Professional Documents
Culture Documents
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.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
The University of Chicago Press is collaborating with JSTOR to digitize, preserve and
extend access to American Journal of Sociology
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Can Class Analysis Be Salvaged? 1
David B. Grusky
Stanford University
Jesper B. Srensen
University of Chicago
The ongoing retreat from class analysis can be attributed to the de-
clining appeal of aggregate representations of class coupled with the
virtual absence of any disaggregate alternatives. When local solidar-
ities are ignored, the weakness of conventional aggregate models is
easily misinterpreted as evidence of generic destructuration, and
standard postmodernist formulations are accordingly difficult to re-
sist. Although local structuration is often regarded as sociologically
trivial, the available evidence suggests that such class analytic pro-
cesses as closure, exploitation, and collective action emerge more
clearly at the level of disaggregate occupations than conventional
aggregate classes.
It has long been fashionable to claim that conventional class analysis,
Marxian or otherwise, is somehow deficient and in need of revision. In
recent years, such criticisms of the class analytic framework have esca-
lated, with many scholars now feeling sufficiently emboldened to argue
that the concept of class is ceasing to do any useful work for sociology
(Pahl 1989, p. 710). Although this retreat from class analysis was antici-
pated in the late 1950s by Nisbet (1959) and again in the 1980s by Gorz
1
This article was presented in draft form to the Association for the Historical Study of
Social Mobility (Birmingham, England), the Social Class and Politics Working Group
(Washington, D.C.), and the American Sociological Association (New York). The re-
search reported here was carried out with support from the Deans Research Fund
of Stanford University, the Presidential Young Investigator Program of the National
Science Foundation (SES-8858467), and the Stanford Center for the Study of Families,
Children, and Youth. We are most grateful for the comments of Steven Brint, Clem
Brooks, Bruce Carruthers, Terry Clark, Mariah Evans, Eliot Freidson, Mark Grano-
vetter, Robert Hauser, Jerald Herting, Paul Kingston, Seymour Lipset, Jeff Manza,
Robert Mare, Aldon Morris, Steve Rytina, Aimee-Noelle Swanson, Jan Pakulski,
Charles Ragin, Eric Rice, Arthur Stinchcombe, Robert Szafran, Szonja Szelenyi, Da-
vid Weakliem, Kim Weeden, and Nancy Weinberg. We also profited from the detailed
and often passionate critiques of our AJS reviewers. Direct all correspondence to Da-
vid B. Grusky, Department of Sociology, Stanford University, McClatchy Hall, Stan-
ford, California 94305-2047.
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
(1982) and other recanting Marxians (e.g., Offe 1985; Lukes 1984; Hobs-
bawm 1981), the present round of anticlass rhetoric is unprecedented in
its popularity, especially in Europe where class analysis has historically
enjoyed a privileged position (Pakulski and Waters 1996a, 1996b, 1996c;
Pakulski 1996; Lee and Turner 1996; Scott 1996; Clark 1996; Casey 1995;
Barnes 1995, chap. 7; Joyce 1995; Kingston 1994; Clark and Lipset 1991).
This development constitutes a striking repudiation of our disciplinary
heritage; in fact, it was not so long ago that commentators as mainstream
as Stinchcombe (1973) could allege, without generating much in the way
of controversy, that social class was the one and only independent variable
of sociological interest. In the present article, we reenter these ongoing
debates, not by providing yet another critique or defense of conventional
class analysis, but rather by attempting to rebuild it in ways that may
usefully redirect future research.
The existing literature is short on such reconstructive efforts. At pres-
ent, virtually all protagonists adopt one of three analytic positions, two
of which arguably take contemporary anticlass critiques too seriously,
whereas the third one fails, in our view, to take them seriously enough.
1. The dominant form of revisionism, at least within American sociol-
ogy, rests on the claim that long-standing cleavages of class, occupation,
and status are giving way to new cleavages generated in institutional are-
nas (e.g., schools, organizations) that have become more prominent with
the transition to postindustrialism. These new cleavages have been de-
fined in disparate ways; for example, Meyer (1994, p. 734) claims that
researchers mistakenly use terms like class or socioeconomic status when
education is the main phenomenon at issue, while Baron (1994, p. 390)
suggests that organizational affiliation is . . . becoming increasingly im-
portant, and Dunleavy (1980, p. 378) argues that the emerging cleavage
between state and public employment cannot be directly assimilated into
notions of occupational and social class (also see Saunders 1990; Brint
1984; Hodson 1978; Habermas 1975; OConnor 1973).
2. The deconstructionists have suggested, by contrast, that the prior
reformulations gloss over more fundamental declines in the effects of all
structural cleavages, with the implication thus being that new theories,
perhaps more cultural than structural, [are] in order (Davis 1982, p. 585;
also see Beck 1992; Bell 1987). This line of argumentation underlies stan-
dard forms of postmodernism that seek to represent new social move-
ments (e.g., feminism) as the vanguard force behind future change in strat-
ification systems (Larana, Johnston, and Gusfield 1994; Eyerman 1994;
Eder 1993; Inglehart 1990, 1977; Laclau and Moffe 1985; Offe 1985; Tour-
aine 1981; Melucci 1980). In all such accounts, the labor movement is
represented as a fading enterprise rooted in the old conflicts of the work-
place and industrial capitalism, whereas new social movements are as-
1188
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
1189
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
TABLE 1
Realist Nominalist*
Aggregated ..................... Reputational models (e.g., Marxian and neo-Marxian Hollingshead occupational Weberian and neo-
Warner, Meeker, and models (e.g., Wright classification (Hollings- Weberian models (e.g.,
Eells 1949) 1985) head and Redlich 1958) Edwards 1943; Erikson
and Goldthorpe 1992)
Disaggregated ................ Strong form (Grusky and Prestige and socioeconomic Weak form (Grusky and
Srensen, this discussion) scales (e.g., Treiman Srensen, this discussion)
1977; Duncan 1961)
* We include under this designation class analytic formulations that treat class awareness, identification, and action as historically contingent and probabilistic.
The Erikson-Goldthorpe classification scheme clearly has some realist features to it. However, insofar as it is intended to capture indirectly the structure of market and
work conditions, one might argue that its nominalist tendencies are dominant. In this context, it does not much matter whether the work conditions of interest are construed
broadly (Goldthorpe 1987, p. 40) or narrowly (Erikson and Goldthorpe 1992, pp. 3547), since in either case the constituent categories are viewed as proxies for more
fundamental class attributes. The class analysis proffered by Goldthorpe is, therefore, more limited than that of Marxian provenance, since it neither assumes exploitative
relations between classes nor a gemeinschaftlich model of working-class action based on the local solidarities of workplace or community (Goldthorpe and Marshall 1992,
p. 384).
1191
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
2
As argued by Bourdieu (1984, p. 101), occupations are only approximate indicators
of underlying class boundaries, since even highly detailed classifications may not per-
fectly identify sets of agents who are placed in homogeneous conditions of existence.
1192
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
and hence the best single indicator of class (Blau and Duncan 1967, p.
6; also see Parsons 1954; Sorokin 1927). This claim was seen as sufficiently
self-evident to make more detailed explication unnecessary; in fact, when
Wright (1979, p. 12) completed his comprehensive review of competing
models of class, he properly concluded that there are relatively few sus-
tained theoretical reflections on the logic of linking class to positions
within the technical division of labor. The absence of a credible defense
of this default model rendered it especially susceptible to the antioccupa-
tion critiques that took off in the 1980s and continue today. In the neo-
Marxian variant of such critiques, occupations are assumed to be merely
technical in character, whereas real classes capture the social relations
of production (Wright 1980, p. 177). By contrast, the radical deconstructi-
onist accounts of Conk (1978) and others (e.g., Casey 1995) represent occu-
pational classifications as statistical conveniences that reflect, in large
part, the interests and assumptions of the classifiers themselves (i.e., statis-
ticians) rather than the operation of more fundamental technical or social
boundaries. The neo-Marxian and deconstructionist accounts proceed,
then, from strikingly different premises, yet they coincide insofar as both
represent conventional occupational categories in largely nominal terms.
The purpose of the present section is to reconsider the logic of these ac-
counts and thereby provide a more sustained defense of the sociological
(i.e., occupation-based) model of class than has heretofore been available.
We might begin by asking why neo-Marxians so frequently characterize
the division of labor as purely technical (see, esp., Wright 1980). As best
we can determine, the intention here is to imply that functional groupings
in the division of labor are nominal in character and are unlikely, there-
fore, to become social groupings that constitute meaningful bases of social
action and identification. This viewpoint is aptly summarized in the ral-
lying cry that technical features do not entail social features (Abercrom-
bie and Urrie 1983, p. 109). If one were to take the concept of a technical
division of labor seriously, it would further suggest (a) that the configura-
tion of technical tasks mechanically determines the boundaries between
occupations, (b) that the rise and fall of occupations can likewise be attrib-
uted to the forces of technological advance and obsolescence, and (c) that
the incumbents of occupations form purely nominal unions held together
by virtue of shared task demands. Under such a formulation, the relation-
ship between occupational structure and function becomes quite unprob-
lematic, with the former being effectively determined by the latter (see
Abbott 1988, pp. 3558; Turner and Hodge 1970). We suspect that as-
sumptions of this sort inform at least some of the more celebrated critiques
of the sociological model of class. At the same time, we are reluctant to
push our characterization too far, since it perforce rests on rather sketchy
1193
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
3
For all its neo-Marxist credentials, this characterization of the division of labor would
nonetheless be unappealing to Marx (1963), since he rejected all efforts to interpret
labor arrangements as the outcome of some simple technical imperative. The alterna-
tive account favored by Marx is itself partial and problematic. Indeed, whereas Marx
(1963) emphasized the role of capitalists in fashioning the division of labor, we shall
suggest that occupational associations are equally involved in jurisdictional struggles
over functional niches.
4
This approach implies that jurisdictional settlements may well differ by country. At
the same time, there are obvious limits to such variability, not merely because the
usual forces of cross-national diffusion are at work, but also because the technical
structure of tasks renders certain jurisdictional solutions more likely than others.
1194
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
1195
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
mapped into one and only one class. However, insofar as such simplifying
assumptions continue to be relied upon, our approach requires class ana-
lysts to identify the dominant jurisdictional settlements at the disaggregate
level.
SOCIOPOLITICAL ACCOUNTS
We next consider the realist claim that disaggregate classes are closed
and self-aware sociopolitical groupings that act collectively and imply a
specific style of life. When scholars seek to market a particular class model,
they typically proceed by first representing certain analytic criteria as es-
pecially crucial (e.g., social closure) and then showing that these criteria,
if so privileged, lead inevitably to the preferred mapping. The wide range
of class models on offer can therefore be understood as reflecting corre-
spondingly wide differences of opinion as to which trait of class society
is regarded as primary or most important from the viewpoint of causal
dependence (Ossowski 1963, p. 138; see also Grusky 1994, p. 6). In the
present context, we can afford to be more inclusive than usual in selecting
the criteria of interest, if only because the case for disaggregation can be
made quite broadly. We shall thus proceed by examining the full comple-
ment of criteria that are conventionally considered in advancing sociopo-
litical models of class (i.e., identification, awareness, closure, collective ac-
tion, lifestyles).
Identification
It is natural to begin by considering the subjective domain of class sys-
tems. For all their differences, Marx and Durkheim both saw the future
of capitalism (or industrialism) in subjectively realist terms, with Marx
assuming that individuals would become aware of and identify with ag-
gregate classes, while Durkheim assumed much the same with respect to
disaggregate occupations. Moreover, both Marx and Durkheim antici-
pated a great clearing operation in which competing sources of solidarity
(e.g., regional, ethnic) would gradually wither away, thus allowing
production-based identities to become central and all-encompassing
(Marx 1964, pp. 48081; Durkheim 1933, pp. 1518, 2429). The question
that naturally arises is whether either of these realist stories might plausi-
bly be maintained in light of contemporary theory and evidence on iden-
tity formation.
In surveying the extensive and conflicting literature on such matters,
one is immediately reminded that sociological data leave ample room for
predispositions to affect conclusions, no matter how value free the analyst
attempts or purports to be. Although one can find, then, a great many
1196
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
1197
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
tional affiliations (e.g., Zabusky and Barley 1996; Mortimer and Lorence
1995; Freidson 1994, pp. 8991; Epstein 1990; Van Maanen and Barley
1984; Vaught and Smith 1980; Mortimer and Simmons 1978). The forego-
ing research may be seen as bearing out the classical Durkheimian view
that occupational affiliations are sufficiently powerful to become cross-
situational identities of the essentialist variety (see Laclau and Mouffe
1985). This is not to suggest that all competing sources of identity have
cleared away as Marx and Weber anticipated. In fact, the postmodern
presumption that identities are structured by an increasingly broad range
of sources (e.g., sexual preference) may well have merit, yet we ought not
forget that detailed occupations remain fundamental sources of identity
for a great many workers, especially those in the professional, technical,
craft, and service sectors. There is no evidence here of a zero-sum game
in which the (presumed) strengthening of racial, ethnic, and related soli-
darities occurs in tandem with the decline of production-based affiliations.
The postmodernist dismissal of the productive realm accordingly falls
short when the analysis is properly located at the disaggregate level. By
way of defense, some postmodernists might nonetheless suggest that, just
as aggregate solidarities are withering away, so too are disaggregate ones
as craft occupations are deskilled, rationalized, and thereby stripped of
their totalizing and identity-bestowing features. This formula is mis-
leading on many counts but most obviously because true occupational
communities can be found in manual and nonmanual sectors alike. The
professions, for example, typically provide a worldview that is preemi-
nently craftlike in its totalizing fusion of work and leisure (see Mills 1956).
In this regard, sociologists should presumably appreciate that the much
vaunted sociological perspective is merely one of many occupationally de-
termined worldviews, all of which override naive or lay interpretations
of social reality and thus provide incumbents with a deeply felt sense of
being apart and different (Van Maanen and Barley 1984, p. 300). The
available evidence suggests that such intense identification is characteris-
tic of occupations requiring extended training, esoteric skills, or a strong
service orientation. By this account, some craft communities are no doubt
eroded by the forces of deskilling, but new communities are simulta-
neously arising as esoteric skills and related solidarity-inducing factors
surface elsewhere in the division of labor (Van Maanen and Barley 1984,
p. 302).
Awareness
When studying the subjective aspects of stratification, it is conventional
to consider not merely the social identities through which individuals un-
derstand themselves, but also the cognitive maps with which they perceive
1198
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
5
The results presented by Jackman and Jackman (1983, pp. 2231) suggest that some-
what greater consistency emerges when respondents are asked to sort detailed occupa-
tions into a preexisting aggregate classification.
6
It is often argued that scholars should search long and hard for that underlying
strain of consistency and continuity (Bott 1957, p. 17) running across this great diver-
sity of popular images (see, esp., Vanneman and Cannon 1987). Although the average
class mapping is surely of interest, we ought not forget that individual-level variability
around this average is evidently substantial.
1199
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
7
The results cited here pertain to contemporary Australia and Hungary (Evans et al.
1992, table 2).
1200
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
1201
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
Social Closure
The foregoing subjectivist models were once dominant in sociology (e.g.,
Warner et al. 1949) but have now been largely superseded by class analytic
approaches that rest on such concepts as exclusion, social closure, and
positional property (Brint 1994; Freidson 1994, pp. 8084; Manza 1992;
Murphy 1988; Collins 1979; Stinchcombe 1979; Parkin 1979; Giddens
1973; Lopreato and Hazelrigg 1972; Weber [1922] 1968). In models of this
sort, the emphasis shifts not only to the closure rules by which class incum-
bents protect their privileges, but also to the countervailing actions of
excluded parties that challenge these rules and the inequality that they
legitimate. The two exclusionary devices that closure theorists (esp. Parkin
1979) conventionally emphasize are those of private property and creden-
tialism: the former operates by preventing general access to the means
of production and its fruits, whereas the latter operates by controlling
and monitoring entry to key positions in the division of labor (Parkin
1979, p. 48). While closure theory provides, therefore, a new sociological
language for understanding interclass relations, the actual class mappings
posited by closure theorists ultimately prove to be standard aggregate fare.
For example, Parkin (1979, p. 58) proposes a two-class solution for modern
capitalism, with the exclusionary class comprising those who control pro-
ductive capital and professional services and the subordinate class com-
prising all those who are excluded from these positions of control (also,
Giddens 1973, pp. 10712; Weber 1968, pp. 3027). This formulation can-
not be easily reconciled with the post-Marxist rhetoric that Parkin (1979)
has elsewhere deployed. As Manza (1992, p. 289) notes, Parkin has taken
great pleasure in critiquing Marxian class models that ignore intraclass
fragmentation, yet in the end his own representation of classes is equally
vulnerable to such criticism. If there is indeed any inconsistency here, it
only serves to underline our more general point that aggregate formula-
tions are so dominant in sociology that even their critics cannot help but
deploy them.8
The question that emerges here is whether an aggregate formulation is
a necessary feature of closure theory or merely a superfluous adjunct. The
latter interpretation is arguably more plausible; that is, if closure theory
could somehow be reinvented without the coloration of class analytic con-
vention, its authors would likely emphasize that the real working institu-
tions of closure (i.e., professional associations, craft unions) represent the
interests of occupational incumbents and thus impose barriers to occupa-
8
This characterization of Parkin (1979, pp. 11213) ought not be pushed too hard,
since at some points in the text he emphasizes that intra-class relations [should] be
treated as conflict phenomena of the same general order as inter-class relations, and
not as mere disturbances or complications within a pure class model.
1202
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
tional entry. These local associations also establish and enforce jurisdic-
tional settlements that create task monopolies by preventing other occupa-
tions from providing competing services. By contrast, there are no
analogous organizations that advocate on behalf of an aggregate exclu-
sionary class, nor are there jurisdictional settlements or institutionalized
barriers to entry that are truly aggregate in scope. The closure rules of ad-
vanced capitalism are not, therefore, defined broadly but are rather im-
posed by particularistic associations representing the credential-holders
themselves (Murphy 1988, p. 174).
The long-standing attempt to identify an aggregate exclusionary bour-
geoisie (Parkin 1979) is accordingly misleading. After all, such conceptual
efforts not only misrepresent the analytic level at which exclusionary bar-
riers are drawn, but they also ignore the many pockets of exclusionary
closure that are found within craft, service, and other subordinate occupa-
tions. The unfortunate tendency within closure theory has thus been to
ignore craft unions altogether or to treat them as weak and inconsequen-
tial sources of closure. In justifying this practice, Parkin (1979, p. 57) has
stressed that professional closure is based on state-guaranteed monopolies
of the kind that craft occupations only rarely secure, while Murphy (1988,
p. 188) has further emphasized that unions seek to redress disadvantage
rather than extend or elaborate it. These points of distinction may render
unions a less objectionable form of closure, but from our standpoint what
is most relevant is the social fact of closure itself rather than its standing
relative to some moral compass.
This line of reasoning suggests that closure theorists ought not focus so
exclusively on the relatively rare case of monopolistic guilds. As noted by
Gouldner (1979), the union shop can also be an effective exclusionary tool,
and so too can various forms of expert closure in which control over hiring
and firing is ceded to occupational representatives or incumbents. The
latter form of control is ubiquitous; for example, the professional associa-
tions of the knowledge trades cannot legally prevent academic employ-
ers from hiring uncredentialed academics, but in practice hiring decisions
are controlled by occupational representatives who ensure that employees
have the appropriate training and credentials (Freidson 1994, p. 116).
There is great variability, then, in the methods by which closure is secured
and in the effectiveness of these methods. As is well known, some occupa-
tions have clearly failed to realize their exclusionary objectives (e.g.,
nurses), while others have neither pursued nor even articulated such ob-
jectives. Although many occupations are quite unformed, we would none-
theless emphasize that, insofar as formal rules of closure have been estab-
lished, they are typically implemented at the level of occupations.
This revised version of closure theory is inconsistent with the conven-
tional practice of carrying out aggregate analyses of class mobility. The
1203
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
results from these analyses have been variously interpreted, but it is be-
coming increasingly fashionable to characterize modern exchange as so
fluid, permeable, and inchoate that exclusionary institutions do not seem
to be designed first and foremost to solve the problem of class reproduc-
tion (Parkin 1979, p. 62). In our view, conclusions of the latter sort should
be regarded as premature, since they are based on data that are highly
aggregate and hence likely to conceal exclusionary practices operating at
the occupational level. There are at least two ways in which occupation-
specific credentials protect incumbents from the hazards of the market-
place: they provide tenure by guaranteeing the competence of credential
holders for the duration of their careers, and they require training (e.g.,
advanced schooling) that is so arduous that only workers at the beginning
of their careers will likely undertake the necessary investment. The re-
sulting barriers to inflow and outflow are clearly occupational in nature
and are thereby concealed in mobility analyses pitched exclusively at the
aggregate level. When conventional mobility tables are disaggregated, one
indeed finds that the holding power of many detailed occupations (e.g.,
doctor, lawyer, carpenter) is quite strong, whereas that of aggregate classes
is invariably weak (Srensen and Grusky 1996a; cf. Evans and Laumann
1983).9 If, then, the amount of intragenerational closure revealed in con-
ventional research falls short of what reproduction theorists promised
(e.g., Bourdieu and Passeron 1977), this is not because modern forms of
closure are altogether unreliable but rather because modern analyses mis-
specify the analytic level at which closure is secured.
The case for disaggregation is perhaps less persuasive when the focus
shifts to intergenerational closure and the more complex institutional
forces that underly it. As Parkin (1979, pp. 6066) rightly emphasizes, the
closure rules of credentialing organizations may protect current incum-
bents from the ravages of the market, but they are not so obviously per-
fected for the purpose of transferring privilege from one generation to the
next. The offspring of professional parents, for example, cannot them-
selves become professionals without passing through the same arduous
credentialing regimen that is required of the general population, with all
the uncertainty and potential for failure that this implies. We would obvi-
ously not suggest that parents are powerless in the face of such demand-
side universalism; however, insofar as they can facilitate intergenerational
9
Although the main barriers to mobility are thus found at the disaggregate level,
these barriers might still be insufficiently developed to produce strong individual
commitment to the long-term interests of a profession (Evans and Laumann 1983,
p. 25). In this regard, the forces for closure may well work principally at the disaggre-
gate level, but the resulting restrictions on mobility flows cannot be expected to assume
castelike proportions.
1204
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
Collective Action
In recent years, the aggregate classes identified by Marxian theory have
shown a decided reluctance to act collectively on behalf of their (pre-
sumed) interests, thus leading to considerable neo-Marxian hand-
wringing as well as more constructive neo-Marxian efforts to salvage
class-based models of collective action. The principal scholarly responses
have been (1) to redraw the lines of cleavage along which classes for
themselves will likely form (e.g., Poulantzas 1974), (2) to conceptualize
the relationship between class structure and action as highly probabilistic
and contingent (e.g., Wright 1985, pp. 12324), (3) to emphasize the role
of political movements and parties in activating these contingent interests
(e.g., Korpi 1983), and (4) to discard the class analytic enterprise altogether
by locating the main locus of interest formation outside the realm of pro-
duction (see, e.g., Larana et al. 1994). The recent history of class analytic
scholarship might thus be represented as a series of increasingly overt
challenges to the Marxian orthodoxy on collective action. The minor revi-
sionism of position 1 has in fact been largely abandoned by now, whereas
the more substantial revisionism of positions 2 and 3 continues to claim
adherents (e.g., Goldthorpe and Marshall 1992), and the explicit post-
Marxism of position 4 is obviously in ascendancy.
Unfashionable though it may be, we suggest that position 1 ought not
be rejected outright, nor should it be weakened with escape clauses that
render class action merely contingent or probabilistic. In adopting the for-
mulations of the other three positions, stratification scholars have appar-
1205
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
ently overlooked the simple fact that much collective action flows quite
unproblematically out of structurally defined groupings, albeit only if such
groupings are understood in less aggregate terms than is conventionally
the case. The three principal types of collective action at the level of unit
occupations are downwardly directed closure strategies designed to re-
strict access to occupational positions, lateral competitive struggles be-
tween occupational associations over functional niches in the division of
labor, and upwardly directed collective action oriented toward securing
occupation-specific benefits (e.g., monopoly protection) from the state and
from employers. We have already discussed downward and lateral forms
of collective action in some detail. As for upward forms, the main point
that should be stressed is that interoccupational struggle is clearly the
norm among professionals, so much so that only rarely have their associa-
tions supported one another in efforts to secure state-granted prerogatives
of licensing or credentialing. In fact, Abbott (1988, p. 175) was hard-
pressed to locate a single example of a professional association publicly
defending the licensing efforts of another profession, whereas he found a
great many examples of negative testimony coming from immediately
competing professions. The case for sectionalism is somewhat weaker
within the manual sector; nonetheless, many unions in the United States
are organized at the level of detailed occupations, while cross-occupa-
tional combinations of craft unions (e.g., the American Federation of La-
bor) have been too weak to undertake much in the way of meaningful
classwide action.10 These considerations led Krause (1971, p. 87) to con-
clude long ago that there has historically been more occupation-specific
consciousness and action than cross-occupational combination (also see
Freidson 1994, pp. 7591; Van Maanen and Barley 1984, pp. 33133).
The division of labor is rife, then, with collective action that occurs so
predictably as to eliminate any need to deploy the rhetoric of contingency.
To be sure, much work remains in understanding the conditions under
which occupations engage in collective action, but at least the discipline
can be assured of having sufficient such behavior to explain. We must
concede, of course, that occupation-based approaches lack the tidy closure
of orthodox Marxian models in which all parties to conflict are themselves
classes or their direct representatives. Although collective action of the
lateral variety may approximate closure of this sort, the remaining forms
10
The unions of the Congress of Industrial Organizations (CIO) were originally orga-
nized as aggregates of many occupations. In the end, the rise of CIO unions only
turned inward (Dunlop 1956) the long-standing struggle between occupational and
industrial forms of organization, with sectional interests ultimately reemerging in the
form of skilled trade committees that represented particular occupations within each
CIO union (Weber 1969).
1206
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
11
The recent work of Kohn and Slomczynski (1990) can be read as combining the
foregoing approaches.
1207
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
1208
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
STRUCTURAL ACCOUNTS
The preceding discussion suggests that disaggregate occupations can be
meaningful sociopolitical communities of precisely the sort that class ana-
lysts once sought. However, given that contemporary scholarship has in-
variably defaulted to the aggregate level, such gemeinschaftlich concep-
tions of class have by now been largely abandoned (Holton and Turner
1989), and neo-Marxians of the present day thus adopt the more limited
objective of mapping out structural locations that have the potential to
become sociopolitical communities in the future (see also Holton 1996).
Under this formulation, much attention is conventionally focused on iden-
tifying the underlying axes of exploitation, since these are assumed to con-
stitute the objective bases of antagonistic interests (see Wright 1989) that
12
While some political sociologists (e.g., Nieuwbeerta and De Graaf, n.d.) have sought
to disaggregate conventional classifications, the resulting categories remain extremely
heterogeneous and cannot, therefore, be represented as occupations of a meaningful
sort (also, see Nieuwbeerta 1995).
1209
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
13
The labor theory of value is not without its defenders, but it no longer has litmus-
test status in determining whether a scholar self-identifies as neo-Marxian.
14
The concept of rent has been defined in various ways, but for our purposes these
differences can be safely glossed over. In the discussion that follows, the complicating
effects of organizational exploitation (Wright 1985) will be ignored since we accept
the view of Srensen (1994) that managerial positions are not rent generating and
cannot, therefore, be represented as exploitative.
15
This test should of course be based on the present value of cumulated lifetime earn-
ings.
1210
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
16
We are ignoring here the inequality that arises by virtue of effort, native ability,
and compensating differentials.
1211
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
17
The licensing of occupations has become strictly regulated in some states, yet such
legislative reform has so far addressed the conditions under which licenses may be
obtained rather than the legitimacy of closure itself.
1212
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
Sociotechnical Change
The debates within the sociotechnical domain have in the past unfolded
in highly stylized fashion, with Marxians and neo-Marxians arguing that
industrialism promotes classwide solidarity by gradually obliterating all
distinctions of labor (Marx 1964, p. 480), while postindustrial theorists
counter that ongoing sociotechnical change only serves to further differen-
tiate the working class into diverse occupations that differ in skill, life
chances, and sociopolitical interests (Aronowitz and DiFazio 1994; Gallie
1991; DiPrete 1988; Spenner 1995, 1985, 1983; Braverman 1974; Dahren-
dorf 1959). In classical versions of postindustrial theory, the emerging cat-
egories of highly skilled labor (e.g., professional, technical, service) are
seen as resistant to deskilling because (a) the tasks generated under postin-
dustrialism are intrinsically complex and difficult to differentiate and (b)
the division of labor within these emerging sectors is occupationally orga-
nized and therefore protected from most forms of managerial rationaliza-
tion. For example, Ellul (1967, p. 162) writes that technique as a general
phenomenon always gives rise to an aristocracy of technicians, while
Bennis and Slater (1968, p. 66) refer to bands of specialists held together
by the illusion of a unique identity, and Bell (1973, p. 145) observes that
the organization of professionals will be a major feature of post-industrial
society. The resulting postindustrial story thus suggests, if only implicitly,
a Durkheimian solidarity of specialty that emerges even as classwide
forms of solidarity weaken or wither away. Against this intellectual back-
drop, the postmodernist account may be seen as simultaneously embrac-
ing the postindustrial view that classwide forms of solidarity are unviable,
while also rejecting the corollary claim that persisting local solidarities
account, at least in part, for the breakdown of such aggregate solidarities.
In this sense, postmodernists unfairly emphasize the purely destructive
effects of skill differentiation, thus making it possible to represent the
realm of production as increasingly formless and unstructured (see Za-
1213
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
busky and Barley 1996; Friedson 1994, pp. 99102; Van Maanen and Bar-
ley 1984).
The current fashion is to approach these long-standing debates about
task deskilling from a post-Fordist perspective (Piore and Sabel 1984). As
Amin (1994) notes, the post-Fordists suggest that early industrialism did
indeed promote much craft deskilling, yet this process is alleged to be
reversing itself as Fordist factories are gradually supplanted by small-
scale production, flexible specialization, and a rejuvenated artisanal sec-
tor, all of which serve to reintroduce those distinctions of manual labor
that Marx promised would disappear (also, see Hirst and Zeitlin 1991).
The foregoing account may be seen, then, as a freshened form of postin-
dustrialism in which the forces of upgrading and reskilling are presumed
to play out not merely in the professional, technical, and service categories
but in the craft sector as well. In this context, one might expect postmod-
ernists to view post-Fordism with some antipathy, yet in fact these two
accounts are often conflated in the literature. For example, the new times
post-Fordism of Hall (1988) and his colleagues (e.g., Hall and Jacques
1989) becomes virtually indistinguishable from conventional postmodern-
ism, since it emphasizes that sociotechnical changes weaken aggregate sol-
idarities and generate a new stratification order based on lifestyle, taste,
and culture rather than categories of social class (Hall 1988, p. 24). This
account rests on the characteristic postmodernist assumption that an in-
creasingly fragmented productive realm (Hall 1988, p. 24) necessarily
weakens all forms of solidarity within the division of labor.
While the pathbreaking work of Piore and Sabel (1984) clearly has
merit, we would necessarily take issue with these more elaborated ac-
counts that attempt to smuggle in poststructuralism under a post-Fordist
banner. If one accepts the core post-Fordist claim that flexible specializa-
tion breathes new life into artisanal production (e.g., Piore and Sabel
1984), the appropriate implication is not that all production-based solidar-
ities shall wither away but rather that such solidarities are increasingly
localistic. In any standard post-Fordist account, the new and emerging
forms of craft production are assumed to require worker solidarity and
communitarianism (Piore and Sabel 1984, p. 278), and the rejuvenated
artisanal sector therefore brings early industrial craft communities into
the 21st century on the basis of a newly decentralized production process
(Aronowitz and DiFazio 1994, p. 98). The end result, then, is a manifestly
prostructuralist account whereby modern craft workers are increasingly
bound to an often familial community [that] promotes both greater con-
trol and a sense of belonging (Aronowitz and DiFazio 1994, p. 97). We
do not wish to suggest here that a post-Fordist model should be adopted
wholesale (see, e.g., Tomaney 1994); however, insofar as this model now
1214
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
18
In some versions of post-Fordism, the local communities so created are presumed
to cut across purely occupational boundaries, and the resulting account therefore be-
gins to diverge from our own. We address this discrepant variant of post-Fordism in
the following section.
1215
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
Organizational Change
As sociologists so frequently point out, the division of labor is not mechan-
ically determined by purely technical requisites, but rather it reflects the
organizational environment within which work is designed and completed
(e.g., Edwards 1979). The rise of industrialism in the 18th century can be
attributed, for example, to the spread of vertical strategies of coordination
that fragmented tasks into increasingly simple jobs and thus rendered
them amenable to purely administrative or bureaucratic oversight (Weber
1968). By contrast, preindustrial craft workers defined and organized the
production process themselves, and the division of labor was accordingly
controlled by self-regulating occupational experts rather than organiza-
tionally empowered administrators (Zabusky and Barley 1996, pp. 188
92; Barley 1995; Freidson 1994, pp. 1034; Abbott 1989). The obvious
question that then arises is whether vertical methods of control will con-
tinue to encroach on occupationally defined labor as postindustrialism
evolves.
This question cannot be as easily answered as some postmodernists
seemingly suggest. In fact, one can identify incipient organizational theo-
ries on either side of this debate, with the contemporary literature thus
encompassing both postoccupational theories describing the gradual with-
ering away of functionally defined positions and revisionist theories sug-
gesting that a new occupationally oriented logic of production is on the
rise. The former literature, which is clearly dominant, rests on the claim
that contemporary organizations are relying increasingly on teamwork,
cross-training, and multiactivity jobs that break down conventional skill-
based distinctions (Casey 1995; Baron 1994; Drucker 1993; Peters 1992,
1991; Kanter 1992, 1989; Kilman and Covin 1988; Atkinson and Meager
1986; cf. Pollert 1991).19 These new polyvalent jobs are created either by
19
In some cases, such skill fusion may ultimately serve to strengthen occupational
structuration (rather than weaken it), if only because the newly created jobs may
combine formerly distinct jobs within an existing occupation and thus eliminate in-
traoccupational divisions and conflict. This outcome is clearly more likely when occu-
pational associations are strong enough to successfully defend existing jurisdictional
settlements.
1216
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
1217
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
Associational Change
The final institutions of interest to us are the various intermediary associa-
tions (e.g., trade unions, professional associations) that characteristically
develop at the site of production. Within the Marxian framework, the
long-standing concern has been that trade union consciousness is intrinsi-
cally sectional, thus requiring intellectuals and party functionaries to carry
out supplementary ideological work that presumably cultivates more en-
compassing class-based interests (esp. Lenin 1927). This Marxian concern
appears now to have been well founded. If the history of guilds, unions,
and related production-based associations is reevaluated from the long
view, it is evident that true classwide organization emerged for only a
brief historical moment and that postmodern forms are reverting back
to localism and sectionalism. The widely documented difficulties facing
contemporary unions should be interpreted accordingly; namely, despite
an evident weakening in the encompassiveness of union movements
(Visser 1988, p. 167), there is much evidence suggesting that purely local
unions and associations have by no means lost their hold over workers
(e.g., the American Federation of Teachers). In many countries, central-
ized bargaining between national unions and employers is indeed on the
decline, yet decentralized negotiations have taken their place as instru-
mental collectivism, based on sectional self-interest, becomes the order of
the day (Marshall et al. 1988, p. 7; Hancke 1993; Hyman 1991; Crouch
1990). This interpretation, if borne out, does not speak to destructuration
per se but rather to increasing disaggregation and differentiation of associ-
ational forms.
The professional sector has given rise to organizational forms that are
yet more localistic. As Parkin (1979) points out, professionals eschew all
types of interoccupational confederation, whereas they typically seek out
sectional associations that can defend jurisdictional claims and thereby
protect against incursions by neighboring occupations. In assessing the
future of professionalization, one must consider (a) the power of the tradi-
tional professions (e.g., law, medicine) to resist state and corporate en-
croachment on their jurisdictional claims, (b) the effects of economic and
other demand-side factors on the size of the professional labor force, and
(c) the prospects for full or partial professionalization of new and emerging
technical occupations. We cannot review the massive literature on (a) in
any detail; however, it bears emphasizing that theories of deprofessionali-
zation, while currently in vogue (e.g., Krause 1996), are by no means uni-
1218
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
versally accepted in the field (see, esp., Freidson 1994). Moreover, even if
one concedes that traditional professions are no longer powerful guilds
(Krause 1996), this development is counterbalanced by increases in the
size of the professional sector and the consequent growth in the number
of workers who find themselves in classlike groupings. Without predicting
the professionalization of everyone (see Wilensky 1964), one might also
anticipate long-term growth in the number of occupations that achieve
partial professionalization, since new closure movements can be readily
and profitably built around some of the abstract skills emerging in the
labor market. The foregoing forces imply a simple extrapolation of past
trends toward more licensing, registration, and certification of occupations
(Freidson 1994, p. 120; cf. Marsden 1986). As this process unfolds, opposi-
tional movements may possibly emerge and stall further closure projects,
yet there is relatively little in the contemporary political arena that might
now be interpreted as incipient antiprofessionalism. This conclusion
serves, then, to emphasize our larger point that contemporary commenta-
tors have often overlooked the many forces for structuration operating at
lower analytic levels.
We therefore suggest that the future of local solidarities is more ambigu-
ous than standard poststructuralist formulas allow. In modern academia,
scholars are of course rewarded for waging argument through selective
and coherent accounts, yet we have resisted here all temptation to advance
a positive revisionist argument. Instead, we have merely emphasized that,
because of the great many competing forces at work, the discipline should
withhold judgment on matters of trend until further evidence accumu-
lates. The following section outlines a new research agenda that will hope-
fully reopen the debate on empirical grounds.
CONCLUSIONS
In his much-celebrated preface to The Division of Labor in Society, Dur-
kheim (1933, p. 28) suggested that occupational associations would gradu-
ally become intercalated between the state and the individual, thus pro-
viding an organizational counterbalance to the threat of class formation
on one hand and that of state tyranny on the other. Although this line of
argumentation is frequently rehearsed by scholars of Durkheim, it has
never been treated as a credible developmental model by conventional
class analysts. This is not to suggest that the forces of occupational differ-
entiation and association have gone altogether unappreciated; however,
rather than being represented as important sources of structuration in
their own right, they are typically invoked for the negative purpose of
refuting Marxian and neo-Marxian class models (e.g., Dahrendorf 1959,
pp. 4851). In this regard, we would note that the Marxian project became
1219
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
so influential in the academic world that virtually all scholars found them-
selves obliged to contend with it, if only for purposes of debate or refuta-
tion (see Goldthorpe [n.d.] for related comments). As the Marxian project
falls out of favor, it is typically assumed that the discipline has no alterna-
tive but to resort to some sort of Weberianism or poststructuralism, nei-
ther of which has typically paid sufficient attention to occupation-level
structuration. The purpose of the present article has been to outline a
quasi-Durkheimian third road that might have flourished had the contem-
porary academic world not so completely embraced Marx and the various
Marxisms that he inspired. While our approach provides, then, a structur-
alist alternative to postmodernism, it nonetheless leaves open the possibil-
ity that contemporary identities are shaped by an ever-widening range of
postmodern sources (e.g., gender), some of which constitute important
bases of solidarity in their own right. It is surely worthwhile to study
these postmodern forms of structuration, but unit occupations rightly take
center stage insofar as one wishes to map structuration at the site of pro-
duction.20
In laying out this case for disaggregation, we have largely ignored cross-
national variability in local structuration, but not because we believe such
variability to be either trivial or inconsequential. To the contrary, we sus-
pect that convergence theories (e.g., Erikson and Goldthorpe 1992) may
be rather less appealing when disaggregate analyses are attempted, since
national idiosyncrasies are necessarily concealed through the abstracting
and aggregating operations of class analysis. The case of Germany, for
example, provides a revealing example of the extent to which local institu-
tional forms can support and sustain disaggregate structuration. As class
analysts have long stressed, Germany has a well-developed system of vo-
cational training and apprenticeship, both of which serve to encourage
occupation-specific investments and promote professional commitment
and craftsmanship (e.g., Blossfeld 1992; Kappelhoff and Teckenberg 1987;
Carroll and Mayer 1986; also, see Krause 1996, chap. 6). In systems of
this sort, workers must invest in a single trade early in their careers, and
the correspondingly high costs of retraining produce relatively closed oc-
cupational groupings. If the German system reveals, then, the limits of
disaggregate structuration, the case of Japan conversely reveals the extent
to which such structuration can be institutionally suppressed. The stan-
20
We have largely ignored the complicating effects of property ownership (but see the
section on collective action). In justifying this omission, we would follow the post-
Marxian convention (e.g., Parsons 1970) in suggesting that the interests and lifestyles
of the petty bourgeoisie are much affected by cross-cutting occupational ties and that
the latent function of professional associations is to weaken intraoccupational cleav-
ages stemming from differences in employment status (Durkheim 1933).
1220
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
21
In its traditional form, the Japanese personnel system ties wages and promotions to
seniority (nenko-joretsu seido), thereby reducing the incentive for interfirm mobility.
1221
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
1222
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
Rytina 1993, 1992). Moreover, given that much of the cross-national vari-
ability in local structuration is concealed through aggregation, we may
find that standard convergence hypotheses (e.g., Featherman, Jones, and
Hauser 1975) are no longer sustainable once mobility data are properly
disaggregated.
Are background effects adequately controlled with conventional class or
socioeconomic measures?The merits of competing scales and aggrega-
tions are of course frequently debated (e.g., Grusky and Van Rompaey
1992); however, all such debate presumes that unit occupations are not
directly usable and must be subjected to some sort of scaling or aggrega-
tion, if only to render the analysis tractable. It is high time to put this
assumption to test by exploring whether disaggregation changes the esti-
mated returns to schooling, the effects of racial discrimination, and the
great many other sociological coefficients that likewise depend crucially
on the assumption that background effects are fully purged.
How are attitudes, lifestyles, and dispositions shaped at the site of pro-
duction?In most cases, the burgeoning literature on such matters pro-
ceeds either by aggregating or scaling occupations (e.g., Kohn and
Schooler 1983), and the local subcultures that are generated at the unit
occupational level are thereby muted and obscured. If these subcultures
are identified properly, the correspondence between objective conditions
and lifestyles may be sufficiently strengthened to cast doubt on the post-
modernist retreat from production-based groupings.
We are thus suggesting nothing short of a complete remapping of the
stratification system and its consequences. If our institutionalist account
is on the mark, such a remapping will not lead to an infinite regress in
which ever-increasing disaggregation, no matter how extreme, proves in
the end to be sociologically necessary. Instead, the returns to disaggrega-
tion will at some point likely diminish, since the inveterate splitter will
eventually disaggregate beyond the social boundaries that are institution-
alized in the labor market and defended by closure-generating associa-
tions.22 The resulting categories nonetheless take on gemeinschaftlich form
in only some sectors of the occupational structure; and, consequently, the
strong model of class holds partially and may not necessarily diffuse any
more widely in the future, the claims of Durkheim (1933) notwithstanding.
The practical task of operationalizing our approach is further complicated
22
There are any number of large-scale surveys (e.g., Current Population Survey, Occu-
pational Changes in a Generation) that might be used to assess the costs of aggregation
in terms of explanatory power forgone. Although such an approach obviously pre-
cludes disaggregating beyond the occupational detail available in conventional classi-
fications, it nonetheless allows for far more detailed analyses than have been carried
out to date.
1223
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
insofar as the preferred level of aggregation differs across the various ana-
lytic criteria that are conventionally used to define classes (e.g., social clo-
sure, exploitation, identification). Although we have shown that, regard-
less of the criterion invoked, aggregate class categories are difficult to
defend, we have not sought to demonstrate that all class analytic criteria
imply disaggregation of exactly the same kind or extent. Indeed, the class
analytic conceit that the same classes are operative in all contexts is
suspiciously convenient, and the field would therefore do well to examine
this premise empirically rather than take it on assumption. If a single
disaggregate classification does prove adequate to the task, this result
would be consistent with our argument, as advanced above, that social
closure is a demographic precondition for exploitation and other forms of
classlike behavior.
We are, finally, advocating class models that take contemporary pat-
terns of collective action explicitly into account. In its strong form version,
the classification exercise should involve more than simply aggregating
technically similar occupations, since proceeding as such fails to acknowl-
edge the organizational barriers and combinations (e.g., unions, associa-
tions) that serve to convert technical distinctions into more nearly social
ones. By contrast, neo-Marxians are typically content to identify the struc-
ture of classes in themselves, thus obliging them to resort to ever more
ingenious accounts of the continuous and wholesale discrepancies be-
tween class position and class behavior (Parkin 1979, p. 113). Under our
alternative approach, we acknowledge from the start that occupations are
socially constructed and define class categories accordingly, thereby cir-
cumventing the problems that are created by the neo-Marxian divorce of
structure from action. This third road provides, then, needed relief from
the extreme forms of postmodernism that are now so much in fashion, if
only for want of a plausible sociological alternative.
REFERENCES
Abbott, Andrew. 1988. The System of Professions: An Essay on the Division of Expert
Labor. Chicago: University of Chicago Press.
. 1989. The New Occupational Structure: What Are the Questions? Work
and Occupations 16:27391.
Abercrombie, Nicholas, and John Urry. 1983. Capital, Labour, and the Middle Classes.
London: Allen & Unwin.
Althusser, Louis. 1969. For Marx. London: Verso.
Amin, Ash. 1994. Post-Fordism: Models, Fantasies, and Phantoms of Transition.
Pp. 139 in Post-Fordism: A Reader, edited by Ash Amin. Oxford: Blackwell.
Ardagh, John. 1988. Germany and the Germans. London: Penguin.
Aronowitz, Stanley. 1992. The Politics of Identity: Class, Culture, Social Movements.
New York: Routledge.
Aronowitz, Stanley, and William DiFazio. 1994. The Jobless Future. Minneapolis:
University of Minnesota Press.
1224
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
1225
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
Calhoun, Craig. 1994. Social Theory and the Politics of Identity. Pp. 936 in Social
Theory and the Politics of Identity, edited by Craig Calhoun. Oxford: Black-
well.
Caplow, Theodore. 1954. The Sociology of Work. Minneapolis: University of Minne-
sota Press.
Carroll, Glenn R., and Karl U. Mayer. 1986. Job-Shift Patterns in the FRG: The
Effects of Social Class, Industrial Sector, and Organizational Size. American Socio-
logical Review 51:32341.
Casey, Catherine. 1995. Work, Self, and Society. London: Routledge.
Clark, Terry N. 1996. The Debate over Are Social Classes Dying? 1996. Working
paper. Conference on Social Class and Politics, Woodrow Wilson Center, Washing-
ton, D.C.
Clark, Terry N., and Seymour M. Lipset. 1991. Are Social Classes Dying? Interna-
tional Sociology 6:397410.
Clement, Wallace, and John Myles. 1994. Relations of Ruling: Class and Gender in
Postindustrial Societies. Montreal: McGillQueens University Press.
Cole, Robert E. 1979. Work, Mobility, and Participation. Berkeley and Los Angeles:
University of California Press.
Coleman, Richard P., and Lee Rainwater with Kent A. McClelland. 1979. Social
Standing in America. London: Routledge & Kegan Paul.
Collins, Randall. 1979. The Credential Society: An Historical Sociology of Education
and Stratification. New York: Academic Press.
Conk, Margo A. 1978. Occupational Classification in the United States Census: 1870
1940. Journal of Interdisciplinary History 9:11130.
Coxon, Anthony P. M., and Charles L. Jones. 1979a. Class and Hierarchy: The Social
Meaning of Occupations. New York: St. Martins.
. 1979b. Measurements and Meanings: Techniques and Methods of Studying
Occupational Cognition. New York: St. Martins.
Crompton, Rosemary. 1993. Class and Stratification: An Introduction to Current De-
bates. Cambridge: Polity.
Crouch, Colin. 1990. Afterword to European Industrial Relations, edited by Guido
Baglioni and Colin Crouch. London: Sage.
Dahrendorf, Ralf. 1959. Class and Class Conflict in Industrial Society. Stanford,
Calif.: Stanford University Press.
. 1986. Society and Democracy in Germany. London: Weidenfeld &
Nicolson.
Davis, James A. 1982. Achievement Variables and Class Cultures: Family, Schooling,
Job, and Forty-Nine Dependent Variables in the Cumulative GSS. American So-
ciological Review 47:56986.
De Soto, Clinton, and Frank Albrecht. 1968. Cognition and Social Orderings.
Pp. 50411 in Theories of Cognitive Consistency: A Sourcebook, edited by Rob-
ert P. Abelson, Elliot Aronson, William J. McGuire, Theodore M. Newcomb, Mil-
ton J. Rosenberg, and Percy H. Tannenbaum. Chicago: Rand McNally.
Devine, Fiona. 1997. Social Class in America and Britain. Edinburgh: Edinburgh
University Press.
DiPrete, Thomas A. 1988. The Upgrading and Downgrading of Occupations: Status
Redefinition versus Deskilling as Alternative Theories of Change. Social Forces
66:72546.
Dore, Ronald P. 1973. British FactoryJapanese Factory. London: Allen & Unwin.
Drucker, Peter. 1993. Post-Capitalist Society. New York: Harper.
Duncan, Otis D. 1961. A Socioeconomic Scale for All Occupations. Pp. 10938
in Occupations and Social Status, edited by Albert J. Reiss, Jr. New York: Free
Press.
Dunleavy, Patrick. 1980. The Political Implications of Sectoral Cleavages and the
1226
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
1227
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
1228
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
Holton, Robert J. 1996. Has Class Analysis a Future? Pp. 2641 in Conflicts about
Class: Debating Inequality in Late Industrialism, edited by David J. Lee and Bryan
S. Turner. London: Longman.
Holton, Robert J., and Bryan S. Turner. 1989. Max Weber on Economy and Society.
London: Routledge & Kegan Paul.
Hout, Michael, Clem Brooks, and Jeff Manza. 1993. The Persistence of Classes in
Post-Industrial Societies. International Sociology 8:25977.
Hughes, Everett C. 1958. Men and Their Work. New York: Free Press.
. 1971. The Sociological Eye: Selected Papers. Chicago: Aldine-Atherton.
Hyman, Richard. 1991. European Unions: Towards 2000. Work, Employment, and
Society 5:62139.
Illich, Ivan. 1980. Toward a History of Needs. New York: Bantam Books.
Inglehart, Ronald. 1977. The Silent Revolution: Changing Values and Political Styles
among Western Publics. Princeton, N.J.: Princeton University Press.
. 1990. Culture Shift in Advanced Industrial Society. Princeton, N.J.: Princeton
University Press.
Ishida, Hiroshi. 1993. Social Mobility in Contemporary Japan. Stanford, Calif.: Stan-
ford University Press.
Jackman, Mary R. 1994. The Velvet Glove: Paternalism and Conflict in Gender, Class,
and Race Relations. Berkeley and Los Angeles: University of California Press.
Jackman, Mary R., and Robert W. Jackman. 1983. Class Awareness in the United
States. Berkeley and Los Angeles: University of California Press.
Jackson, J. A. 1970. Professions and Professionalization. Cambridge: Cambridge Uni-
versity Press.
Jackson, Robert Max. 1984. The Formation of Craft Labor Markets. New York: Aca-
demic Press.
Johnston, Russell, and Paul R. Lawrence. 1988. Beyond Vertical Integrationthe
Rise of the Value-Adding Partnership. Harvard Business Review 88:94101.
Joyce, Patrick. 1995. Class. Oxford: Oxford University Press.
Kahl, Joseph A., and James A. Davis. 1955. A Comparison of Indexes of Socio-
Economic Status. American Sociological Review 20:31725.
Kalmijn, Matthijs. 1994. Assortative Mating by Cultural and Economic Occupa-
tional Status. American Journal of Sociology 100:42252.
Kanter, Rosabeth M. 1989. When Giants Learn to Dance: Mastering the Challenge
of Strategy, Management, and Careers in the 1990s. New York: Simon & Schuster.
. 1992. The Challenge of Organizational Change. New York: Free Press.
Kappelhoff, Peter, and Wolfgang Teckenberg. 1987. Intergenerational Mobility and
Career Mobility in the Federal Republic and the United States. Comparative Stud-
ies of Social Structure 17:349.
Kilman, Ralph, and Theresa Covin. 1988. Corporate Transformations: Revitalizing
Organizations for a Competitive World. San Francisco: Jossey Bass.
Kingston, Paul W. 1994. Are There Classes in the United States? Pp. 341 in Re-
search and Social Stratification and Mobility, vol. 13. Edited by Robert Althauser
and Michael Wallace. Greenwich, Conn.: JAI Press.
Kluegel, James R., and Eliot R. Smith. 1986. Beliefs about Inequality. New York:
Aldine De Gruyter.
Kocka, Jurgen. 1981. Class Formation, Interest Articulation, and Public Policy: The
Origins of the German White-Collar Class in the Late Nineteenth and Earlier
Twentieth Centuries. Pp. 6381 in Organizing Interests in Western Europe, edited
by Suzanne Berger. Cambridge: Cambridge University Press.
Kohn, Melvin L., and Carmi Schooler with Joanne Miller, Karen A. Miller, Carrie
Schoenbach, and Ronald Schoenberg. 1983. Work and Personality: An Inquiry into
the Impact of Social Stratification. Norwood, N.J.: Ablex.
1229
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
Kohn, Melvin L., and Kazimierz M. Slomczynski. 1990. Social Structure and Self-
Direction. Oxford: Blackwell.
Korpi, Walter. 1978. The Working Class in Welfare Capitalism: Work, Unions, and
Politics in Sweden. London: Routledge & Kegan Paul.
. 1983. The Democratic Class Struggle. London: Routledge.
Kraus, Vered, E. O. Schild, and Robert W. Hodge. 1978. Occupational Prestige in
the Collective Conscience. Social Forces 56:900918.
Krause, Elliott A. 1971. The Sociology of Occupations. Boston: Little, Brown.
. 1996. Death of the Guilds: Professions, States, and the Advance of Capitalism,
1930 to the Present. New Haven, Conn.: Yale University Press.
Laclau, Ernesto, and Chantal Mouffe. 1985. Hegemony and Socialist Strategy: To-
wards a Radical Democratic Politics. London: Verso.
Ladd, Everett Carll, Jr. 1979. Pursuing the New Class: Social Theory and Survey
Data. Pp. 10122 in The New Class? edited by B. Bruce-Briggs. New Brunswick,
N.J.: Transaction.
Lamont, Michele. 1992. Money, Morals, and Manners. Chicago: University of Chicago
Press.
Larana, Enrique, Hank Johnston, and Joseph R. Gusfield. 1994. New Social Move-
ments: From Ideology to Identity. Philadelphia: Temple University Press.
Larson, Magali Safartti. 1977. The Rise of Professionalism. Berkeley and Los Angeles:
University of California Press.
Lawrence, Paul R., and Davis Dyer. 1983. Renewing American Industry. New York:
Free Press.
Lee, David J. 1996. Weak Class Theories or Strong Sociology? Pp. 24553 in Con-
flicts about Class: Debating Inequality in Late Industrialism, edited by David J.
Lee and Bryan S. Turner. London: Longman.
Lee, David J., and Bryan S. Turner, eds. 1996. Conflicts about Class: Debating In-
equality in Late Industrialism. London: Longman.
Lenin, Vladimir I. 1927. Collected Works of V. I. Lenin. New York: International.
Lipset, Seymour M. 1994. The State of American Sociology. Sociological Forum 9:
199220.
. 1996. American Exceptionalism: A Double-Edged Sword. New York: Norton.
Lopreato, Joseph, and Lawrence E. Hazelrigg. 1972. Class, Conflict, and Mobility:
Theories and Studies of Class Structure. San Francisco: Chandler.
Lukes, Steven. 1984. The Future of British Socialism? Pp. 26983 in Fabian Essays
in Socialist Thought, edited by Ben Pimlott. London: Heinemann.
Manza, Jeff. 1992. Classes, Status Groups, and Social Closure: A Critique of Neo-
Weberian Social Theory. Pp. 275302 in Current Perspectives in Social Theory,
edited by Ben Agger. Greenwich, Conn.: JAI Press.
Manza, Jeff, and Clem Brooks. 1996. Does Class Analysis Have Anything to Contrib-
ute to the Study of Politics? Theory and Society 25:71724.
Marsden, David. 1986. The End of Economic Man? Custom and Competition in Labor
Markets. New York: St. Martins Press.
Marshall, Gordon. 1997. Repositioning Class: Social Inequality in Industrial Socie-
ties. London: Sage.
Marshall, Gordon, David Rose, Howard Newby, and Carolyn Vogler. 1988. Social
Class in Modern Britain. London: Unwin Hyman.
Marx, Karl. 1948. Manifesto of the Communist Party. New York: International Pub-
lishers.
. 1963. The Poverty of Philosophy. New York: International Publishers.
. 1964. Selected Works, vol. 1. Moscow: Progress Publishers.
Melucci, Alberto. 1980. The New Social Movements: A Theoretical Approach. So-
cial Science Information 19:199226.
Meyer, John W. 1994. The Evolution of Modern Stratification Systems. Pp. 730
1230
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
1231
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
1232
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
Class Analysis
1233
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms
American Journal of Sociology
1234
This content downloaded from 128.227.155.59 on Tue, 31 Oct 2017 06:25:42 UTC
All use subject to http://about.jstor.org/terms