New YorkWiley

Lipset Seymour M 1960 Political Man The SocialBases of Politics London UK Heinemann

Lipset Seymour M and Reinhard Bendix 1959Social Mobility in Industrial Society London UKHeinemann

Lockwood David 1958 The Blackcoated Worker AStudy in Class Consciousness London UK Allenand Unwin

mdashmdashmdash 1992 Solidarity and Schism The Problemof Disorder in Durkheimian and Marxist SociologyOxford UK Clarendon Press

Marshall Thomas H 1950 Citizenship and SocialClass And Other Essays Cambridge UKCambridge University Press

McKibbin Ross 1998 Classes and CulturesEngland 1918ndash1951 Oxford UK OxfordUniversity Press

Moles Abraham A 1971 Sociodynamique de laCulture Paris Mouton

Office for National Statistics 2005 The NationalStatistics Socio-Economic Classification UserManual Basingstoke UK Palgrave Macmillan

Pakulski Jan and Malcolm Waters 1996 The Deathof Class London UK Sage

Park Alison and Paula Surridge 2003 ldquoChartingChange in British Valuesrdquo Pp 131ndash59 in BritishSocial Attitudes The 20th Report edited by APark J Curtice K Thomson L Jarvis and CBromley London UK Sage



Tue 22 Jan 2008 150742

Peterson Richard A 1997 ldquoThe Rise and Fall ofHighbrow Snobbery as a Status Markerrdquo Poetics2575ndash92

Peterson Richard A and Roger M Kern 1996ldquoChanging Highbrow Taste From Snob toOmnivorerdquo American Sociological Review61900ndash07

Peterson Richard A and Albert Simkus 1992 ldquoHowMusical Tastes Mark Occupational Status GroupsrdquoPp 152ndash86 in Cultivating Differences SymbolicBoundaries and the Making of Inequality editedby M Lamont and M Fournier Chicago ILUniversity of Chicago Press

Rose David and David J Pevalin eds 2003 AResearcherrsquos Guide to the National StatisticsSocio-economic Classification London UK Sage

Rose David David J Pevalin and Karen OrsquoReilly2005 The NS-SEC Origins Development andUse Basingstoke UK Palgrave Macmillan

Runciman Walter G 1997 A Treatise on SocialTheory Vol 3 Applied Social Theory CambridgeUK Cambridge University Press

Shavit Yossi Richard Arum and Adam Gamoraneds 2006 Expansion Differentiation andInequality of Access in Higher Education AComparative Study Stanford CA StanfordUniversity Press

Shavit Yossi and Walter Muumlller eds 1998 FromSchool to Work A Comparative Study of

Educational Qualifications and OccupationalDestinations Oxford UK Clarendon Press

Shils Edward [1962] 1975 ldquoClassrdquo Pp 249ndash55 inCenter and Periphery Essays in MacrosociologyChicago IL University of Chicago Press

Skelton Adrienne Ann Bridgwood KathrynDuckworth Lucy Hutton Clare Fenn ClaireCreaser and Adrian Babbidge 2002 Arts inEngland Attendance Participation and Attitudesin 2001 Research report 27 London UK The ArtsCouncil of England

Weakliem David L and Anthony F Heath 1994ldquoRational Choice and Class Votingrdquo Rationalityand Society 6243ndash70

Weber Max [1922] 1968 Economy and SocietyBerkeley and Los Angeles CA University ofCalifornia Press

Weeden Kim A and David B Grusky 2005 ldquoTheCase for a New Class Maprdquo American Journal ofSociology 111141ndash212

Weininger Elliot B 2005 ldquoPierre Bourdieu on SocialClass and Symbolic Violencerdquo Pp 116ndash65 inApproaches to Class Analysis edited by E OWright Cambridge UK Cambridge UniversityPress

Wright Erik Olin 1997 Class Counts ComparativeStudies in Class Analysis Cambridge UKCambridge University Press



Tue 22 Jan 2008 150742

esp 1984) have attracted much attention par-ticularly his attempt to rethink and indeed over-come Weberrsquos opposition between class andstatus (1984xii) that is by treating status as thesymbolic aspect of class structure that is itselfdeemed to be not reducible to economic rela-tions alone (cf Weininger 2005)

In a previous article we sought to reassert theconceptual value of the distinction betweenclass and status and to argue on empiricalgrounds that in present-day British society atleast a status order is still discernible (Chan andGoldthorpe 2004) In the present article wehave two further complementary aims We seekto show again on empirical grounds how in dif-ferent areas of social life the stratification ofoutcomes whether seen as life-chances or aslife-choices may predominantly occur on thebasis of either class or status In this way wethen hope to clarify and reinforce the case fortreating class and status as different forms ofstratification that exert their effects throughquite distinct social processes or mechanismsBoth require recognition if a full understandingof the structuring of social inequality in con-temporary societies is to be obtained

CLASS AND STATUS

Taking a broadly Weberian position we regarda class structure as one formed by the social rela-tions of economic life or more specifically byrelations in labor markets and production unitsThus a primary level of differentiation of classpositions is that which sets apart employersself-employed workers and employees In mod-ern societies however further differentiation

must be recognized among employees in termsof their relations with employers as these areregulated by the (implicit as well as explicit)terms of their employment contracts In recentyears a fairly wide consensus has emerged atleast among sociologists engaging in compar-ative empirical research (eg Blossfeld Millsand Bernadi 2006 Breen 2004 Shavit Arumand Gamoran 2006 Shavit and Muumlller 1998)to treat class operationally on these lines on thebasis of the EGP or CASMIN class schema(Breen 2005 Erikson and Goldthorpe 1992Erikson Goldthorpe and Portocarero 1979)1

Table 1 shows the versions of the schemathat we use in this article The following pointsshould be noted First in the formation of theschema employment status (employer self-employed employee etc) and occupation aretaken as proxies for employment relations it issupposed that individuals with similar employ-ment status and occupation are likely to be sub-ject to similar forms of employment regulationand thus to have similar class positions Thetheoretical basis for this approach is set out byGoldthorpe (2007 vol II chap 5) drawing onthe analysis of problems of work-monitoring and


Table 1 Versions of the Class Schema

National StatisticsOriginal Version Socioeconomic Classification

I Professionals and managers higher grade 1 Higher managers and professionalsII Professionals and managers lower grade 2 Lower managers and professionalsIIIa Routine nonmanual employees higher grade 3 Intermediate employeesIIIb Routine nonmanual employees lower gradeIVac Small employers and proprietors (including farmers) 4 Small employers and own-account workersIVb Self-employed workersV Technicians and supervisors of manual workers 5 Lower supervisors and techniciansVI Skilled manual workers 6 Semi-routine workersVII Nonskilled manual workers 7 Routine workers

Note NS occupational unit groups in combination with employment status codes are allocated to classes follow-ing algorithms available for the original version at httpwwwcfacukCAMSISDataBritain91html and forNSndashSEC at httpwwwstatisticsgovukmethods_qualityns_secdownloadsNS-SEC_User_2005pdf

1 An important alternative class schema is thatproposed from a Marxist standpoint by Wright It isbeyond the scope of this article to compare the twoclass schemata Suffice to note that Wright (199737)himself recognizes that although the two schematahave clearly differing theoretical origins ldquoas a prac-tical set of operational categories the [Wright] classstructure matrix || does not dramatically differ fromthe class typology used by Goldthorperdquo


Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742



Rank

01

02

03

04

05

06

07

08

0910

1112

13

14

15

1617

18

19

20

21

2223

2425

26

27

Title



specialist managers













sales workers

health workers






Code

HP

APB

SM

TPE

GMA

API

SET

FRC

OMOAOA

NCCAPH

SEC

OCW

BSR

CCWMPS

PDM

SW

HW

PSW

PSPRWS

CWSDC

SMO

TO
























keepers



Score

05643

05337

05107

05017

04114

03116

03115

02559

0235502274

0223802228

01539

01443

01193

01097ndash0453

ndash0625

ndash1151

ndash2121

ndash2261

ndash2288ndash2974

ndash3261ndash3353

ndash4072

ndash4114



Tue 22 Jan 2008 150742







28

29

30

31


Table 2 (continued)





general laborers

SMC

SMM

PMO

GL





ndash5014

ndash5121

ndash5589

ndash5979




Tue 22 Jan 2008 150742



RESULTS










Tue 22 Jan 2008 150742







Model 1 Model 2

0 SE 0 SE






Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742




Age Age Squared

0 SE 0 SE N0







inco

me

inco

me

ageage

inco

me

inco

me

ageage


Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742



Rank

01

02

03

04

05

06

07

08

0910

1112

13

14

15

1617

18

19

20

21

2223

2425

26

27

Title



specialist managers













sales workers

health workers






Code

HP

APB

SM

TPE

GMA

API

SET

FRC

OMOAOA

NCCAPH

SEC

OCW

BSR

CCWMPS

PDM

SW

HW

PSW

PSPRWS

CWSDC

SMO

TO
























keepers



Score

05643

05337

05107

05017

04114

03116

03115

02559

0235502274

0223802228

01539

01443

01193

01097ndash0453

ndash0625

ndash1151

ndash2121

ndash2261

ndash2288ndash2974

ndash3261ndash3353

ndash4072

ndash4114



Tue 22 Jan 2008 150742







28

29

30

31


Table 2 (continued)





general laborers

SMC

SMM

PMO

GL





ndash5014

ndash5121

ndash5589

ndash5979




Tue 22 Jan 2008 150742



RESULTS










Tue 22 Jan 2008 150742







Model 1 Model 2

0 SE 0 SE






Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742




Age Age Squared

0 SE 0 SE N0







inco

me

inco

me

ageage

inco

me

inco

me

ageage


Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742



Rank

01

02

03

04

05

06

07

08

0910

1112

13

14

15

1617

18

19

20

21

2223

2425

26

27

Title



specialist managers













sales workers

health workers






Code

HP

APB

SM

TPE

GMA

API

SET

FRC

OMOAOA

NCCAPH

SEC

OCW

BSR

CCWMPS

PDM

SW

HW

PSW

PSPRWS

CWSDC

SMO

TO
























keepers



Score

05643

05337

05107

05017

04114

03116

03115

02559

0235502274

0223802228

01539

01443

01193

01097ndash0453

ndash0625

ndash1151

ndash2121

ndash2261

ndash2288ndash2974

ndash3261ndash3353

ndash4072

ndash4114



Tue 22 Jan 2008 150742







28

29

30

31


Table 2 (continued)





general laborers

SMC

SMM

PMO

GL





ndash5014

ndash5121

ndash5589

ndash5979




Tue 22 Jan 2008 150742



RESULTS










Tue 22 Jan 2008 150742







Model 1 Model 2

0 SE 0 SE






Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742




Age Age Squared

0 SE 0 SE N0







inco

me

inco

me

ageage

inco

me

inco

me

ageage


Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742



Rank

01

02

03

04

05

06

07

08

0910

1112

13

14

15

1617

18

19

20

21

2223

2425

26

27

Title



specialist managers













sales workers

health workers






Code

HP

APB

SM

TPE

GMA

API

SET

FRC

OMOAOA

NCCAPH

SEC

OCW

BSR

CCWMPS

PDM

SW

HW

PSW

PSPRWS

CWSDC

SMO

TO
























keepers



Score

05643

05337

05107

05017

04114

03116

03115

02559

0235502274

0223802228

01539

01443

01193

01097ndash0453

ndash0625

ndash1151

ndash2121

ndash2261

ndash2288ndash2974

ndash3261ndash3353

ndash4072

ndash4114



Tue 22 Jan 2008 150742







28

29

30

31


Table 2 (continued)





general laborers

SMC

SMM

PMO

GL





ndash5014

ndash5121

ndash5589

ndash5979




Tue 22 Jan 2008 150742



RESULTS










Tue 22 Jan 2008 150742







Model 1 Model 2

0 SE 0 SE






Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742




Age Age Squared

0 SE 0 SE N0







inco

me

inco

me

ageage

inco

me

inco

me

ageage


Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742







28

29

30

31


Table 2 (continued)





general laborers

SMC

SMM

PMO

GL





ndash5014

ndash5121

ndash5589

ndash5979




Tue 22 Jan 2008 150742



RESULTS










Tue 22 Jan 2008 150742







Model 1 Model 2

0 SE 0 SE






Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742




Age Age Squared

0 SE 0 SE N0







inco

me

inco

me

ageage

inco

me

inco

me

ageage


Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742



RESULTS










Tue 22 Jan 2008 150742







Model 1 Model 2

0 SE 0 SE






Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742




Age Age Squared

0 SE 0 SE N0







inco

me

inco

me

ageage

inco

me

inco

me

ageage


Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742







Model 1 Model 2

0 SE 0 SE






Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742




Age Age Squared

0 SE 0 SE N0







inco

me

inco

me

ageage

inco

me

inco

me

ageage


Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742




Age Age Squared

0 SE 0 SE N0







inco

me

inco

me

ageage

inco

me

inco

me

ageage


Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742




Age Age Squared

0 SE 0 SE N0







inco

me

inco

me

ageage

inco

me

inco

me

ageage


Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742







0 SE 0 SE 0 SE






Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742







Model 1 Model 2


0 SE 0 SE 0 SE 0 SE





Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742













Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742







0 SE 0 SE


N 2421 2441R2 13 17




Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742










Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742







CONCLUSIONS



liber

taria

nndashau

thor

itaria

n sc

ale



Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742

APPENDIX

















REFERENCES





















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742









































Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742



















Tue 22 Jan 2008 150742

Documents

Class and Status: The Conceptual Distinction and its ... · of class and status as two qualitatively different forms of social stratification retained currency through the 1970s,