Вы находитесь на странице: 1из 16

Who Were the Yugoslavs?

Failed Sources of a Common Identity in the Former Yugoslavia


Author(s): Dusko Sekulic, Garth Massey, Randy Hodson
Source: American Sociological Review, Vol. 59, No. 1, (Feb., 1994), pp. 83-97
Published by: American Sociological Association
Stable URL: http://www.jstor.org/stable/2096134
Accessed: 07/07/2008 18:43

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=asa.

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 organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

http://www.jstor.org
WHO WERE THE YUGOSLAVS?
FAILED SOURCES OF A COMMON IDENTITY IN
THE FORMER YUGOSLAVIA*

DUSKOSEKULIC GARTHMASSEY
George Mason University Universityof Wyoming
RANDYHODSON
Indiana University

Yugoslavia's leaders believed that a policy of equality among the many nationalities in
Yugoslavia, in tandem with CommunistParty hegemony, would allow nationalism within
Yugoslavia to exist, mature, andfinally diminish as a political force withoutjeopardizing
the political stability and economic development of the country as a whole. Consequently
the identification of people with their nationality was accepted to the neglect of an iden-
tity associated with the state as a whole. The expectation that a shared political agenda
and the modernization of the society would weaken nationalism as a political force was
not met. Instead, economic and political rivalries among the Yugoslav republics intensi-
fied nationalist feelings. In the early 1990s Yugoslavia's experiment in building a multi-
national state was replaced with open hostilities and warfare among the South Slavs. We
identifyfour routes to Yugoslav self-identification and analyze the significance of these
using survey data from 1985 and 1989, just prior to the break up of Yugoslavia. Urban
residents, the young, those from nationally-mixed parentage, Communist Party mem-
bers, and persons from minority nationalities in their republic were among those most
likely to identify as Yugoslavs. None of these factors, however, proved sufficient to over-
ride the centrifugal forces of rising nationalism. Implications for political integration in
Eastern Europe and the former Soviet Union are discussed.

T he dreamof nineteenth-century
national- and Warwick 1983; Collins and Waller 1992;
ism was thata common language,history, Enloe 1973; Nielsen 1985). In addition,indus-
and habits of everydaylife justified the forma- trialization and the establishment of modern
tion of distinct nation states that would repre- mass institutions,such as education and mass
sent and protect ethnically homogeneous communications, have been emphasized by
groups(Gertz 1973;Isaacs 1975;Rokkan1975; state builders,in partwith the expectationthat
Smith 1986). In the twentieth century states these would erode the differencesupon which
have developed a variety of political mecha- national identities were based (Davis 1978;
nisms to accommodatenationaldiversitywithin Hodson, Sekulic, and Massey forthcoming).In
their borders, including legal recognition of some states, such as France, nationalpolicies,
minoritynationalities,proportionateseating in industrialization, and mass institutions pro-
legislatures,andpolicies favoringeconomically duced integrated national identities (Tilly
disadvantagednationalities and areas (Cohen 1975). In other cases the integrative conse-
*Directall correspondence quencesof these processes have been slower to
to DuskoSekulic,In- develop; and in others, these processes may
ternationalInstitute,4001 N. FairfaxDrive,Suite
have exacerbatedratherthan eased nationalri-
450, George Mason University,Arlington,VA
22203.This researchwas supportedin partby the valrieswithinstates(BelangerandPinard1991;
NationalScience Foundationthroughthe United Hechter 1976; Olzak 1992; Tudjman 1981).
States-YugoslaviaJoint Fundfor Scientific and Now, at the end of the twentiethcentury,a re-
TechnologicalCooperation. We thankYorkBrad- surgence of claims, rivalries, and conflicts
shaw,BurkeGrandjean, CharlesJelavich,Dorothy among national groups within states threaten
Watson,and threeanonymousASRreviewersfor the state-buildingaccomplishmentsof past de-
theircommentson previousdrafts. cades in many partsof the world.

AmericanSociological Review, 1994, Vol. 59 (February:83-97) 83


84 AMERICANSOCIOLOGICALREVIEW

To establish a new, ethnically diverse state it fered no particularnational identity (Petrovic


is seen as critical that the people adopt a com- 1983). Table 1 presents the percentagesof the
mon identity as citizens of that state, as mem- population of Yugoslavia identifying as
bers of a unified political system composed of Yugoslav in 1961, 1971, and 1981.
groupsotherwisediverse in language,religion, The modest decline in self-identification as
customs, ethnicity,or historicalexperience. In a Yugoslav for the countryas a whole between
such states,however,thereis a tensionbetween 1961 and 1971 was primarily the result of a
ethnic or nationalgroups wanting to maintain decline in Yugoslav identifiers in Bosnia-
their own sovereignty within the state and the Hercegovina (hereafterreferredto as Bosnia)
state's need to integratesuch groups into a co- in 1971. According to Ramet (1984:144-49),
hesive political unit that represents a shared high Yugoslav self-identificationin Bosnia in
ideology and vision of the future (Kedouire 1961 occurred because Moslems refused to
1960; Kohn 1961; Smith 1986). Symptomatic identify themselves with dominant national
of this tension is the struggleover identity,with groups (i.e., Serbs or Croats).The 1971 census
the particularnationalgroupstryingto preserve was the first to allow "Moslem"as a national-
their identities and the state attemptingto im- ity, and many Moslem Bosnians switchedfrom
pose a new identity assimilating and subsum- the "Yugoslav"to the "Moslem" category in
ing the more particularidentities. 1971. Moslems in Yugoslavia,many of whom
The formerYugoslaviais an instance where live in Bosnia, are Slavs who adoptednot only
the integrative processes of state identify for- the Islamicfaith,but also embracedmany other
mation have failed. Many of the underpinnings cultural and linguistic features of the Turkish
necessaryto integrativeprocesses were present people who, duringthe period of the Ottoman
in Yugoslavia, but they proved insufficient in Empire, controlled much of what would later
the face of economic downturnsand resurgent become Yugoslavia.
nationalist forces. We identify four potential Apart from Bosnia between 1961 and 1971
routes that led people in the formerYugoslavia and Kosovo, self-identification as a Yugoslav
to identify themselves as members of a multi- shows a generalpatternof increase from 1961
nationalstate ratherthan as membersof a spe- to 1981, especially in Croatia and the Vojvo-
cific nationality.We use data from two major dine, and in Bosnia between 1971 and 1981.
social surveys done in Yugoslaviain 1985 and Among the republics and provinces, Kosovo
1989 (just prior to the dissolution of the state) showed the lowest Yugoslav identification in
to examine the forces that facilitatedor under- 1971 and 1981, with most people in Kosovo
mined the emergence of a shared Yugoslav identifying themselves as either Albanian or
identity. Serbian.One factorthatencouragedidentifica-
tion as a Yugoslavwas the heterogeneityof na-
ROUTES TO YUGOSLAVSELF- tionalities within the republic or province
IDENTIFICATION (Breuilly 1982; Djilas 1991). Contactbetween
differentnationalitiesis increasedin heteroge-
The firstYugoslavcensus afterWorldWarI did neous settings leading to greatersocial mixing
not record nationality, although the official and intermarriage(Isaacs 1975). Bosnia and
name of the newly formed state was "The the Vojvodine were the former Yugoslavia's
Kingdom of the Serbs, Croats,and Slovenes," most nationallydiverse areas, and both repub-
clearly indicating its multinationalcharacter. lics had larger than average proportions of
This promptedhistoriansto constructa picture people identifying as Yugoslavs.
of Yugoslaviabased on religious affiliationand The goal of our analysis is to identify the so-
language spoken-imperfect substitutes for cial forces that influenced the creation of a
self-identification (Cohen and Warwick 1983: Yugoslav identity. Behind the social fact of
app.A). After WorldWarII, nationalitywas re- "Yugoslav identification" lies a diversity of
corded in the decennial census, but there was motives, inclinations,and rationales.We iden-
no Yugoslav category."Yugoslav"was first in- tify four sets of factors encouragingincreased
cluded in the third post-war census in 1961. self-identificationas a Yugoslav.These sets of
Officially this category was reserved for "na- factors constitutea tentativetypology of iden-
tionally noncommitted persons," and was tification with the state in multi-nationalsoci-
treatedas a residualcategoryfor those who of- eties and reveal the complexities and contra-
FAILEDSOURCESOF A COMMONIDENTITYIN YUGOSLAVIA 85

Table 1. Percentages of Adult Population of Yugoslavia Identifying Themselves as Yugoslavs in Yugoslavia and
within Each Republic and Province: 1961, 1971, and 1981

PercentageIdentifying as Yugoslav Predominant


Geographic Area 1961 1971 1981 Nationality in 1981

All of Yugoslavia 1.7 1.3 5.4 36.3% Serbian

Republics and Provinces


Croatia .4 1.9 8.2 75.1% Croatian
Serbia .2 1.4 4.8 85.4% Serbian
Bosnia/Herzegovina 8.4 1.2 7.9 39.5% Moslem
Kosovo .5 .1 .1 77.4% Albanian
Macedonia .1 .2 .7 67.0% Macedonian
Montenegro .3 2.1 5.3 68.3% Montenegro
Slovenia .2 .4 1.4 90.5% Slovenian
Vojvodina .2 2.4 8.2 54.3% Serbian

Sources: Statisticki Bilten SFRJ (No. 1295), 1982, Beograd, Yugoslavia: GovernmentPrinting Office. Statisticki
GodisnjakSFRJ, 1981, Beograd, Yugoslavia: GovernmentPrintingOffice.

dictions involved in attemptsto create a new unified people (Edelman 1971:164-68; Horo-
nationalidentity. witz 1985). The Warof National Liberationin
Yugoslaviahad been waged in orderto end for-
eign occupation. Ideologically, the Yugoslav
Modernization
Partisans, who took power at the war's end,
Yugoslavia experienced the transitionfrom a stressedthe unity of all nationalitiesin the fed-
primarilyagriculturalto a primarilyindustrial eral republic. Nationality as a divisive force
society during the post-World War II period was condemnedby patriots,who remembered
(Bozic, Cirkovic,Ekmecic, and Dedijer 1973). the partitioning of Yugoslavia during World
Urbanizationand increasingeducationand lit- War II and the foreign-inspired internecine
eracy were expected to diminish the salience warfare that cost hundreds of thousands of
of national identities as intergroupcontact in- lives. Thus, to identify as a Yugoslav was to
creased,as a sharednationalhistorydeveloped, condemn the forces that betrayedthe memory
and as a prosperous national future emerged of the war and to identify with the efforts of
(Deutsch 1969:27). With industrialization the Partisansto create a progressive, socialist
came geographic mobility and greatercontact society.' This legacy was carried forward in
among nationalitiesin urbanareas,and identi- Yugoslavia by the CommunistParty and also
fication as a Yugoslav became a means of eas- through workplace and community organiza-
ing social relations among individuals from tions (Burg 1983; Seroka and Smiljkovic
disparatenationalbackgroundsby minimizing 1986).
culturalbarriersand distinctions.
DemographicFactors
Political Participation
Yugoslav self-identificationalso provided in-
Creating symbolic representationsof the new dividuals a means of avoiding competing
state as well as providingopportunitiesto par- claims to their national allegiance. This was
ticipate in the ritualsof history and patriotism especially importantfor the childrenof nation-
are among the first efforts of new governments ally-mixedmarriages,whereeach parentmight
(Chirot 1988:72; Smith 1986). Tales of mili-
tary sacrifice and victory, identification of 1 For furtherdiscussion of Partisaninspired loy-
common enemies, and images of shared des- alties to the concept of Yugoslavia, see Connor
tiny are promoted to support the image of a (1984:540 ff).
86 AMERICANSOCIOLOGICAL
REVIEW

expect their child to recognize their particular Also, defensive Yugoslavism does not neces-
nationalidentity.By identifyingas a Yugoslav, sarily imply the same diminutionof religious
one could resist claims thatothers might make and cultural differences characteristic of as-
on one's identity and thus avoid potentialcon- similation (Alba 1990; Archdeacon 1983; Wa-
flicts. Yugoslav identificationalso provided a ters 1990).
way of breakingwith an increasinglydiscred-
ited past, especially among youngerpersons-
BUILDINGTHE YUGOSLAVSTATE
it was a protest against traditionalnationalist AMID NATIONALDIVERSITY
politics that seemed to be at the heart of the
region's problems (Banac 1984). The recogni- The YugoslavPartisanswho assumedpower in
tion that much of Yugoslaviawas less prosper- 1945 understoodthat a policy of a unified na-
ous than the rest of Europe-an observation tion-state,based on "imperialistic"attemptsto
often reflected in Yugoslav popularculture- deny nationhoodto the many peoples making
encourageda Yugoslav identity as a reflection up Yugoslavia,would fail (CohenandWarwick
of hopes for greaterintegrationinto the Euro- 1983; Shoup and Hoffman 1990). By defeat-
pean community.An importantstep in this di- ing the Chetniksand their ideology of Serbian
rection was the abandonmentof particularistic, domination and condemning the Ustasa's vi-
traditional notions and movement toward a sion of a GreaterCroatia,the League of Com-
vague notion of "Europeanism." Yugoslav munists of Yugoslavia(LCY) satisfied the na-
identification seemed closer to this ideal than tional aspirationsof the majorityof people in
more narrowethnic or nationalidentifications. Yugoslavia and gained widespread support
among people for whom the ideals of a com-
munist or socialist society had little relevance
Majority/MinorityStatus
(Cohen 1982). At the same time the concept of
Yugoslav self-identificationcould also serve as a protectivefederationwas attractiveto smaller
a way to resist assimilatinginto a dominantna- nationalgroups,such as Macedonians,who felt
tional group, as was the case for Moslems in threatenedby Greece and Bulgaria.The tactic
Bosnia in the 1961 census who identified as of the LCY,describedby Connor(1984:19) as
Yugoslavsratherthanas Croatsor Serbs (Stan- "strategicnationalism,"was to recreateYugo-
ovcic 1988). Persons of a minoritynationality slavia after the war, but without insisting on a
could claim a Yugoslav identity to resist pres- highly centralizedYugoslav nation-stateremi-
sure from the majorityto assimilateinto the lo- niscent of the Serbian-dominatedstate of the
cal dominantnationality.For example, Serbian interwarperiod.
nationalists often interpretedYugoslav identi- Politically this objective entailed establish-
fication by the Serb minorityliving in Croatia ing a federationof equal nations(equal regard-
as a defensive response to unfavorabletreat- less of geographic size or population) while
ment by the Croatmajority(Tomasevic 1975). launching a policy of centralism within the
Croatiannationalistsmade the same arguments LCY itself. The programof federalismwas an
on behalf of Croatsin the Vojvodina(Bilandzic importantmeans of winning support in post-
1985). Identifying as a Yugoslav thus avoided war Yugoslavia;it allowed the LCY to orches-
either assimilating into the majorityor label- tratethe creationof a new society that would,
ing oneself as a minority.A similarmotivation in time, transformthe lives of people and lead
for adopting a sharedidentity as a citizen of a to the replacementof politics based on national
state was reportedby Isaacs (1975) for India: identities with loyalty and identification with
"The educatedex-Untouchablesdearlywished the new Yugoslav state (Horowitz 1985). Na-
to shed their own group affiliations and their tionalismbased on regional identities was also
own group names, and, if they only could, be- expectedto decreasebecauseof the geographic
come 'Indian' and be nothing else" (p. 81). mixing of nationalitiesthat had occurreddur-
Defensive identification as a Yugoslav has ing WorldWarII and because of increasedmo-
some similarities to the strategy of "passing" bility that was partof the process of urbaniza-
among ethnic and racial groups in the United tion and modernizationbeing enacted by the
States, although it is a distinct strategyin that LCY (Bilandzic 1985).
the minoritygroup does not seek to be assimi- Under Tito's leadership,the LCY was san-
lated, but rather to be identified as neutral. guine about the possible centrifugal tenden-
FAILEDSOURCESOF A COMMONIDENTITYIN YUGOSLAVIA 87

cies of federalism leading to demands for a ence, and the League of Communists in the
looser confederation. It was confident these Vojvodina pursued greater autonomy from
tendencies could be controlledby the central- Serbia. Based on the organization of the
ized party system (Denitch 1976; Lederer League of Communistsinto republicanwings,
1969:434-37; Rusinow 1977:33; Zwick "nationalismnot only pervadedthe apparatus
1983:80). In 1953 the LCY began a series of but, on many occasions, turnedthe party into
initiatives aimed at maintainingpolitical inte- the principal battlefield of ethnonational
gration in the face of growing nationalism struggle"(Connor 1984:555).
(Ramet 1984:55-63). In the 1953 constitution Constitutionalamendmentsadoptedin 1967
the Chamberof Nationalities was eliminated reduced the power of the LCY at the federal
and the Chamberof Producerswas formed as level, in partas a reactionto the fear of increas-
a vehicle for the political representationof the ing Serbian domination of the Party. Hence-
worker self-management system first estab- forth the LCY functioned more "like an asso-
lished in 1950 (Terzuolo 1982). Five years ciation of eight regionalPartymachines than a
later, Kardelj (1960), one of the most vocal centralized system of socialist leadership"
supportersof worker self-management,wrote (Cohen and Warwick 1983:145). Along with
that "on the basis of inexorable socio-eco- economic development,the LCY's role in po-
nomic tendencies . . . [there] will be even litical indoctrination,including control of the
greater cultural merger of the Yugoslav media, was still expected to erode the strength
peoples" (p. 54). He based his conclusion on of nationalistsentimentseven as regionalrival-
several expectations. First, worker self-man- ries and moves toward confederation were
agement would accelerate the pace of eco- gaining strength(Tomc 1988). During this pe-
nomic development. Second, particularistic riod the ability of Tito to maintain ultimate
loyalties (including nationality) would give controlthroughthe LCY elite effectively coun-
way to working-class solidarity as people termandedthe increasing autonomy of local
found political avenues open to them as repre- Partyorgans on mattersextending beyond lo-
sentatives of self-managed enterprises.Third, cal importance.
worker self-management,as a form of class- A new constitution in 1974 increased the
based decentralizedpower in Yugoslav soci- trend toward a looser confederationas many
ety, would reduce the threatof any one nation- of the responsibilitiesand prerogativesof the
ality controlling the fate of any other (Cohen federalgovernmentwere divestedto the repub-
and Warwick1983:74-76). lics. A collective presidency was established
along with the right of any republic to veto a
decision by the presidency.The latterproved a
Devolution of Power to the Republics
majordevice in reducingthe power of the fed-
Duringthe 1950s Yugoslaviaexperienceda ris- eral government.
ing GNP along with uneven regional develop- The system of workerself-managementcon-
ment, setting the stage for economic national- tributedto the decentralizedpolitical and eco-
ism between regions (Cohen and Warwick nomic structureof Yugoslavia. Worker self-
1983:77). The gap between the republics- managementreducedthe power of the state to
measured in productivity and personal in- organizeand coordinatethe economy and gave
come-widened (Lydall 1989:186-96). At the primaryeconomic power to republics and en-
same time, workerself-managementbecame a terprisesthemselves.After 1954 the League of
vehicle for the expression of local ratherthan Communistsin each republicexercised consid-
class interests(Cohen and Warwick1983:76). erable influence over high-level appointments
From the mid-1960s onwardthe communist within firms, and local Party nomenklatura
partiesin the variousrepublicssaw themselves moved easily between the elite positions in
as representing their constituent "nations": firms and positions within the Party bureau-
Croatians in Croatia, Macedonians in Mace- cracy (Sekulic 1990). The weakness of federal
donia, Serbiansin Serbia,and so forth (Ramet control over the economy furtherfragmented
1984; Cohen and Warwick 1983). In Bosnia the LCY and encouragedthe League of Com-
the republic'sLeague of Communistspursued munistsin each republicto take on an increas-
policies designed to "protect" Moslems by ingly particularistic, and often nationalistic,
counterbalancingSerbian and Croatianinflu- agenda (Denitch 1991:77-79).
88 AMERICANSOCIOLOGICAL
REVIEW

For tactical reasons Tito and the Partisan post-war Yugoslavia (Lydall 1989; Denitch
leadership that ruled Yugoslavia after the war 1991). In January1990, the League of Com-
had approachedthe issue of Yugoslavismindi- munists of Yugoslaviaceased to exist, even as
rectly.The expectationthatthe LCY could pro- a symbolic unifying element, when the Slo-
vide political unity in the context of multina- veniandelegationwalkedout of the Fourteenth
tional identities was supported by the belief NationalCongress.
that this move would buy time for economic System failure bred not only distrust, but
developmentto erode particularisticidentities. provided political opportunitiesfor ambitious
The supportfor nationalism,includingthe rec- individuals to link the distress of the people
ognition of the "lessernations"(Montenegrans, with nationaldifferencesand historicalresent-
Macedonians, and Moslems), was also an at- ments (Cviic 1990; Devetak 1988). Newly
tempt to undercutSerbia'sdominanceover the emergingleadersand formerCommunistParty
other republics and nationalities (Rusinow leaders promotednationalistpride and offered
1985); by recognizing more peoples as "na- solutions marked by cultural atavism and
tionalities"the aspirationsof Serbiannational- sometimes by a desire to emulate more afflu-
ism could be checked. Increasedurbanization, ent nations. To many Westernobservers (e.g.,
reduced isolation of ruralareas, higher educa- Voirst 1991) the question, "Who are the
tional attainment,an open opportunitystruc- Yugoslavs?"was asked as an expressionof dis-
ture, worker-managedenterprises,and nearly belief that the idea of Yugoslavia could so
two generationsof living as a single state were quickly be abandoned, first by the Slovenes
expected to reduce the political strengthof na- and Croatians, soon to be followed by the
tionalism,leaving it its place culturaltraditions Bosnian Moslems, and finally by the people of
and ethnic pride held in common by all South every nationalgroupin the formerYugoslavia.
Slavic people. The question, "Who were the Yugoslavs?"
What actually transpired was increased raises important questions about commonly
fragmentation of identities and the develop- held assumptions regardingthe capability of
ment of political rivalries associated with na- states to foster unity among people with di-
tionalist claims. Yugoslav identificationcame verse cultures and historical experiences
to be seen as a threat to the republic-level (Vuskovic 1982; Tomc 1988). The question
CommunistParties that were increasinglygo- also poses a dilemmafor modernizationtheory,
ing in separate directions as federatedYugo- which assumesthatthe structuralconditionsof
slavia began to unravel.The stage for collapse economic growthand its attendantinstitutional
was set by growing economic gaps between frameworkwill negate particularisticloyalties
republics, economic nationalism,a weak cen- and provide sufficient rewards for people to
tral government, and the political fragmenta- adoptthe common outlooks, goals, and identi-
tion of the LCY ties of a multinational state (Hodson et al.
forthcoming; Nielsen 1985; see also Ragin
The Collapse of Yugoslavia 1979).

TodayYugoslaviahas disintegrated.The South


Slavs's experimentin building a unified state DATA AND VARIABLES
has failed as the variousnationalitiesdeny their Data
common interests and seek to forge smaller,
more nationally homogeneous states than the We used informationfrom two surveys in our
former Yugoslavia (Banac 1992). The eco- analysis.The first survey was conductedin the
nomic crisis of the 1980s was an important fall of 1984 and the winter of 1985 by the In-
catalyst for the disintegrationof the union- stitutefor Social Researchin Zagreb,Yugosla-
living standardsdeclined by at least a quarter, via. Using a disproportionatestratifiedrandom
and inflation reachedmore than 2,500 percent samplingframework,approximately3,600 ac-
in 1989. The legitimacy of the LCY was tively employed men and women in Croatia
openly questioned by Communist leaders were interviewed,about400 from each of nine
themselves, and most Yugoslavs desired to occupationalgroupings.The nine groups were
radicallyalteror abandonthe system of worker political functionaries,managersanddirectors,
self-managementthathad been the hallmarkof intellectuals and professionals, service work-
FAILEDSOURCESOF A COMMONIDENTITYIN YUGOSLAVIA 89
Table 2. Summary Statistics for Dependent and IndependentVariables by Republic: Yugoslavia, 1985 and 1989
Surveys

1985 Survey 1989 Survey


Variable Croatia Croatia Bosnia Serbia
Percent who identify as "Yugoslav" 10.6 9.0 14.4 4.6

Modernization
Urban residence (1 through3) 2.2 2.0 1.9 1.9
Education (years completed) 9.2 9.9 9.6 10.0
Read news (2 through 8) 5.0 4.6 4.4 4.4
Political Participation
Communist Party (percentage) 23.6 22.9 37.4 38.3
Work organization(percentage) 17.0 21.0 21.0 22.5
Community organization(percentage) 15.0 15.7 19.2 20.4
Demographic Factors
Age (years) 38.7 39.4 38.1 40.5
Nationally-mixed parentage(percentage) 10.0 10.2 9.0 5.8
Majority/MinorityStatus
Croat parentage(percentage) 71.4 72.7 17.6 0.3
Serb parentage(percentage) 14.4 12.8 28.3 84.6
Other homogenous parentage(percentage) 4.2 4.3 45.1 9.3
Number of cases 3,619 2,040 1,569 2,617

ers, three skill levels of manualworkers,self- (N = 1,569), and Serbia(N = 2,617). The addi-
employed "artisans,"and peasants. Based on tion of Bosnia and Serbia in the 1989 survey
the 1985 census of Croatia,weights were ap- allow us to compareCroatiawith the more na-
plied to this sample to replicate the distribu- tionally heterogeneous Bosnia and the more
tion of occupationsin the active workingpopu- homogeneousSerbia.
lation. After eliminating cases for which data Questions relevant for the analysis of
are missing, the resultantsample for the analy- Yugoslav self-identification were repeated in
sis contains 3,619 cases. Womenconstituteap- both the 1985 and 1989 surveys. Minor modi-
proximately one-third of the sample, concor- fications in the wording of some questions in
dant with the distributionof women in the paid 1989 do not appearto have affected the results
labor force. significantlyin that similarmodels fit the data
In the winterof 1989-1990 a second survey from both 1985 and 1989.
of randomly selected households was con-
ducted, this time in all six republics and the
Dependent Variable
two "autonomousprovinces"of the formerYu-
goslavia. The survey was conducted by the The dependent variable, self-identification, is
Consortium of Social Research Institutes of measuredby answers to a question asking the
Yugoslavia.This survey design yielded greater respondents' national identification. Most
numbersof respondentsover 60 years of age people answeredCroat,Serb,Moslem, or some
than did the 1985 Croatiansurvey and, unlike other nationality or ethnicity. In 1985 in
the earlier survey, included unemployed per- Croatia,10.6 percentof respondentsresponded
sons. Unemployed persons and respondents "Yugoslav"(see Table 2); by 1989 this figure
over 69 years of age (the upperlimit in 1985) had droppedto 9.0 percent.In 1989 in Bosnia,
were eliminated from our analysis to increase the level of Yugoslav self-identification was
the comparabilityof the two samples. more than 50 percent higher than in Croatia
We analyze data from 1989 for the three (14.4 percent),while in Serbia it was about 50
largest republics, Croatia(N = 2,040), Bosnia percentlower than in Croatia(4.6 percent).
90 AMERICANSOCIOLOGICAL
REVIEW

IndependentVariables asked to respondon the same 4-level scale as


in 1989 to questions asking how often they
Our analysis suggests that four sets of factors read newspapersand, separately,news maga-
influence the likelihood of identifying as a zines. These two variablesfrom the 1985 sur-
Yugoslav: modernization, political participa- vey were summed to again create an index of
tion, demographicfactors,andmajority/minor- readingthe news rangingfrom 2 to 8. The in-
ity status in a republic. dex for Croatia has a slightly higher mean
Modernization.We measure modernization value in 1985 than in 1989, and is slightly
using variables that tap urbanism,education, lower in Bosnia and Serbia in 1989 than in
and access to the media. In 1985 respondents Croatia.
were asked if they lived in a village, town, or Political participation. Participationin the
city. The majorityansweredcity, with substan- political system of Yugoslavia is expected to
tial minorities reporting that they lived in increase the likelihood of identifying as a
towns and villages. In 1989 respondentswere Yugoslav.We measure three types of political
asked if they lived in a village, a village center, involvement:membershipin the LCY,holding
a town, a town center,a city, or a regionalcen- office in workplaceorganizations,and holding
ter. The 1989 question formatgenerateda bet- office in community organizations. Each is
ter distributionof responses than did the 1985 coded as a dichotomousvariable.Membership
format. It was impossible, however, to reclas- in the LCY was not an elite status;individuals
sify the 1989 responses to approximate the who were LCY membersshouldnot be equated
1985 distributionand also remain faithful to with LCY officials, who held elite positions
the verbal options as they were presented in and who disproportionatelyenjoyed the privi-
1989. We chose to collapse the 1989 responses leges to which such power provided access
into the verbally analogous categories of the (Massey, Hodson, and Sekulic 1992). Never-
1985 survey-village (= 1), town (= 2), and theless, Partymembershipshould still indicate
city (= 3)-resulting in a somewhat different a greatercommitmentto explicitly articulated
distributionthanin 1985. The averageresiden- state goals. For both survey years, LCY mem-
tial location in both 1985 and 1989 is a town. bership is coded "yes" for those who either
The slight reductionin the mean of urbanism were membersat the time of the surveyor were
between 1985 and 1989 in Croatiais likely an membersin the past. Considerablyhigher per-
artifactof the change in questionformat. centages report being, or having been, mem-
To measure the effects of educationand ac- bers of the LCY in Bosnia and Serbia than in
cess to the media, we use the level of educa- Croatia.
tional attainmentof respondentsanda measure Ourothertwo measuresof political involve-
of how frequentlythey read the news. Educa- ment are holding office in political organiza-
tion is coded as years completedand averaged tions in the workplace and in the community.
10.0 in Serbia and 9.9 in Croatiain 1989. In Workplaceand communityorganizationswere
Bosnia, average years of schooling was only frequently dominated by Party members and
slightly lower at 9.6 years. The measure of served as conduits for Party goals and agen-
reading the news is based on two questions das. Such organizations, however, were also
askedin each survey.In the 1989 survey,news- frequently"captured"by local interestsoperat-
paper reading is measuredby a question ask- ing outside official Partymandates(Bilandzic
ing whether respondents read the newspaper 1985). In 1985 questions about holding office
never (= 1), monthly (= 2), weekly (= 3), or in workplace organizationswere asked sepa-
daily (= 4). Laterin the questionnaire,respon- ratelyfor workerself-managedenterprisesand
dents were asked to identify their three most for otherenterprises.Positive answersto either
common leisure activities. Some identified yielded a code of 1 (0 if both answers were
"readingnews" as their most importantleisure negative). In 1989 a summaryquestion asked
activity (= 4), others as their second most im- if the respondentoccupied an elected position
portant activity (= 3), or their third most im- at the workplace(yes = 1). Similarly,in 1985
portant activity (= 2), or not at all (= 1). Re- separate questions were asked pertaining to
sponses to these two questions were summed, holding office in organizationsin the commu-
creating an index of reading the news that nity dealing with either social or political ac-
ranges from 2 to 8. In 1985, respondentswere tivities. Office holding in eitheryielded a code
FAILEDSOURCESOF A COMMONIDENTITYIN YUGOSLAVIA 91

of 1. In 1989 these questions were combined, modernization,political participation,demo-


but a later question asked if respondentswere graphic factors, and majority/minoritystatus.
active in communityorganizations.A positive Logistic regression is appropriatefor a binary
response to either of these questions in 1989 dependent variable and allows utilization of
resulted in a code of 1. In 1989 in Croatia,a both categorical and continuous independent
higher percentage of respondents indicated variables. The regression coefficients from a
they held office in workplace organizations logistic regression can also readily be trans-
than in 1985, but there were similar levels of lated into easily interpretableodds indicating
participation in community organizations in the change in the likelihood of the dependent
1985 and 1989. The higher percentagereport- variable (identifying as a Yugoslav) given a
ing participationin work organizationsin 1989 unit shift in an independentvariable.
than in 1985 may have occurredbecause the
1985 questionnaire was administered at the
RESULTS
workplace where such participation would
have been harderto manufacture.Bosniansand The logistic regression coefficients estimating
Serbiansreportedhigherlevels of participation the effects of the modernization,political par-
in communityorganizationsthanCroatians(as ticipation, demographicfactors, and majority/
well as higher levels of Party membership). minoritystatus on Yugoslav self-identification
Participationin work organizationswas more for the combinedsample arepresentedin Table
nearlyequal across the threerepublicsin 1989. 3. This model also yields coefficients estimat-
Demographic factors. The third factor we ing the net contrastsin Yugoslav self-identifi-
expect to influence identifying as a Yugoslav cation between Croatiain 1985 and 1989 and
is representedby the demographiccharacteris- between Croatia and the republics of Bosnia
tics of age and nationally-mixed parentage. and Serbiain 1989.
The average age in each subsample ranges The model estimated for the total sample is
from 38 to 40. We determined nationally- highly statisticallysignificant-the variablesin
mixed parentageby comparinga respondent's the model reduce the chi-square from the
answers to questions about the nationalityof baseline model (with only the intercept in-
his or her motherandfather.In Croatiain 1985, cluded) by 1,573. Five of the 10 independent
10.0 percentof respondentsreportedthat their variablesare statisticallysignificantat the .001
parents were of different nationalities. For level. Modernizationtheories of identity for-
1989 this figure is 10.2 percent. Nationally- mationfind supportin a significanturbanresi-
mixed parentage occurred at only a slightly dence effect, but the effects of the other two
lower level in Bosnia (9.0 percent).The level modernizationvariables,educationandreading
of nationally-mixed parentage was signifi- the news, arenot significant.Partymembership
cantly lower in Serbia at 5.8 percent. strongly increases the likelihood of Yugoslav
Majority/minoritystatus. The final factorwe self-identification.Participationin community
expect to influence identificationas a Yugoslav organizationsalso has a significantpositive ef-
is the position of respondent'sparentsas a ma- fect on identifying as a Yugoslav, but partici-
jority or minority nationalityin their republic. pationin work organizationshas no significant
We expect persons of minority parentage to effect. Both of the demographicvariables,age
self-identify as Yugoslavs at a higherrate than and nationally-mixedparentage,arehighly sig-
persons of majority parentage. Persons of nificant.2
Croatian parentage were in the majority in
Croatiaand persons of Serbianparentagewere 2 Non-linearity in the age and education effects
in the majorityin Serbia. In Bosnia, Moslems was evaluated using 5-year cohorts and individual
were numericallydominant,followed by Serbs years of educational attainment.Dummy variables
and Croats,althoughall threegroupswere well representing these categories were added to the
represented. equation separatelyfor age and education. No sig-
nificant nonlinearitieswere found for either age or
education in either 1985 or 1989: Chi-square for
METHOD age categories (d.f. = 9) was 12.12 in 1985 and
13.86 in 1989; chi-squarefor education categories
We use logistic regression to analyze the pat- (d.f. = 18) was 24.77 in 1985 and 27.56 in 1989. Of
tern of self-identifying as a Yugoslavbased on the 54 coefficients tested to evaluate nonlinearity
92 AMERICANSOCIOLOGICAL
REVIEW

Table 3. Coefficients for the Logistic Regression of approximatelya 3 percentdecrease in the like-
Yugoslav Self-Identification on Selected Inde- lihood of identifying as a Yugoslav with each
pendent Variables: Yugoslavia, 1985 and 1989
year of advancingage. The odds of self-identi-
Independent Standard fying as a Yugoslav increase by about 60 per-
Variable Coefficient Error Odds cent for each increasingstep of urbanismfrom
village to town to city. Those who participate
Modernization
in community organizations are about one-
Urbanism .463*** (.056) 1.59 thirdmore likely to self-identify as a Yugoslav
Education -.020 (.014) .98
thanthose who do not participate.The odds of
Read news .019 (.036) 1.02
identifyingas a Yugoslavare approximately60
Political Participation percenthigherfor Partymembersthanfor non-
Communist Party .472*** (.096) 1.60 members.
Work organization .015 (.106) 1.01 To evaluate change across time in Yugoslav
Community .293* (.123) 1.34 self-identification, we compared respondents
organization in the 1985 Croatiansample to respondentsin
the 1989 Croatiansample. In 1985, Croatians
Demographic Factors were about 12 percentmore likely to self-iden-
Age -.034*** (.004) .97 tify as Yugoslavs than in 1989, but this coeffi-
Nationally-mixed 2.449*** (.115) 11.58 cient is not statistically significant and there-
parentage fore may not be a reliableestimate.Differences
Majority/MinorityStatus between republics in Yugoslav self-identifica-
Croat parentage -.949*** (.127) .39 tion are indicated by the contrasts between
Serb parentage -.003 (.121) 1.00 Croatia and Bosnia and between Croatia and
Other homogenous .000 1.00 Serbiain 1989. Bosnians were 35 percentmore
parentage(baseline) likely to self-identify as Yugoslavs than were
Croatians, and this difference is statistically
Net Contrasts
significantat the .05 level. Serbianswere only
Croatia 1985 .116 (.121) 1.12 35 percent as likely to identify as Yugoslav as
Croatia 1989 (baseline) .000 1.00
Croatiansin 1989, and this difference is sig-
Bosnia 1989 .296* (.133) 1.35
nificantat the .001 level.
Serbia 1989 -1.048*** (.148) .35
The effect of Croatian parentage on self-
Intercept -2.24*** identificationas a Yugoslavis significantat the
x2 (d.f. = 13) 1,572.7*** .001 level. However, since Table 3 is based on
-2 log likelihood 4,501 the combinedsampleacrossthe 1985 and 1989
Number of cases 9,845 surveysand acrossCroatia,Bosnia, and Serbia,
this coefficient cannot be used to evaluate the
<.05 ** <.01 *** < .001 (two-tailed tests)
hypothesis of defensive Yugoslavism which
suggests that minoritystatus in a republic will
Logistic regressioncoefficients are more in- encourage self-identification as a Yugoslav.
terpretablewhen translatedinto odds by using The appropriatetests of this hypothesis is pre-
the logistic coefficient as the exponent for the sented later in Table 4, where the model is
naturallog function. For example, for nation- evaluatedseparatelyfor each republicand time
ally-mixed parentage,e2449equals 11.58. The period.
lattercoefficient indicatesthat, net of the other It is possible that the various factors in our
factors in the model, a one-unit shift in nation- model of Yugoslavself-identificationhave dif-
ally-mixed parentage(from not having nation- ferent effects across time or across republics
ally-mixed parentage to having nationally- ratherthan uniformeffects as assumed for the
mixed parentage)increases the odds of identi- model evaluated in Table 3. Possible differ-
fying as a Yugoslavby more than 11 times. The ences in the model across time and republics
effect of age can be interpretedas indicating are evaluatedin Table4. This model is also re-
quired in order to test the hypothesis that mi-
in age and educationeffects in 1985 and 1989, only nority groups within republics have a greater
2 were statistically significant at the .05 level, and likelihood of identifying as Yugoslavs. The
these formed no interpretablepattern. model fits each of the four subsampleswell and
FAILEDSOURCESOF A COMMONIDENTITYIN YUGOSLAVIA 93

Table 4. Odds Ratios from Logistic Regression Analysis of Yugoslav Self-Identification by Year and Republic:
Yugoslavia, 1985 and 1989

Odds
1985 Survey 1989 Survey
IndependentVariable Croatia Croatia Bosnia Serbia

Modernization
Urban residence 1.54*** 1.17 2.21***.ttt 1.70**
Education .95* 1.01 1.02 1.02
Read news .90 1.04 1.27** 1.25*

Political Participation
Communist Party 1.58** 2.09** 1.59* 1.00#,t
Work organization 1.49*,tt .67 .73 .97
Community organization .85 1.62 1.19 2.20**

Demographic Factors
Age .97*** .96*** .97** .95***
Nationally-mixed parentage 9.82*** 9.68*** 16.08*** 13.60***

Majority/MinorityStatus
Croat parentage .29*** .30** l.22#tt 15.32**.ttt
Serb parentage 1.12 1.31 .98 .73
Other homogeneous parentage 1.00 1.00 1.00 1.00

X2(d.f. = 10) 581.4*** 384.7*** 333.0*** 256.1***


-2 log likelihood 1,860 850 961 714
Number of cases 3,619 2,040 1,569 2,617
Chow test for interactioneffects 18.2 (reference group) 42.9*** 34.2***
(d.f. = 10)
*p <.05 ** < .01 ***p< .001 (two-tailed t-tests)
tp < .05 ttp < .01 tttp < .001 (two-tailedt-testsindicatinga significantdifferencewhen comparedto Croatiain 1989)
#Coefficient is not significantly different from 1.00, but is significantly different from Croatiain 1989.

is statisticallysignificantin each case; the con- pation in work organizationsbecame a less ef-
tributionto chi-square(d.f. = 10) ranges from fective foundationfor Yugoslav identification.
581.4 in Croatiain 1985 to 256.1 in Serbia in This interpretationmust be made with some
1989. caution, however, given that the overall con-
We used the 1989 survey in Croatia as a trastbetween 1985 and 1989 is not significant.
baseline for evaluating temporaland regional In contrastto the negative findings for tem-
differences. We computed a Chow test to poral differences, the Chow tests for regional
evaluate changes over time by selecting only differencesbetweenCroatiaandBosnia in 1989
the two Croatian subsamples and evaluating andbetweenCroatiaand Serbiain 1989 aresta-
the contributionof a full set of interactionsthat tistically significantat the .001 level (with d.f.
allow the regression coefficients to differ in = 10, %2 = 42.9 and 34.2, respectively). Several
1985 and 1989. These interactionsreduce the interactionsare individuallysignificant.Urban
chi-square(d.f. = 10) by 18.2, which is not sta- residence has a larger positive effect on
tistically significant at the .05 level. The only Yugoslav self-identification in Bosnia, a less
interactionthatis individuallysignificantis for economically developed region, than in
participationin work organizations,which in- Croatia.In Croatiain 1989, each level of in-
creases self-identification as a Yugoslav in creasing urbanismincreases the odds of self-
Croatiain 1985, but not in 1989. This contrast identifying as a Yugoslav by a factor of 1.17
suggests that between 1985 and 1989 partici- (which is not significantlydifferentfrom 1.00).
94 AMERICANSOCIOLOGICAL
REVIEW

In Bosnia, each level of increasing urbanism educationalattainmenthad been thoughtto be


increases the odds of self-identifying as a an importantfactor in helping to generate an
Yugoslav by a factor of 2.21. In Serbia,the ur- increasingly cosmopolitan outlook and to fa-
ban residence effect is statisticallysignificant, cilitate the emergenceof a common identifica-
but it is not significantly differentfrom the ef- tion as Yugoslavs. That education did not in-
fect in Croatia.A second differencein the mod- crease Yugoslav self-identification may attest
els involves regional differences in the effects to republic-levelcontrol of curriculumcontent
of CommunistPartymembership.Partymem- and the teaching of separatenationalhistories
bership increases Yugoslav self-identification in each republic (see Jelavich 1983). Any in-
most in Croatiain 1989, less in Bosnia, and in crease of cosmopolitan attitudes associated
Serbiait has no effect at all. The differencebe- withhighereducationappearsto have been can-
tween Croatia,where Partymembershipmore celled out by the attractionof national identi-
than doubles the likelihood of identifying as a ties as a basis for social identity and political
Yugoslav,and Serbia, where it has no effect at mobilization.The controlof newspapersby lo-
all, is statistically significant at the .05 level. cal and republic-levelPartyorganizationsmay
The third significant interactioninvolves the also be responsiblefor the limited influence of
negativeeffect of homogenousCroatianparent- reading the news on Yugoslav self-identifica-
age on Yugoslav self-identificationin Croatia, tion. The limited effect of educationand access
where Croatiansare the majority,and its posi- to the mass media in encouraging Yugoslav
tive effects in Bosnia and Serbia, where identificationthus suggests the success of the
Croatiansare a minority.The higher level of regionalParties'efforts to representnationalist
Yugoslav self-identification associated with interestsand to condemnYugoslavism.
Croatianparentageis most dramaticin Serbia, The image of partisanYugoslavismbased on
whereCroatiansarea small minority.In Serbia, political participation in Communist Party,
having Croatianparentageincreasesthe likeli- workplace,or communitypolitical structuresis
hood of identifying as a Yugoslav by over 15 supportedby the finding that Party members
times. Respondentswith Serb parentagewere were more likely to identify as Yugoslavs.
also more likely to identify as Yugoslavsif they Similarly,in at least some instances,participa-
lived as a minority in Croatiaratherthan as a tion in workplaceor communityorganizations
memberof the majorityin Serbia,but this con- also increasedthe likelihood of identifying as
trastis not statisticallysignificant. a Yugoslav.Despite the Party'scommitmentto
a multinationalsociety andthe growing nation-
alist agenda of republic-level Party organiza-
DISCUSSION
tions, those directly involved in the Party, at
Identifying as a Yugoslav can occur through least outside of Serbia, were more likely to
several possible routes-through moderniza- subscribe to the view that Communist Party
tion, political participation,demographicfac- members should be above the provincial at-
tors, and majority/minoritystatus. Our analy- tachments of the past and that their personal
sis provides evidence for all four bases of loyalties to the concept of a unified Yugoslavia
Yugoslav identification. should provide an example for others to fol-
Modernizationeffects are evident in the in- low.
fluence of urbanresidencein all threerepublics The significance of age in the analysis sup-
examined.Urbanismis one of the strongestand ports the notion of demographic Yugoslavism.
most consistentdeterminantsof Yugoslaviden- Young people were more likely to reject na-
tification. This finding is consistent with the tional identitiesin preferencefor identifyingas
long-standingrecognitionthatcity living is as- a Yugoslav. In this manner they sought sym-
sociated with cosmopolitan attitudes (Wirth bolic entry into the milieu of modernismasso-
1956). The declining influence of urbanismin ciated with the broaderEuropeanculture.Also
Croatiabetween 1985 and 1989 may reflectthe notable is the powerful effect of nationally-
growth of Croatiannationalismduringthis pe- mixed parentage,which reduced allegiance to
riod in the cultural and political capital of nationalidentities and allowed an opening for
Zagreb.The lack of influence of educationon identification as a Yugoslav. Other demo-
self-identificationas a Yugoslav,however,does graphicfactorssuch as migrationandincreased
not support modernization theory. Increased contact with other nationalitiesmay also have
FAILEDSOURCESOF A COMMONIDENTITYIN YUGOSLAVIA 95

been encapsulatedin the effects of nationally- The civil and military conflicts now under-
mixed parentage. way in the formerYugoslavia and threatening
The concept of Yugoslavidentityas a defen- to erupt in other parts of EasternEurope and
sive strategyfor minoritynationalities is sup- the formerSoviet Union unmistakablyexhibit
ported by higher rates of Yugoslav identifica- elements based on nationalist fears, resent-
tion among those with Croatparentsin Bosnia ments, and hostilities. Emergingleaders of in-
and, most dramatically,Serbia. A similar pat- creasingly independentrepublics and states in
tern, though not statisticallysignificant,exists these areas identify themselves with particular
for Serbs, with Yugoslav identificationbeing nationalitiesand claim to representthe inter-
more likely for Serbs living in Croatiaas a mi- ests of all those similarly identified, both
nority group than for Serbs living in Serbia as within and outside currently defined bound-
a majority group. The difference between aries. The end of Soviet hegemony and the
Croats and Serbs in this regardprovides addi- eclipse of communistpartiesthroughoutEast-
tional supportfor the hypothesis of defensive ern Europe have provided new reminders of
Yugoslavism for minority nationalities.Serbs the strength of such nationalist feelings and
were numerically, politically, and militarily claims.
dominantin the formerYugoslavia,and it ap- The Yugoslav experience suggests that the
pears that they felt less need to subscribe to path to a sharedidentity,either throughpoliti-
Yugoslavism,whetherthey lived inside or out- cal persuasionor throughindustrializationand
side Serbia. Croats, by contrast, when living development,is neithersimple nor assured.In
outsideCroatia,were likely to take on the more Yugoslavia, strong Communist Party leader-
neutraland defensive postureof identifyingas ship and the transformationof the economy
Yugoslavs. The higher aggregate level of from an agriculturalto an industrialbase were
Yugoslav identification in ethnically-mixed expected to erode the traditionalbases for na-
Bosnia than in more homogenous Croatia or tional differences. When worker self-manage-
Serbia lends further support to the image of ment was introducedand the role of the LCY
Yugoslav identificationas a defensive strategy at the federal level diminished, however, eco-
for minorities. nomic nationalismand the creationof bastions
of regional political strengthwere made pos-
sible. Ratherthanexplicitly attackingnational-
CONCLUSIONS
ism as a source of divisiveness and instability,
Only a relatively small proportionof people in Yugoslavia'sleaders,especially at the republic
the former Yugoslavia ever expressed the so- level, endorsed nationalism as a right within
cial identity of being Yugoslavs. Many of the the context of equality among the nations of
social and political underpinningsof an emer- Yugoslavia.
gent shared identify, however, were in place The policies of Yugoslavia'sleaders, at least
and operating to increase the likelihood of as far as the national question is concerned,
Yugoslav identification.These included pow- seemed both expedient and prudent. By em-
erful influences for urbanresidence, intermar- phasizing economic development, workplace
riage, and political participation,even partici- democracy,economic and genderequality,tol-
pation in a CommunistPartyapparatusthat it- erancefor nationaldifferences,and equal legal
self was ambivalent about promoting a rights of all citizens, Tito and his colleagues
Yugoslav identity. It will remain unknown if assumed that time was on their side and that
these forces, given sufficient time, would ever the crises the state would inevitablyface could
have createda strongcollective identityfor the be cast in other than nationalistterms. For the
citizens of the former Yugoslavia as common former Yugoslavia, history has proven these
members of a unified nation-state.While this assumptions wrong. The playing out of con-
failureto establish a sharedidentityamong the flicting nationalistforces in the 1990s in other
people of this region cannot be said to explain parts of EasternEurope and in the former So-
the disintegrationof Yugoslavia,it is apparent viet Union provides a profound challenge to
thata sharedidentitywas not much in evidence the political and social will of the people of
as a mediatingmechanismsustainingYugosla- these areas. And understandingthese forces
via throughdifficult transitionsor slowing its provides a profoundchallenge to sociological
disintegrationinto warringnationalcamps. theory.
96 AMERICANSOCIOLOGICALREVIEW

DusKoSEKULIC is Associate Professor of Sociology Belanger, Sarahand MauricePinard. 1991. "Ethnic


and Professor of International Studies at George Movements and the Competition Model: Some
Mason University.He served as Editor in Chief of Missing Links." American Sociological Review
Revija za Sociologiju (Sociological Review), the 56:446-57.
journal of the Croatian Sociological Association, Bilandzic, Dusan. 1985. Historija Socijalisticke
and President of Zagreb University.His mainfields Federativne Republike Jugoslavije (History of
of interest are social stratification, nationalism, the Socialist Federal Republic of Yugoslavia).
economic sociology, and the problems of transfor- Zagreb,Yugoslavia: Skolska Knjiga.
mation of East EuropeanSocieties. Recentpublica- Bozic, Ivan, Sima Cirkovic, Milorad Ekmecic, and
tions include "Dissolution of Yugoslavia:Interna- Vladimir Dedijer. 1973. Istorija Jugoslavije
tional Consequences"(in NATO: The Challengeof (History of Yugoslavia). Beograd, Yugoslavia:
Change, edited by J. Simon, National Defense Uni- Prosveta.
versity, 1993) and "Nationalism Versus Democ- Breuilly, John. 1982. Nationalism and the State.
racy: Legacies of Marxism" (InternationalJournal New York: St. Martin's.
of Politics, Cultureand Society, vol. 6(1), 1992, pp. Burg, Steven L. 1983. Conflictand Cohesion in So-
113-32). Currentlyhe isfinishing a book titled Na- cialist Yugoslavia.Princeton,NJ: PrincetonUni-
tionalism and the Dissolution of Yugoslavia. versity.
Chirot,Daniel. 1988. Social Change in the Modern
GARTHMASSEY is Professor of Sociology at the Uni- Era. New York: Harcourt,Brace, Jovanovich.
versity of Wyoming. He is a Fulbright Research Cohen, Lenard. 1982. "BalkanConsociationalism:
Fellow at Godollo University in Hungaryfor the Ethnic Representation and Ethnic Distance in
1993-1994 academic year studyingchanging social Yugoslav Elites." Pp. 29-62 in At the Brink of
structure in rural Hungary. His research interests War and Peace: The Tito-Stalin Split in a His-
include social change in socialist andformerly so- toric Perspective,edited by W. S. Vucinich. New
cialist societies, labor, and internationalpolitical York: Brooklyn College.
economy. His collaborations with Sekulic and Cohen, Lenard and Paul Warwick. 1983. Political
Hodson include "Political Affiliation and Social Cohesion in a Fragile Mosaic. Boulder, CO:
Mobility in Socialist Yugoslavia" (in Research in Westview.
Social Stratificationand Mobility, vol. 11, edited by Collins, Randall and David V. Waller. 1992. "The
R. Althauser and M. Wallace), and "NationalTol- Geopolitics of Ethnic Mobilization:Some Theo-
erance in the Former Yugoslavia"(AmericanJour- retical Projections for the Soviet Bloc." Paper
nal of Sociology, forthcoming in May 1994). presented at a conference sponsored by George
Mason and George Washington Universities on
RANDY HoDsoNis Professor of Sociology at Indiana
"The Legacies of the Collapse of Marxism,"22-
University. His main research interests are in so-
25 Mar., Washington,DC
cial stratification and the sociology of work. He is Connor, Walker. 1984. The National Question in
currently studying ethnic relations and changing Marxist-LeninistTheoryand Strategy.Princeton,
patterns of social organization in theformer Yugo- NJ: PrincetonUniversity.
slavia. He is also using survey techniques and the Cviic, Christopher. 1990. "The Backgrounds and
descriptions provided in workplace ethnographies Implicationsof the Domestic Scene in Yugosla-
to study the maintenance and defense of dignity in via." Pp. 89-119 in Problemsof Balkan Security,
the workplace. His recent publications include "Is edited by P. Shoup. Washington, DC: Wilson
WorkerSolidarity Underminedby Autonomy and Center.
Participation?" (American Sociological Review, Davis, Horance. 1978. Towarda Marxist Theoryof
vol. 58, 1993, pp. 398-416) and with Teresa A. Nationalism. New York: Monthly Review.
Sullivan, The Social Organization of Work, Denitch, Bogdan. 1976. Legitimationof a Revolu-
(Wadsworth,2nd ed., forthcoming 1995). tion. New Haven, CT: Yale University.
. 1991. Limits and Possibilities. Minne-
REFERENCES apolis, MN: University of Minnesota.
Deutsch, Karl W. 1969. Nationalism and Its Alter-
Alba, RichardD. 1990. Ethnic Identity:The Trans- natives. New York: Alfred A. Knopf.
formation of White America. New Haven, CT: Devetak, Silvo. 1988. The Equality of Nations and
Yale University. Nationalities in Yugoslavia: Successes and Di-
Archdeacon,Thomas J. 1983. BecomingAmerican: lemmas. Vienna, Austria:Wilhelm Braumuller.
An Ethnic History. New York: Free Press. Djilas, Aleksa. 1991. The Contested Country:
Banac, Ivo. 1984. The National Question in Yugo- Yugoslav Unity and the CommunistRevolution,
slavia. Ithaca:Cornell University. 1919-1953. Cambridge, MA: Harvard Univer-
. 1992. "The FearfulAsymmetryof War: sity.
The Causes and Consequences of Yugoslav De- Edelman, Murray. 1971. Politics as Symbolic Ac-
mise." Daedalus 121(2):1-22. tion. New York: Academic Press.
FAILEDSOURCESOF A COMMONIDENTITYIN YUGOSLAVIA 97

Enloe, CynthiaH. 1973. Ethnic Conflictand Politi- tion and Nation Building." Pp. 562-600 in The
cal Development.Boston, MA: Little, Brown and Formation of Nation States in WesternEurope,
Company. edited by C. Tilly. Princeton,NJ: PrincetonUni-
Gertz, Clifford. 1973. The Interpretation of Cul- versity.
tures. New York: Basic Books. Rusinow, Dennison. 1977. The Yugoslav Experi-
Hechter, Michael. 1976. "Ethnicityand Industrial- ment, 1948-1977. Berkeley, CA: University of
ization: On the Proliferationof the CulturalDi- California.
vision of Labor."Ethnicity3:214-24. 1985. "NationalitiesPolicy and the 'Na-
Hodson, Randy, Dusko Sekulic, and GarthMassey. tional Question."' Pp. 131-65 in Yugoslavia in
Forthcoming."NationalTolerancein the Former the 1980s, edited by P. Ramet. Boulder, CO:
Yugoslavia." American Journal of Sociology Westview.
(May 1994). Sekulic, Dusko. 1990. "Recruitmentfor Elite Posi-
Horowitz, Donald L. 1985. Ethnic Groups in Con- tions." Sociologija 32:105-25.
flict. Berkeley, CA: University of California. Seroka, Jim and Rados Smiljkovic. 1986. Political
Isaacs, Harold R. 1975. Idols of the Tribe. New Organizations in Socialist Yugoslavia. Durham,
York: Harperand Row. NC: Duke University.
Jelavich, Charles. 1983. "SerbianTextbooks: To- Shoup, Paul S. and George W. Hoffman. 1990.
ward GreaterSerbia or Yugoslavia?"Slavic Re- Problems of Balkan Security. Washington, DC:
view 42:601-19. Wilson Center.
Kardelj, Edvard. 1960. Razvoj Slovenackog Smith, Anthony D. 1986. The Ethnic Origins of
Nacionalnog Pitanja (Development of the Nations. Oxford, England:Basil Blackwell.
Slovenian National Question). Beograd, Yugo- Stanovcic, Vojislav. 1988. "History and Status of
slavia: Kultura. Ethnic Conflicts." Pp. 23-40 in Yugoslavia: A
Kedouire,Elie. 1960. Nationalism.London:Hutch- Fractured Federalism, edited by D. Rusinow.
inson. Washington,DC: Wilson Center.
Kohn, Hans. 1961. The Idea of Nationalism. New Terzuolo, Eric R. 1982. "Soviet-Yugoslav Conflict
York: MacMillan. and the Origins of Yugoslavia's Self-Manage-
Lederer, Ivo J. 1969. "Nationalism and the ment System." Pp. 195-218 in At the Brink of
Yugoslavs." Pp. 396-438 in Nationalismin East- War and Peace: The Tito-Stalin Split in a His-
ern Europe, edited by P. F. Sugar and I. J. toric Perspective,edited by W. S. Vucinich. New
Lederer.Seattle, WA: University of Washington. York:Brooklyn College.
Lydall, Harold. 1989. Yugoslaviain Crisis. Oxford, Tilly, Charles. 1975. The Formation of National
England:Clarendon. States in Western Europe. Princeton, NJ:
Massey, Garth,Randy Hodson, and Dusko Sekulic. PrincetonUniversity.
1992. "Political Affiliation and Social Mobility Tomasevic, Jozo. 1975. The Chetniks. Stanford,
in Socialist Yugoslavia."Pp. 233-58 in Research CA: StanfordUniversity.
in Social Stratificationand Mobility,vol. 11, ed- Tomc, Gregor.1988. "Classes,PartyElites and Eth-
ited by R. AlthauserandM. Wallace. Greenwich, nic Groups."Pp. 58-77 in Yugoslavia:A Frac-
CT: JAI. tured Federalism, edited by D. Rusinow. Wash-
Nielsen, Francois. 1985. "Towarda Theory of Eth- ington, DC: Wilson Center.
nic Solidarity in Modem Societies." American Tudjman,Franjo. 1981. Nationalism in Contempo-
Sociological Review 50:133-49. raryEurope.Boulder,CO: East EuropeanMono-
Olzak, Susan. 1992. The Dynamics of Ethnic Com- graphs,University of Colorado.
petition and Conflict. Stanford, CA: Stanford Voirst, Milton. 1991. "The Yugoslav Idea." New
University. Yorker(18 Mar.):58-79.
Petrovic, Ruza. 1983. "The National Composition Vuskovic, Boris. 1982. "Tko su Jugoslaveni?"
of the Population." Yugoslav Survey 24 (3 (Who Are the Yugoslavs?). Nase Teme26:1701-
Aug.):21-34. 23.
Ragin, Charles C. 1979. "EthnicPolitical Mobili- Waters, Mary C. 1990. Ethnic Options: Choosing
zation." American Sociological Review 44:619- Identities in America. Berkeley, CA: University
35. of California.
Ramet, Pedro. 1984. Nationalism and Federalism Wirth, Louis. 1956. The Ghetto. Chicago, IL: Uni-
in Yugoslavia,1963-1983. Bloomington, IN: In- versity of Chicago.
diana University. Zwick, Peter. 1983. National Communism.Boulder,
Rokkan, Stein. 1975. "Dimensionsof State Forma- CO: Westview.