Occupational Sociology, Yes: Class Analysis, No: Comment on Grusky and Weeden's "Research

Agenda"
Author(s): John H. Goldthorpe
Source: Acta Sociologica, Vol. 45, No. 3 (2002), pp. 211-217
Published by: Sage Publications, Ltd.
Stable URL: http://www.jstor.org/stable/4194931
Accessed: 19/11/2010 11:31
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=sageltd.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Sage Publications, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to Acta
Sociologica.

http://www.jstor.org

.......
..
... ..
.:. ::
ACTA SOCIOLOGICA2002

COMMENT

Occupational Sociology, Yes: Class Analysis, No:

Comment on Grusky and Weeden's Research
Agenda
John H. Goldthorpe
NuffieldCollege,Oxford,UK
social groupings' to form 'around functional
niches in the division of labour' (p. 203).
I have two comments to make on this
account. First, there appears to be no place in it
for a tradition of class analysis within which
many European sociologists would see themselves as working. This tradition developed long
before the collapse of Marxism, and in fact has
stood as a major alternative to, and basis for
criticism of, Marxism - including at times when
Marxism has been in high academic fashion.
Thus, I for one have not 'discarded' Marxist or
1. The intellectual context
marxisanttheories of class for the simple reason
An initial question is that of why G&Wshould that I never adhered to them, and indeed have
think that class analysis is in need of recon- consistently opposed them (e.g. Goldthorpe
struction, at all events of the very radical kind 1972, 1990). This non-Marxist tradition of
that they propose. Their reading of the present class analysis - with which American sociolosituation would appear to go something like this gists have never easily engaged - is, I would
(cf. Grusky & Sorensen 1998). With the failure maintain, alive and well, not in need of radical
of the 'Marxist project', politically as well as reconstruction but developing according to its
intellectually, class analysts have 'discarded own dynamic.'
This brings me to my second comment.
comprehensive class-based theories of history'
(Grusky and Weeden 2001:205) and have Within the non-Marxist tradition, an interest in
indeed so scaled down their ambitions that class formation, broadly understood, has always
their work now amounts to 'little more than been present. However, not only is the historical
describing the micro-level association between inevitability of class formation rejected but so
class membership and life chances of various too in effect is the idea that class formation in
kinds' (p. 212). In particular, issues of class G&W'smore limited sense of structuration has
formation or, as G&W would rather say, of to be empirically demonstrated before class
'structuration', are avoided. Consequently, it analysis can take on any significant intellectual
becomes difficult to mount any strong resis- interest or socio-political relevance. In other
tance to claims of the 'death of class' or to words, the existence of this tradition serves as a
oppose the postmodernist 'retreat from the basis for questioning G&W's assumption that
productive sphere' (p. 214). If class analysis is structuration - i.e. the process through which
to be salvaged, G&Wthen contend, it will have economic categories become 'real social group- is the sine
to be 'ratcheted down' to a level where there is a ings' (cf. Giddens, 1973:107-112)
.2
better chance of 'structuration' being observed. qua non of class analysis.
It is true that, as G&W remark, for
And most promising here, in their view, is the
occupational level, where it is possible for 'real sociologists working in the non-Marxist tradi-

Grusky and Weeden (G&W) have outlined an

imaginative and potentially highly rewarding
research agenda: one that could well revitalize
the field of occupational sociology. However,I do
not believe that this agenda can serve as an
adequate substitute for class analysis. In what
follows, I develop this argument under a series
of headings.

212

ACTA SOCIOLOGICA2002

tion a major concern is with the association,

and actual causal connection, between class
and differential life-chances - with how class
influences what happensto people.G&W'ssomewhat dismissive attitude towards this concern
presumably arises from the fact that it is not one
to which the degree of class formation need
have any great relevance. However, pace G&W,
there is also a complementary concern with the
association, and causal connection, between
class and differential life-choices - with how
class influences howpeopleact. And while in this
case it is accepted that the degree of class
structuration in G&W'ssense may be crucial, as
with some forms of class-based collective action,
in regard to other class-linked regularities in
action, widely revealed by empirical research.
the importance of such structuration is not
presupposed. It is rather treated as a central
empirical and theoretical issue.

VOLUME45

such groups - i.e. their specific subcultures,

norms, institutions, organizations, etc. 3
I do not of course dispute that what G&W
propose is one possible explanatory approach.
But I do question whether it is the only viable
one, as they would maintain, or indeed that
which, as far as class analysis is concerned, has
at the present time the greatest potential.
Within the European tradition to which I
have referred,an alternative approach has been
on the following lines. The starting-point is with
concepts of class that are quite openly 'fashioned' by sociologists rather than by actors. The
aim is then to show how different class
positions, as understood via these concepts,
create different sets of constraints and opportunities for the individuals who hold them; and, in
turn, to explain class-linked variation, both in
what happens to individuals and in how they
typically act, via causal processes that originate
in their different 'class situations'. Moreover,as
regards variation in action, the :furtheraim has
2. Real versus nominal groupings
been to bring out how this can be understood in
terms of what, after Popper (see 1994: esp. ch.
G&W believe that class analysis should be 8), might be called the 'logic of the situation': or,
founded on 'real social groupings', recognized that is, as reflecting courses of action that, given
by and meaningful to the individuals they particular class situations, are rational, at least
comprise, rather than on mere aggregations in a subjective sense, and therefore intelligible
'fashioned by academics' (p. 208). The opposi- (verstdndlich)(cf. Goldthorpe 2000: ch. 6).
tion that they here set up, I would then note, is
As a classic example of this approach, one
not simply that of realism as against nominal- could cite David Lockwood's The Blackcoated
ism, which is the formulation they favour. It is Worker(1958). In this study Lockwood's conalso that of actors' (or lay) concepts as against cern was with the well-documented reluctance
researchers' (or scientific) concepts.
of clerks, in later 19th- and earlier 20th-century
The issues that thus arise are complex and Britain, to ally themselves with manual workers
contested. G&Wdo not in fact enter into them, in the trade union movement and in support for
and neither for present purposes will I, although the Labour Party. Through a detailed empirical
I have done so elsewhere (Goldthorpe2000: ch. examination of the class situation of clerks, as
1). However, the position I hold is that while decomposed into their 'market' and 'work'
sociologists may in certain lines of investigation situations, Lockwood was able to demonstrate
need to understand actors' concepts and to that Marxist claims of the 'proletarianisation'of
trace the relationship between these concepts clerical labour were much exaggerated and that
and their own, it is always permissible, and most clerks remained significantly differentiated
quite often essential, that sociologists should from manual workers. In turn, the lack of
develop and make operational concepts that are enthusiasm on the part of clerks for collectivist
far removed from those of 'lay members'. From as opposed to individualistic strategies for
this standpoint, I would then challenge G&W's defending and promoting their interests, or the
claim (p. 209) that, in seeking to go beyond distinctive features of the collective action in
descriptions of the association between class which they did in some cases engage, could not
and life-chances or life-choices to the specifica- be dismissed as mere expressions of 'false
tion of causal processes, sociologists will not in consciousness'. They were, rather, to be underfact get far 'without abandoning aggregate stood as quite rational responses to their class
formulations', and (I take it) aiming, instead. situation.
to express all explanatory accounts in terms of
Lockwood was indeed sensitive to differindividuals' memberships of groups with which ences existing among clerks: for example,
they actually identify and of the properties of among bank clerks, local government clerks,

Commenton Gruskyand Weeden 213

railway clerks, and so on. But what, for the

present discussion, is chiefly notable is that in
his analyses these more specific occupational
groupings - though no doubt 'more real', in
G&W'ssense, than clerks as a class - play only a
secondary role. They are treated, one could say,
as a source of occupationalvariation on a class
theme;and as variation that, while interesting in
itself, can only be adequately understood in
relation to this theme.4
The class schema that I have developed,
with various colleagues, over the past twenty
years or so (variously known as the 'Goldthorpe', 'Erikson-Goldthorpe-Portocarero' or
'CASMIN'schema) was in its earliest formulation much influenced by Lockwood's ideas of
market situation and work situation. Subsequently, its theoretical basis has been refined
and elaborated in terms of employment relations. Class positions are seen as being differentiated initially by employment status - i.e.
into those of employers, self-employed workers
and employees; but then further, in the case of
employees, by typical forms of employment
contract, explicit and implicit (Goldthorpe
2000: ch. 10). However, the use of the schema
in class analysis is still on essentially the same
lines as those pursued by Lockwood. That is to
say, the aim is to bring out the constraints and
opportunities typical of different class positions,
and especially as these bear on individuals'
economic security, stability and prospects; and
then, on this basis, to construct testable
explanations of empirical regularities of interest.
For example, Evans (1993) has followed
this approach in providing an explanation of
class differences in political partisanship and of
changes in the pattern of these differences over
the life-cycle. And, in similar fashion, Richard
Breen and I (Goldthorpe 1996; Breen & Goldthorpe 1997; Breen 2001) have developed an
explanation of persisting class differences in
educational choice and attainment. In explanations of the kind in question, it is not then
necessary to claim or to suppose that classes are
'real social groups'. Little appeal to what G&W
would subsume under 'structuration' is
required. Regularities are seen as arising from
the extent to which individuals respond in
similar ways to similar class situations, not
from the extent to which they are influenced by
class-specific subcultural values or social
norms, except perhaps where norms serve
instrumentally as useful 'rules of thumb' (see
further, Breen & Goldthorpe 1997).5 And,
indeed, from this standpoint 'culturalist'

accounts based on the study of real social

groups may be called into question as not
causally adequate to the regularities to be
explained, as, say, in the case of accounts of
class differentials in education that are derived
from school ethnographies (e.g. Willis 1977).
3. Occupational closure and class mobility
G&Wbelieve that social mobility is 'governed by
the deeply institutionalized boundaries' that
actually exist between occupations (p. 208)
rather than by the divisions of class schemata
formed by social scientists. More specifically,
what should be seen as crucial to the explanation of mobility rates and patterns are the
'closure tactics' engaged in by members of
occupational groups or their organizations such as the intergenerational transmission of
businesses or job rights or licensing, certification and job demarcation - supplemented in
some instances by the influence of 'occupational
communities' with their own distinctive
'organic cultures' and life-styles. For this
among other reasons, the central concern of
the research agenda that G&Wpropose is then
to map out the 'local structuration' of occupations across modern societies in a systematic
way (p. 207).
As I remarked at the outset, I would see
such a project as one that could be of great
value and in particular in helping the sociology
of occupations to move beyond its tradition of
rather ad hoc case studies, the shortcomings of
which G&Wrecognize. However, as regards the
explanation of mobility, my accompanying
critical point - that their research agenda
cannot substitute for class analysis - also
applies.
Occupational closure or, in other words,
the restriction or regulation of the supply of
labour to an occupation can be a significant
factor in determining the relative rewards
associated with that occupation - its members'
rates of pay, conditions of employment, and so
forth. And we no doubt need to know more both
about how extensively and how effectively
closure practices do in fact operate in this
respect. However, what we can learn in this way
about mobility processes will be quite limited.
We may gain a better understanding of why in
some occupations there is a marked propensity
for intergenerational 'inheritance', or succession, in the sense of children following directly
in their parents' footsteps. But research into

VOLUME
45

214 ACTASOCIOLOGICA
2002

class mobility, at all events, is about a good deal

more than this. We want to be able to explain,
for example, not so much why doctors' children
have a high propensity to become doctors or
coal miners' sons to become coal miners, but
rather why those doctors' children (the majority) who do not become doctors are far more
likely to move into other kinds of professional or
managerial employment instead of becoming
manual wage-workers and, correspondingly,
why the sons of coal miners (again the majority)
who do not become miners are far more likely to
move into some other kind of wage-work than
into professional or managerial positions. And
in these respects, the closure practices of specific
occupations would seem of little relevance.
It is certainly true that, as G&W argue,
analyses of mobility made at the level of class
categories will tend to conceal what goes on at
the more detailed occupational level. But it does
not then follow from this, as they would
suppose, that a 'problem' for class analysis
arises, the solution to which is 'to disaggregate
until realist occupations are secured' (p. 209).
Not only does the remedy prescribedseem worse
than the disorder diagnosed but, in this case,
'decomposition' would actually guarantee,
rather than stave off death, so far as class
analysis is concerned.
The study of mobility at the class level is, I
would maintain, an entirely viable undertaking
in so far as three conditions are met: (i) the
classes fashioned by sociologists are not merely
arbitrary aggregations but have a clear conceptual grounding, and one that can be
implemented with adequate criterionvalidity i.e. so that the classification does adequately
capture what, conceptually, it is supposed to
capture; (ii) analyses of mobility based on the
classes distinguished reveal meaningful empirical regularities - this being in turn indicative of
their construct validity; and (iii) the way in
which classes are conceptualized itself provides
a starting-point for developing explanations, in
terms of social action - i.e. micro-to-macro
explanations (Coleman 1990) - for the regularities displayed.
Mobility research using the class schema
earlier referred to does, I believe, meet these
conditions to an encouraging degree. The
criterion validity of the schema was extensively
examined, with generally favourable results,
prior to a new instantiation of it being adopted
in 2001 as the official British socio-economic
classification (Rose & O'Reilly(eds.) 1997; Rose
& O'Reilly1998: see also Rose & Pevalin 2002),

and such work is now being extended in

connection with a proposed EU classification.
Through the application of the schema, it has
been possible to demonstrate important regularities, especially in relative mobility rates,
across both time and space and, further, distinctive 'mobility characteristics' of the classes
themselves (see e.g. Erikson& Goldthorpe 1992
and subsequent work within or following on the
CASMIN project).6 And at least some initial
efforts have been made at formulating possible
explanations of these regularities by reference to
the 'mobility strategies' taken to be typical of
members of different classes and in turn
intelligible in the context of the constraints
that their class situations impose and the
opportunities they afford (e.g. Goldthorpe
2000: ch. 11). This latter approach, it may be
added, provides a clear contrast with, as well as
alternative to, the focus proposed by G&Won
the 'closure tactics' of particular occupational
groups.
4. Collective class action
G&W observe (p. 205) that for Marxists and
neo-Marxists, social closure is of interest 'not
because it provides a vehicle for pursuing
purely local interests . . . but rather because it
allegedly facilitates the development of classwide interests and grander forms of interclass
conflict. But, they further comment, the
'aggregate classes identified by contemporary
sociologists have so far shown a decided
reluctance to act in accord with such theorizing'.
What I would note here are two implicit
suggestions, or at least assumptions. The first is
that those sociologists who have identified
aggregate classes do share alike in Marxist or
marxisant expectations about collective class
action. As against this, I have already stressed
the existence of a non-Marxist tradition of class
analysis to which G&W'sanimadversions do not
apply.The second suggestion or assumption, on
which I will concentrate here, is that collective
class action has to be conceived in essentially
Marxist terms: i.e. as involving class structuration, in the sense of the transformation of Klasse
an sich into Klassefiirsich, and in turn the largescale mobilization of individuals in action of a
dissident and conflictual, if not revolutionary,
character.
However. for those European sociologists
who, while not Marxists, have still sought to

Commenton Gruskyand Weeden 215

link class analysis to issues of political economy,

the focus of interest has been on a form of
collective class action quite different from that
envisaged under what might be called 'the
storming of the Winter Palace' model. This is
action taken through organizations - chiefly,
though not exclusively, working-class trade
unions - in the context of neo-corporatist or
'social partnership' institutional arrangements,
the rationale of which is in fact to accommodate
rather than to heighten class conflict. Most
typically, trade-union 'centrals' bargain with
governments and the peak organizations of
employers over regulating the wage and other
demands of their memberships so as to conform
with national economic strategy in return for
what they would regard as concessions and
advances, especially in social welfare and fiscal
policy (see e.g. Schmitter & Lehmbruch 1979;
Goldthorpe (ed.) 1984; Alvarez et al. 1991).
For the effectiveness of collective action of
this kind, the important requirements are that
the rank-and-file should be ready to trust their
national leaders and to accept the restraints
that the latter have to 'deliver'to their bargaining partners - and even under conditions in
which immediate gains could in fact be made at
the occupational or workplace level. So far,
then, as class formation is concerned, what is
chiefly important is the extent to which mobility
patterns are consistent with classes possessing
demographic identity (Goldthorpe 1987: esp.
ch. 12) or, if G&Wwould prefer,simply 'mediate'
structuration7 - together with such organizational factors as forms of rank-and-file representation and leadership accountability.
'Proximate' structuration, on which G&W
focus their attention, is a far more dubious
requirement. Elster's definition of class consciousness (1985:347) has here a particular
aptness: i.e. 'the ability to overcome the freerider problem in realizing class interests' essential to which, at the level of 'real social
groups', may in fact be class-oriented inaction.
In other words, in so far as structuration does
occur chiefly on the basis of occupations or
workplaces, it may lead to the strategic pursuit
of class interests being subverted by attempts to
advance interests of a more local and sectional
kind.
Now G&W do in fact at one point give
recognition to this argument. They remark (p.
214) that 'low-level structuration . .. is sometimes assumed to undermine aggregate organizational forms', and in this regard they contrast
the US case with that of Sweden where class-

based organization reflects the absence of

'competing local structuration'. However, they
then add that while Sweden may thus give some
support to a 'neo-Marxian' position, it is
'unclear' whether class organization of the
kind in question extends beyond Sweden or at
all events Scandinavia.
In response to this, two things might be
said. First, even if Sweden or Scandinavia could
be regarded as distinctive and even if (another
point G&W might have wanted to make) the
forms of neo-corporatism that operated there in
the postwar years have now broken down, this
does not detract from their theoretical significance - any more than the decay of craft and
other kinds of occupational unionism in countries such as the UK removes interest in the
forces that once sustained them. Secondly,
though, it is not in fact the case that neocorporatism was confined to Sweden or Scandinavia, nor yet that neo-corporatist institutions
and practices are now in uniform decline. At
least in 'co-ordinated' as opposed to 'liberal'
market economies, issues of the level at which
economic interests are aggregated and represented remain quite crucial and there is no
reason for supposing that a 'ratcheting down' to
that of real social groups, occupational or
otherwise, is inevitable, nor, conversely, that
higher-level, and including class-based, strategies are henceforth precluded (cf. Iversen et al.
(eds.) 2000; Hall & Soskice (eds.) 2001).
5. Conclusion
G&Wobserve (p. 214) that in 'characterizing
stratification systems of the past, sociologists
have typically relied on categories that were
embedded in the fabric of society (e.g. estates,
castes)' and that the resort to nominal categories is 'peculiarly modern'. This may be true,
but why should the modern (more social
scientific?) approach be thought mistaken? It is
not evident, to say the least, that the studies we
have of estate or caste stratification are sociologically superior to those of modern class
structures. And there are at all events good a
priori reasons for supposing that in modern
gesellschaftlichsocieties, where job and residential mobility are frequent and multiplex social
networks uncommon, to base the study of
stratification solely on gemeinschaftlichentities
will prove unduly limiting.
I welcome G&W'sresearch agenda, even if
for reasons that do not correspond with their

45
VOLUME

2002
216 ACTASOCIOLOGICA

own motivations, and I appreciate their honest

recognition that, so far, they are dealing largely
in speculation and that, as research proceeds, it
may turn out that even at the occupational level
a high degree of (proximate) structuration is
only rarely to be found - as would be my own
expectation. But this only increases my fear that
in treating their agenda as one for class analysis
rather than for occupational sociology, they
may end up not rescuing class analysis from its
present detractors but rather selling it seriously
short. For, as I have tried to show, there is no
reason, methodological or theoretical, why class
analysis need be dependent on the success of a
'search for structuration' of the kind in which
G&Wpropose to engage.

reconstructions of situational logical to the more formal

development of some kind of 'rational choice' or 'rational
action' explanations.
6I am somewhatpuzzledthat G&W(p. 210) should regard
Erikson& Goldthorpe(1992) as a work proposingor defending
any kind of 'convergencethesis'. Whatthey presumablyhave in
mind is something rather different - i.e. its support for a
qualifiedversionof the hypothesis(Feathermanet al. 1975) of a
'basic similarity' in endogenous mobility regimes across all
societieswith marketeconomiesand nuclear familysystems.In
fact, in this regard,Eriksonand I spendmuch time documenting
the idea of specific cross-national 'variations on a theme',
including variationsderivingfrom distinctivenational employment systemssuch as the Germanand Japaneseto which G&W
refer.And again the point may be made that such variationis
scarcelycomprehensiblewithout recognitionof the theme.
7 Class formation in this sense is necessary in that if
mobility were such that classes had no significant degree of
demographiccontinuity,nor then of identity,overtime (which is
not in fact the case). the very idea of class interests would
become problematic.

Acknowledgements

References

I am grateful to RichardBreen, David Grusky,YaojunLi and

Herman van de Werflhorstfor their helpful comments on an
earlierdraftof this paper.

Alvarez, R. M., Garrett, G. & Lange, P. 1991. Government

Partisanship.LaborOrganizationand MacroeconomicPer-

Notes
' It is inadequateto characterizethis traditionsimply as
'Weberian'or 'neo-Weberian'as, it seems, G&Wintend(p. 214).
Other influential figures who should be recognized include
JosephSchumpeter,KarlRenner,FritzCroner,TheodorGeiger.
David Lockwoodand (early) Ralf Dahrendorf.The traditionis
not of course without its internal differencesand conflicts,any
more than is the Marxisttradition.Its lack of influenceor even
recognitionin Americanworkin the fieldof social stratification
is probably related to the degree of polarizationthat exists
between the 'socioeconomicstatus attainment' and ,narxisant
schools, which G&W'spaperitselfreflects.
2 I assume here and throughout that in their use of the
conceptof 'structuration'G&Ware indeedfollowingGiddens,to
whom they severaltimes refer.However,it may also be observed
that while they acknowledge Giddens' distinction between
'mediate' and 'proximate'(or 'local' ) structuration,it is on
the latter that they concentrate.I would myselfpreferto work
with the idea of class formation.the basic level of which is seen
as demographic- i.e. the formationof classes as collectivities
with a recognizableidentity over time - without however,the
relation between socio-cultural and socio-politicalformation
being regardedas necessarily hierarchical(see further below
and also Goldthorpe1987:329-345 esp.).
3 This claim is made specificallyin regardto researchinto
social mobility,but it is, I assume, one that G&Wwouldwish to
maintain as a quite general one.
4 Lockwoodalso complementshis account of the class - i.e.
marketand work - situation of clerks with a considerationof
their 'status situation'. However, this is likewise treated as
secondary(cf. 1958:211), and Lockwoodis in any event more
concerned here with 'the position of the individual in the
hierarchyof prestigein the societyat large'(19 58:15) than with
the involvementof individualsin 'real'status groups.
' Perhapsthe main developmentin this approachin the
recentpast has been the attemptto go beyondrelativelyinformal

formance. American Political Science Review, 85. 539-556.

Breen, R. 2001. A Rational Choice Model of Educational

Inequality.Estudio2001/166, InstitutoJuan March,Madrid.
Breen, R. & Goldthorpe,J. H. 1997. ExplainingEducational
Differentials:Towards a Formal Rational Action Theory.
Rationality and Society, 9, 2 75-305.
Coleman, J. 1990. Foundationsof Social ThleoryCambridge, Mass.

HarvardUniversityPress.
Elster, J. 1985. Making Sense of Marx. Cambridge: Cambridge

UniversityPress.
Erikson, R. & Goldthorpe, J.H. 1992. The Constant Fluix:A Study
of Class Mobility in Industrial Societies. Oxford: Clarendon

Press.
Evans,G. 1993. Class,Prospectsand the Life-Cycle:Explaining
the Associationbetween Class Position and PoliticalPreferences. Acta Sociologica. 36. 263-276.

Featherman,D.,Jones,F.L.&Hauser,R. M. 19 7 5. Assumptions
of Social MobilityResearchin the US: the Case of Occupational Status. Social ScietnceResearch1,4, 329-360.
Giddens. A. 1973. The Class Structure of the Advanced Societies.

London:Hutchinson.
Goldthorpe,J. H. 1972. Class, Status and Party in Modern
Britain:SomeRecentInterpretations,Marxistand Marxisant.
Archives europeennes de sociologie.13. 342-3 72.
G,oldthorpe, J. H. (ed.) 1984. Order and Conflict in Contemporary

Capitalism.Oxford:ClarendonPress.
Goldthorpe J. H. 1987. Social Mobility and Class Struicture in

ModernBritaini.2nd ed. Oxford:ClarendonPress.

Goldthorpe,J. H. 1990. Employment.Class and Mobility:A
Critique of Liberal and Marxist Theories of Long-Term
Change.InH. Haferkamp& N. J.Smelser(eds.),Modernization
andSocialChange.Berkeley:Universityof CaliforniaPress.
Goldthorpe,J. H. 1996. ClassAnalysisand the Reorientationof
Class Theory:The Case of PersistingDifferentialsin Educational Attainment. British Jourinalof Sociology, 47, 481-5(05.
Goldthorpe, J.H. 2000. On Sociology: Numnbers,Narratives and thle

Oxford:OxfordlUniversity
Integrationof Researchand Theoryl.
Press.
Grusky,I). B & Sorensen, J. B. 1998. Can Class Analysis be
Salvaged' American Journal of Sociology, 103, 1187-1234.

Grusky.D. B. & Weeden, K. A. 2(0(01.DecompositionWithout

Commenton Gruskyand Weeden 217

Death: A Research Agenda for a New Class Analysis. Acta
Sociologica, 44, 203-218.

National Statistics and Economic and Social Research

Council.

Hall, P. A. & Soskice, D. (eds.) 2001. Varietiesof Capitalism. Rose, D. & O'Reilly, K. 1998. The ESRC Review of Government
Oxford:OxfordUniversityPress.
SocialClassifications.
London:OfficeforNationalStatisticsand
Economicand Social ResearchCouncil.
Iversen,T., Pontusson,J. & Soskice,D. 2000. Unions,Employers
andCentralBanks.Cambridge:CambridgeUniversityPress.
Rose, D. & Pevalin, D. J. 2002. Confirming Classes: Validatingthe
Lockwood, D. 1958. The Blackcoated Worker. London: Allen &

Unwin.
Popper, K. 1994. TheMyth of the Framework.London: Routledge.
Rose, D. & O'Reilly, K. (eds.) 1997. Constructing Classes: Towardsa
New Social Classification for the UK. London: and Office for

National Statistics Socio-Economic Classification. London: Office

for National Statistics.

Schmitter, P. & Lehmbruch G. (eds.) 1979. TrendsTowards
CorporatistIntermediation. London: Sage.
Willis, P. 1977. Learning to Labour.Farnborough: Saxon House.

A Class Analysis for the Future? Comment on

Grusky and Weeden: 'Decomposition Without
Death: A Research Agenda for a New Class
Analysis'
Gunn Elisabeth Birkelund
Institute for Sociology and Geography,Universityof Oslo, Norway
As the Marxianprojectfalls out of favour,scholars
have thereforesettled into some versionof
Weberianismor postmodernism,neitherof which
pays much attention to occupation-level
structuration.We have outlineda quasi-Durheimian
third roadthat refocusesattention on localforms of
structurationwithin the divisionof labor (Grusky &
Weeden 2001:214)

Grusky & Sorensen 1998). In the 1990s we

witnessed strong voices of anti-class criticism,
against which various representatives of class
analysis defended the importance of class, yet
with concessions such as arguing in favour of a
weaker form of class than previously advocated.
This debate on classes can be divided into two
parts, one on whether class matters at all in
modern societies, and another on how we best
measure class in modern societies. Whereas the
former debate on the fruitfulness of class
and
Grusky
In order to revitalize class analysis,
Weeden (hereafter G&W)launch a new orienta- altogether involves anti-class critics and class
tion and approach to class. Their main argu- defenders, the latter debate activates class
ment is that the concept of class should be defenders, with different perspectives on class
disaggregated into occupational categories, (such as Wright 1997; or Erikson & Goldthorpe
such as carpenters, teachers, and so on. 1992). It is mainly this second debate that G&W
Occupations are gemeinschaftliche groupings, address,though they are well aware that a better
G&Wargue, that are institutionally embedded understanding of class might have a bearing on
in society. At the level of occupations, we will the first debate as well. There are many issues on
find occupation-based organizations, such as which I would agree with GW However, even if
unions, that participate in producing social their new approach to class analysis is interestidentity, social closure, various types of out- ing, I am not quite convinced. I will addresstheir
comes (such as income) and possibly also argument by focusing on three issues:
collective action. If class is defined as occupaI. The realist claim
tions, the scholarly and lay understandings of
II. Gemeinschaft
class correspond as well.
III. Globalization
G&Waddress an important topic (see also