STÖR Who Were the Yugoslavs? Failed Sources of a Common Identity in the Former Yugoslavia Dusko Sekulic, Garth Massey, Randy Hodson American Sociological Review, Volume 59, Issue 1 (Feb., 1994), 83-97. Stable URL: http://links.jstor.org/sici?sici=0003-1224%28199402%2959%3Al%3C83%3AWWTYFS%3E2.0.CO%3B2-J Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://uk.jstor.org/about/terms.html. 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. 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. American Sociological Review is published by American Sociological Association. Please contact the publisher for further permissions regarding the use of this work. Publisher contact information may be obtained at http://uk.jstor.org/journals/asa.html. American Sociological Review ©1994 American Sociological Association JSTOR and the JSTOR logo are trademarks of JSTOR, and are Registered in the U.S. Patent and Trademark Office. For more information on JSTOR contactjstor@mimas.ac.uk. ©2002 JSTOR http://uk.jstor.org/ Tue Sep 10 12:04:26 2002 WHO WERE THE YUGOSLAVS? FAILED SOURCES OF A COMMON IDENTITY IN THE FORMER YUGOSLAVIA* Dusko Sekulic Garth Massey George Mason University University of Wyoming Randy Hodson Indiana University Yugoslavia's leaders believed that a policy of equality among the many nationalities in Yugoslavia, in tandem with Communist Party hegemony, would allow nationalism within Yugoslavia to exist, mature, and finally diminish as a political force without jeopardizing 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 identity 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 intensified nationalist feelings. In the early 1990s Yugoslavia's experiment in building a multinational state was replaced with open hostilities and warfare among the South Slavs. We identify f our 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 members, 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 override the centrifugal forces of rising nationalism. Implications for political integration in Eastern Europe and the former Soviet Union are discussed. The dream of nineteenth-century nationalism was that a common language, history, and habits of everyday life justified the formation of distinct nation states that would represent and protect ethnically homogeneous groups (Gertz 1973; Isaacs 1975; Rokkan 1975; Smith 1986). In the twentieth century states have developed a variety of political mechanisms to accommodate national diversity within their borders, including legal recognition of minority nationalities, proportionate seating in legislatures, and policies favoring economically disadvantaged nationalities and areas (Cohen * Direct all correspondence to Dusko Sekulic, International Institute, 4001 N. Fairfax Drive, Suite 450, George Mason University, Arlington, VA 22203. This research was supported in part by the National Science Foundation through the United States-Yugoslavia Joint Fund for Scientific and Technological Cooperation. We thank York Brad-shaw, Burke Grandjean, Charles Jelavich, Dorothy Watson, and three anonymous ASR reviewers for their comments on previous drafts. and Warwick 1983; Collins and Waller 1992; Enloe 1973; Nielsen 1985). In addition, industrialization and the establishment of modern mass institutions, such as education and mass communications, have been emphasized by state builders, in part with the expectation that these would erode the differences upon which national identities were based (Davis 1978; Hodson, Sekulic, and Massey forthcoming). In some states, such as France, national policies, industrialization, and mass institutions produced integrated national identities (Tilly 1975). In other cases the integrative consequences of these processes have been slower to develop; and in others, these processes may have exacerbated rather than eased national rivalries within states (Belanger and Pinard 1991; Hechter 1976; Olzak 1992; Tudjman 1981). Now, at the end of the twentieth century, a resurgence of claims, rivalries, and conflicts among national groups within states threaten the state-building accomplishments of past decades in many parts of the world. American Sociological Review, 1994, Vol. 59 (February:83-97) 83 84 AMERICAN SOaOLOGICAL REVIEW To establish a new, ethnically diverse state it is seen as critical that the people adopt a common identity as citizens of that state, as members of a unified political system composed of groups otherwise diverse in language, religion, customs, ethnicity, or historical experience. In such states, however, there is a tension between ethnic or national groups wanting to maintain their own sovereignty within the state and the state's need to integrate such groups into a cohesive political unit that represents a shared ideology and vision of the future (Kedouire 1960; Kohn 1961; Smith 1986). Symptomatic of this tension is the struggle over identity, with the particular national groups trying to preserve their identities and the state attempting to impose a new identity assimilating and subsuming the more particular identities. The former Yugoslavia is an instance where the integrative processes of state identify formation have failed. Many of the underpinnings necessary to integrative processes were present in Yugoslavia, but they proved insufficient in the face of economic downturns and resurgent nationalist forces. We identify four potential routes that led people in the former Yugoslavia to identify themselves as members of a multinational state rather than as members of a specific nationality. We use data from two major social surveys done in Yugoslavia in 1985 and 1989 (just prior to the dissolution of the state) to examine the forces that facilitated or undermined the emergence of a shared Yugoslav identity. ROUTES TO YUGOSLAV SELF-IDENTIFICATION The first Yugoslav census after World War I did not record nationality, although the official name of the newly formed state was "The Kingdom of the Serbs, Croats, and Slovenes," clearly indicating its multinational character. This prompted historians to construct a picture of Yugoslavia based on religious affiliation and language spoken—imperfect substitutes for self-identification (Cohen and Warwick 1983: app. A). After World War H, nationality was recorded in the decennial census, but there was no Yugoslav category. "Yugoslav" was first included in the third post-war census in 1961. Officially this category was reserved for "nationally noncommitted persons," and was treated as a residual category for those who of- fered no particular national identity (Petrovic 1983). Table 1 presents the percentages of the population of Yugoslavia identifying as Yugoslav in 1961, 1971, and 1981. The modest decline in self-identification as a Yugoslav for the country as a whole between 1961 and 1971 was primarily the result of a decline in Yugoslav identifiers in Bosnia-Hercegovina (hereafter referred to as Bosnia) in 1971. According to Ramet (1984:144-49), high Yugoslav self-identification in Bosnia in 1961 occurred because Moslems refused to identify themselves with dominant national groups (i.e., Serbs or Croats). The 1971 census was the first to allow "Moslem" as a nationality, and many Moslem Bosnians switched from the "Yugoslav" to the "Moslem" category in 1971. Moslems in Yugoslavia, many of whom live in Bosnia, are Slavs who adopted not only the Islamic faith, but also embraced many other cultural and linguistic features of the Turkish people who, during the period of the Ottoman Empire, controlled much of what would later become Yugoslavia. Apart from Bosnia between 1961 and 1971 and Kosovo, self-identification as a Yugoslav shows a general pattern of increase from 1961 to 1981, especially in Croatia and the Vojvodine, and in Bosnia between 1971 and 1981. Among the republics and provinces, Kosovo showed the lowest Yugoslav identification in 1971 and 1981, with most people in Kosovo identifying themselves as either Albanian or Serbian. One factor that encouraged identification as a Yugoslav was the heterogeneity of nationalities within the republic or province (Breuilly 1982; Djilas 1991). Contact between different nationalities is increased in heterogeneous settings leading to greater social mixing and intermarriage (Isaacs 1975). Bosnia and the Vojvodine were the former Yugoslavia's most nationally diverse areas, and both republics had larger than average proportions of people identifying as Yugoslavs. The goal of our analysis is to identify the social forces that influenced the creation of a Yugoslav identity. Behind the social fact of "Yugoslav identification" lies a diversity of motives, inclinations, and rationales. We identify four sets of factors encouraging increased self-identification as a Yugoslav. These sets of factors constitute a tentative typology of identification with the state in multi-national societies and reveal the complexities and contra- FAILED SOURCES OF A COMMON IDENTITY IN YUGOSLAVIA 85 Table 1. Percentages of Adult Population of Yugoslavia within Each Republic and Province: 1961, 1971, Identifying Themselves as Yugoslavs and 1981 in Yugoslavia and Percentage Identifying as ; Yugoslav Predc National )minant Geographic Area 1961 1971 1981 ityin 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: Štatistickí Eilten SFRJ (No. 1295), 1982, Beograd, Yugoslavia: Government Printing Office. Štatistickí Godisnjak SFRJ, 1981, Beograd, Yugoslavia: Government Printing Office. dictions involved in attempts to create a new national identity. Modernization Yugoslavia experienced the transition from a primarily agricultural to a primarily industrial society during the post-World War II period (Bozic, Cirkovic, Ekmecic, and Dedijer 1973). Urbanization and increasing education and literacy were expected to diminish the salience of national identities as intergroup contact increased, as a shared national history developed, and as a prosperous national future emerged (Deutsch 1969:27). With industrialization came geographic mobility and greater contact among nationalities in urban areas, and identification as a Yugoslav became a means of easing social relations among individuals from disparate national backgrounds by minimizing cultural barriers and distinctions. Political Participation Creating symbolic representations of the new state as well as providing opportunities to participate in the rituals of history and patriotism are among the first efforts of new governments (Chirot 1988:72; Smith 1986). Tales of military sacrifice and victory, identification of common enemies, and images of shared destiny are promoted to support the image of a unified people (Edelman 1971:164-68; Horowitz 1985). The War of National Liberation in Yugoslavia had been waged in order to end foreign occupation. Ideologically, the Yugoslav Partisans, who took power at the war's end, stressed the unity of all nationalities in the federal republic. Nationality as a divisive force was condemned by patriots, who remembered the partitioning of Yugoslavia during World War II and the foreign-inspired internecine warfare that cost hundreds of thousands of lives. Thus, to identify as a Yugoslav was to condemn the forces that betrayed the memory of the war and to identify with the efforts of the Partisans to create a progressive, socialist society.1 This legacy was carried forward in Yugoslavia by the Communist Party and also through workplace and community organizations (Burg 1983; Seroka and Smiljkovic 1986). Demographic Factors Yugoslav self-identification also provided individuals a means of avoiding competing claims to their national allegiance. This was especially important for the children of nationally-mixed marriages, where each parent might 1 For further discussion of Partisan inspired loyalties to the concept of Yugoslavia, see Connor (1984:540 ff). 86 AMERICAN SOCIOLOGICAL REVIEW expect their child to recognize their particular national identity. By identifying as a Yugoslav, one could resist claims that others might make on one's identity and thus avoid potential conflicts. Yugoslav identification also provided a way of breaking with an increasingly discredited past, especially among younger persons— it was a protest against traditional nationalist politics that seemed to be at the heart of the region's problems (Banac 1984). The recognition that much of Yugoslavia was less prosperous than the rest of Europe—an observation often reflected in Yugoslav popular culture— encouraged a Yugoslav identity as a reflection of hopes for greater integration into the European community. An important step in this direction was the abandonment of particularistic, traditional notions and movement toward a vague notion of "Europeanism." Yugoslav identification seemed closer to this ideal than more narrow ethnic or national identifications. Majority/Minority Status Yugoslav self-identification could also serve as a way to resist assimilating into a dominant national group, as was the case for Moslems in Bosnia in the 1961 census who identified as Yugoslavs rather than as Croats or Serbs (Stan-ovcic 1988). Persons of a minority nationality could claim a Yugoslav identity to resist pressure from the majority to assimilate into the local dominant nationality. For example, Serbian nationalists often interpreted Yugoslav identification by the Serb minority living in Croatia as a defensive response to unfavorable treatment by the Croat majority (Tomasevic 1975). Croatian nationalists made the same arguments on behalf of Croats in the Vojvodina (Bilandzic 1985). Identifying as a Yugoslav thus avoided either assimilating into the majority or labeling oneself as a minority. A similar motivation for adopting a shared identity as a citizen of a state was reported by Isaacs (1975) for India: "The educated ex-Untouchables dearly wished to shed their own group affiliations and their own group names, and, if they only could, become Indian' and be nothing else" (p. 81). Defensive identification as a Yugoslav has some similarities to the strategy of "passing" among ethnic and racial groups in the United States, although it is a distinct strategy in that the minority group does not seek to be assimilated, but rather to be identified as neutral. Also, defensive Yugoslavism does not necessarily imply the same diminution of religious and cultural differences characteristic of assimilation (Alba 1990; Archdeacon 1983; Waters 1990). BUILDING THE YUGOSLAV STATE AMID NATIONAL DIVERSITY The Yugoslav Partisans who assumed power in 1945 understood that a policy of a unified nation-state, based on "imperialistic" attempts to deny nationhood to the many peoples making up Yugoslavia, would fail (Cohen and Warwick 1983; Shoup and Hoffman 1990). By defeating the Chetniks and their ideology of Serbian domination and condemning the Ustasa's vision of a Greater Croatia, the League of Communists of Yugoslavia (LCY) satisfied the national aspirations of the majority of people in Yugoslavia and gained widespread support among people for whom the ideals of a communist or socialist society had little relevance (Cohen 1982). At the same time the concept of a protective federation was attractive to smaller national groups, such as Macedonians, who felt threatened by Greece and Bulgaria. The tactic of the LCY, described by Connor (1984:19) as "strategic nationalism," was to recreate Yugoslavia after the war, but without insisting on a highly centralized Yugoslav nation-state reminiscent of the Serbian-dominated state of the interwar period. Politically this objective entailed establishing a federation of equal nations (equal regardless of geographic size or population) while launching a policy of centralism within the LCY itself. The program of federalism was an important means of winning support in postwar Yugoslavia; it allowed the LCY to orchestrate the creation of a new society that would, in time, transform the lives of people and lead to the replacement of politics based on national identities with loyalty and identification with the new Yugoslav state (Horowitz 1985). Nationalism based on regional identities was also expected to decrease because of the geographic mixing of nationalities that had occurred during World War II and because of increased mobility that was part of the process of urbanization and modernization being enacted by the LCY (Bilandzic 1985). Under Tito's leadership, the LCY was sanguine about the possible centrifugal tenden- FAILED SOURCES OF A COMMON IDENTITY IN YUGOSLAVIA 87 cies of federalism leading to demands for a looser confederation. It was confident these tendencies could be controlled by the centralized party system (Denitch 1976; Lederer 1969:434-37; Rusinow 1977:33; Zwick 1983:80). In 1953 the LCY began a series of initiatives aimed at maintaining political integration in the face of growing nationalism (Ramet 1984:55-63). In the 1953 constitution the Chamber of Nationalities was eliminated and the Chamber of Producers was formed as a vehicle for the political representation of the worker self-management system first established in 1950 (Terzuolo 1982). Five years later, Kardelj (1960), one of the most vocal supporters of worker self-management, wrote that "on the basis of inexorable socio-economic tendencies . . . [there] will be even greater cultural merger of the Yugoslav peoples" (p. 54). He based his conclusion on several expectations. First, worker self-management would accelerate the pace of economic development. Second, particularistic loyalties (including nationality) would give way to working-class solidarity as people found political avenues open to them as representatives of self-managed enterprises. Third, worker self-management, as a form of class-based decentralized power in Yugoslav society, would reduce the threat of any one nationality controlling the fate of any other (Cohen and Warwick 1983:74-76). Devolution of Power to the Republics During the 1950s Yugoslavia experienced a rising GNP along with uneven regional development, setting the stage for economic nationalism between regions (Cohen and Warwick 1983:77). The gap between the republics— measured in productivity and personal income—widened (Lydall 1989:186-96). At the same time, worker self-management became a vehicle for the expression of local rather than class interests (Cohen and Warwick 1983:76). From the mid-1960s onward the communist parties in the various republics saw themselves as representing their constituent "nations": Croatians in Croatia, Macedonians in Macedonia, Serbians in Serbia, and so forth (Ramet 1984; Cohen and Warwick 1983). In Bosnia the republic's League of Communists pursued policies designed to "protect" Moslems by counterbalancing Serbian and Croatian influ- ence, and the League of Communists in the Vojvodina pursued greater autonomy from Serbia. Based on the organization of the League of Communists into republican wings, "nationalism not only pervaded the apparatus but, on many occasions, turned the party into the principal battlefield of ethnonational struggle" (Connor 1984:555). Constitutional amendments adopted in 1967 reduced the power of the LCY at the federal level, in part as a reaction to the fear of increasing Serbian domination of the Party. Henceforth the LCY functioned more "like an association of eight regional Party machines than a centralized system of socialist leadership" (Cohen and Warwick 1983:145). Along with economic development, the LCY's role in political indoctrination, including control of the media, was still expected to erode the strength of nationalist sentiments even as regional rivalries and moves toward confederation were gaining strength (Tome 1988). During this period the ability of Tito to maintain ultimate control through the LCY elite effectively countermanded the increasing autonomy of local Party organs on matters extending beyond local importance. A new constitution in 1974 increased the trend toward a looser confederation as many of the responsibilities and prerogatives of the federal government were divested to the republics. A collective presidency was established along with the right of any republic to veto a decision by the presidency. The latter proved a major device in reducing the power of the federal government. The system of worker self-management contributed to the decentralized political and economic structure of Yugoslavia. Worker self-management reduced the power of the state to organize and coordinate the economy and gave primary economic power to republics and enterprises themselves. After 1954 the League of Communists in each republic exercised considerable influence over high-level appointments within firms, and local Party nomenklatura moved easily between the elite positions in firms and positions within the Party bureaucracy (Sekulic 1990). The weakness of federal control over the economy further fragmented the LCY and encouraged the League of Communists in each republic to take on an increasingly particularistic, and often nationalistic, agenda (Denitch 1991:77-79). 88 AMERICAN SOCIOLOGICAL REVIEW For tactical reasons Tito and the Partisan leadership that ruled Yugoslavia after the war had approached the issue of Yugoslavism indirectly. The expectation that the LC Y could provide political unity in the context of multinational identities was supported by the belief that this move would buy time for economic development to erode particularistic identities. The support for nationalism, including the recognition of the "lesser nations" (Montenegrans, Macedonians, and Moslems), was also an attempt to undercut Serbia's dominance over the other republics and nationalities (Rusinow 1985); by recognizing more peoples as "nationalities" the aspirations of Serbian nationalism could be checked. Increased urbanization, reduced isolation of rural areas, higher educational attainment, an open opportunity structure, worker-managed enterprises, and nearly two generations of living as a single state were expected to reduce the political strength of nationalism, leaving it its place cultural traditions and ethnic pride held in common by all South Slavic people. What actually transpired was increased fragmentation of identities and the development of political rivalries associated with nationalist claims. Yugoslav identification came to be seen as a threat to the republic-level Communist Parties that were increasingly going in separate directions as federated Yugoslavia began to unravel. The stage for collapse was set by growing economic gaps between republics, economic nationalism, a weak central government, and the political fragmentation of the LCY. The Collapse of Yugoslavia Today Yugoslavia has disintegrated. The South Slavs's experiment in building a unified state has failed as the various nationalities deny their common interests and seek to forge smaller, more nationally homogeneous states than the former Yugoslavia (Banac 1992). The economic crisis of the 1980s was an important catalyst for the disintegration of the union— living standards declined by at least a quarter, and inflation reached more than 2,500 percent in 1989. The legitimacy of the LCY was openly questioned by Communist leaders themselves, and most Yugoslavs desired to radically alter or abandon the system of worker self-management that had been the hallmark of post-war Yugoslavia (Lydall 1989; Denitch 1991). In January 1990, the League of Communists of Yugoslavia ceased to exist, even as a symbolic unifying element, when the Slovenian delegation walked out of the Fourteenth National Congress. System failure bred not only distrust, but provided political opportunities for ambitious individuals to link the distress of the people with national differences and historical resentments (Cviic 1990; Devetak 1988). Newly emerging leaders and former Communist Party leaders promoted nationalist pride and offered solutions marked by cultural atavism and sometimes by a desire to emulate more affluent nations. To many Western observers (e.g., Voirst 1991) the question, "Who are the Yugoslavs?" was asked as an expression of disbelief that the idea of Yugoslavia could so quickly be abandoned, first by the Slovenes and Croatians, soon to be followed by the Bosnian Moslems, and finally by the people of every national group in the former Yugoslavia. The question, "Who were the Yugoslavs?" raises important questions about commonly held assumptions regarding the capability of states to foster unity among people with diverse cultures and historical experiences (Vuskovic 1982; Tome 1988). The question also poses a dilemma for modernization theory, which assumes that the structural conditions of economic growth and its attendant institutional framework will negate particularistic loyalties and provide sufficient rewards for people to adopt the common outlooks, goals, and identities of a multinational state (Hodson et al. forthcoming; Nielsen 1985; see also Ragin 1979). DATA AND VARIABLES Data We used information from two surveys in our analysis. The first survey was conducted in the fall of 1984 and the winter of 1985 by the Institute for Social Research in Zagreb, Yugoslavia. Using a disproportionate stratified random sampling framework, approximately 3,600 actively employed men and women in Croatia were interviewed, about 400 from each of nine occupational groupings. The nine groups were political functionaries, managers and directors, intellectuals and professionals, service work- FAILED SOURCES OF A COMMON IDENTITY IN YUGOSLAVIA 89 Table 2. Summary Statistics for Dependent and Independent Variables by Republic: Yugoslavia, 1985 and 1989 Surveys Variable 1985 Survey Croatia 1989 Survey Croatia Bosnia Serbia Percent who identify as "Yugoslav" 10.6 Modernization Urban residence (1 through 3) Education (years completed) Read news (2 through 8) Political Participation Communist Party (percentage) Work organization (percentage) Community organization (percentage) Demographic Factors Age (years) Nationally-mixed parentage (percentage) Majority/Minority Status Croat parentage (percentage) Serb parentage (percentage) Other homogenous parentage (percentage) Number of cases 3,619 9.0 2,040 14.4 1,569 4.6 2.2 2.0 1.9 1.9 9.2 9.9 9.6 10.0 5.0 4.6 4.4 4.4 23.6 22.9 37.4 38.3 17.0 21.0 21.0 22.5 15.0 15.7 19.2 20.4 38.7 39.4 38.1 40.5 10.0' 10.2 9.0 5.8 71.4 72.7 17.6 0.3 14.4 12.8 28.3 84.6 4.2 4.3 45.1 9.3 2,617 ers, three skill levels of manual workers, self-employed "artisans," and peasants. Based on the 1985 census of Croatia, weights were applied to this sample to replicate the distribution of occupations in the active working population. After eliminating cases for which data are missing, the resultant sample for the analysis contains 3,619 cases. Women constitute approximately one-third of the sample, concordant with the distribution of women in the paid labor force. In the winter of 1989-1990 a second survey of randomly selected households was conducted, this time in all six republics and the two "autonomous provinces" of the former Yugoslavia. The survey was conducted by the Consortium of Social Research Institutes of Yugoslavia. This survey design yielded greater numbers of respondents over 60 years of age than did the 1985 Croatian survey and, unlike the earlier survey, included unemployed persons. Unemployed persons and respondents over 69 years of age (the upper limit in 1985) were eliminated from our analysis to increase the comparability of the two samples. We analyze data from 1989 for the three largest republics, Croatia (N = 2,040), Bosnia (N = 1,569), and Serbia (N = 2,617). The addition of Bosnia and Serbia in the 1989 survey allow us to compare Croatia with the more nationally heterogeneous Bosnia and the more homogeneous Serbia. Questions relevant for the analysis of Yugoslav self-identification were repeated in both the 1985 and 1989 surveys. Minor modifications in the wording of some questions in 1989 do not appear to have affected the results significantly in that similar models fit the data from both 1985 and 1989. Dependent Variable The dependent variable, self-identification, is measured by answers to a question asking the respondents' national identification. Most people answered Croat, Serb, Moslem, or some other nationality or ethnicity. In 1985 in Croatia, 10.6 percent of respondents responded "Yugoslav" (see Table 2); by 1989 this figure had dropped to 9.0 percent. In 1989 in Bosnia, the level of Yugoslav self-identification was more than 50 percent higher than in Croatia (14.4 percent), while in Serbia it was about 50 percent lower than in Croatia (4.6 percent). 90 AMERICAN SOCIOLOGICAL REVIEW Independent Variables Our analysis suggests that four sets of factors influence the likelihood of identifying as a Yugoslav: modernization, political participation, demographic factors, and majority/minority status in a republic. Modernization. We measure modernization using variables that tap urbanism, education, and access to the media. In 1985 respondents were asked if they lived in a village, town, or city. The majority answered city, with substantial minorities reporting that they lived in towns and villages. In 1989 respondents were asked if they lived in a village, a village center, a town, a town center, a city, or a regional center. The 1989 question format generated a better distribution of responses than did the 1985 format. It was impossible, however, to reclassify the 1989 responses to approximate the 1985 distribution and also remain faithful to the verbal options as they were presented in 1989. We chose to collapse the 1989 responses into the verbally analogous categories of the 1985 survey—village (= 1), town (= 2), and city (= 3)—resulting in a somewhat different distribution than in 1985. The average residential location in both 1985 and 1989 is a town. The slight reduction in the mean of urbanism between 1985 and 1989 in Croatia is likely an artifact of the change in question format. To measure the effects of education and access to the media, we use the level of educational attainment of respondents and a measure of how frequently they read the news. Education is coded as years completed and averaged 10.0 in Serbia and 9.9 in Croatia in 1989. In Bosnia, average years of schooling was only slightly lower at 9.6 years. The measure of reading the news is based on two questions asked in each survey. In the 1989 survey, newspaper reading is measured by a question asking whether respondents read the newspaper never (= 1), monthly (= 2), weekly (= 3), or daily (= 4). Later in the questionnaire, respondents were asked to identify their three most common leisure activities. Some identified "reading news" as their most important leisure activity (= 4), others as their second most important activity (= 3), or their third most important activity (= 2), or not at all (= 1). Responses to these two questions were summed, creating an index of reading the news that ranges from 2 to 8. In 1985, respondents were asked to respond on the same 4-level scale as in 1989 to questions asking how often they read newspapers and, separately, news magazines. These two variables from the 1985 survey were summed to again create an index of reading the news ranging from 2 to 8. The index for Croatia has a slightly higher mean value in 1985 than in 1989, and is slightly lower in Bosnia and Serbia in 1989 than in Croatia. Political participation. Participation in the political system of Yugoslavia is expected to increase the likelihood of identifying as a Yugoslav. We measure three types of political involvement: membership in the LCY, holding office in workplace organizations, and holding office in community organizations. Each is coded as a dichotomous variable. Membership in the LCY was not an elite status; individuals who were LCY members should not be equated with LCY officials, who held elite positions and who disproportionately enjoyed the privileges to which such power provided access (Massey, Hodson, and Sekulic 1992). Nevertheless, Party membership should still indicate a greater commitment to explicitly articulated state goals. For both survey years, LCY membership is coded "yes" for those who either were members at the time of the survey or were members in the past. Considerably higher percentages report being, or having been, members of the LCY in Bosnia and Serbia than in Croatia. Our other two measures of political involvement are holding office in political organizations in the workplace and in the community. Workplace and community organizations were frequently dominated by Party members and served as conduits for Party goals and agendas. Such organizations, however, were also frequently "captured" by local interests operating outside official Party mandates (Bilandzic 1985). In 1985 questions about holding office in workplace organizations were asked separately for worker self-managed enterprises and for other enterprises. Positive answers to either yielded a code of 1 (0 if both answers were negative). In 1989 a summary question asked if the respondent occupied an elected position at the workplace (yes =1). Similarly, in 1985 separate questions were asked pertaining to holding office in organizations in the community dealing with either social or political activities. Office holding in either yielded a code FAILED SOURCES OF A COMMON IDENTITY IN YUGOSLAVIA 91 of 1. In 1989 these questions were combined, but a later question asked if respondents were active in community organizations. A positive response to either of these questions in 1989 resulted in a code of 1. In 1989 in Croatia, a higher percentage of respondents indicated they held office in workplace organizations than in 1985, but there were similar levels of participation in community organizations in 1985 and 1989. The higher percentage reporting participation in work organizations in 1989 than in 1985 may have occurred because the 1985 questionnaire was administered at the workplace where such participation would have been harder to manufacture. Bosnians and Serbians reported higher levels of participation in community organizations than Croatians (as well as higher levels of Party membership). Participation in work organizations was more nearly equal across the three republics in 1989. Demographic factors. The third factor we expect to influence identifying as a Yugoslav is represented by the demographic characteristics of age and nationally-mixed parentage. The average age in each subsample ranges from 38 to 40. We determined nationally-mixed parentage by comparing a respondent's answers to questions about the nationality of his or her mother and father. In Croatia in 1985, 10.0 percent of respondents reported that their parents were of different nationalities. For 1989 this figure is 10.2 percent. Nationally-mixed parentage occurred at. only a slightly lower level in Bosnia (9.0 percent). The level of nationally-mixed parentage was significantly lower in Serbia at 5.8 percent. Majority/minority status. The final factor we expect to influence identification as a Yugoslav is the position of respondent's parents as a majority or minority nationality in their republic. We expect persons of minority parentage to self-identify as Yugoslavs at a higher rate than persons of majority parentage. Persons of Croatian parentage were in the majority in Croatia and persons of Serbian parentage were in the majority in Serbia. In Bosnia, Moslems were numerically dominant, followed by Serbs and Croats, although all three groups were well represented. METHOD We use logistic regression to analyze the pattern of self-identifying as a Yugoslav based on modernization, political participation, demographic factors, and majority/minority status. Logistic regression is appropriate for a binary dependent variable and allows utilization of both categorical and continuous independent variables. The regression coefficients from a logistic regression can also readily be translated into easily interpretable odds indicating the change in the likelihood of the dependent variable (identifying as a Yugoslav) given a unit shift in an independent variable. RESULTS The logistic regression coefficients estimating the effects of the modernization, political participation, demographic factors, and majority/ minority status on Yugoslav self-identification for the combined sample are presented in Table 3. This model also yields coefficients estimating the net contrasts in Yugoslav self-identification between Croatia in 1985 and 1989 and between Croatia and the republics of Bosnia and Serbia in 1989. The model estimated for the total sample is highly statistically significant—the variables in the model reduce the chi-square from the baseline model (with only the intercept included) by 1,573. Five of the 10 independent variables are statistically significant at the .001 level. Modernization theories of identity formation find support in a significant urban residence effect, but the effects of the other two modernization variables, education and reading the news, are not significant. Party membership strongly increases the likelihood of Yugoslav self-identification. Participation in community organizations also has a significant positive effect on identifying as a Yugoslav, but participation in work organizations has no significant effect. Both of the demographic variables, age and nationally-mixed parentage, are highly significant.2 2 Non-linearity in the age and education effects was evaluated using 5-year cohorts and individual years of educational attainment. Dummy variables representing these categories were added to the equation separately for age and education. No significant nonlinearities were found for either age or education in either 1985 or 1989: Chi-square for age categories (d.f. = 9) was 12.12 in 1985 and 13.86 in 1989; chi-square for education categories (d.f. = 18) was 24.77 in 1985 and 27.56 in 1989. Of the 54 coefficients tested to evaluate nonlinearity 92 AMERICAN SOCIOLOGICAL REVIEW Table 3. Coefficients for the Logistic Regression of Yugoslav Self-Identification on Selected Independent Variables: Yugoslavia, 1985 and 1989 Independent Variable Coefficient Standard Error Odds Modernization Urbanism .463*** (.056) 1.59 Education -.020 (.014) .98 Read news .019 (.036) 1.02 Political Participation Communist Party .472*** (.096) 1.60 Work organization .015 (.106) 1.01 Community organization .293* (.123) 1.34 Demographic Factors Age -.034*** (.004) .97 Nationally-mixed parentage 2.449*** (.115) 11.58 Majority/Minority Status Croat parentage -.949*** (.127) .39 Serb parentage -.003 (.121) 1.00 Other homogenous parentage (baseline) .000 1.00 Net Contrasts Croatia 1985 .116 (.121) 1.12 Croatia 1989 (baseline) .000 — 1.00 Bosnia 1989 .296* (.133) 1.35 Serbia 1989 -1.048*** (.148) .35 Intercept -2.24*** X2(d.f. = 13) 1,572.7*** -2 log likelihood 4,501 Number of cases 9,845 *p < .05 **p < .01 ***p < .001 (two-tailed tests) Logistic regression coefficients are more in-terpretable when translated into odds by using the logistic coefficient as the exponent for the natural log function. For example, for nationally-mixed parentage, e2449 equals 11.58. The latter coefficient indicates that, net of the other factors in the model, a one-unit shift in nationally-mixed parentage (from not having nationally-mixed parentage to having nationally-mixed parentage) increases the odds of identifying as a Yugoslav by more than 11 times. The effect of age can be interpreted as indicating in age and education effects in 1985 and 1989, only 2 were statistically significant at the .05 level, and these formed no interpretable pattern. approximately a 3 percent decrease in the likelihood of identifying as a Yugoslav with each year of advancing age. The odds of self-identifying as a Yugoslav increase by about 60 percent for each increasing step of urbanism from village to town to city. Those who participate in community organizations are about one-third more likely to self-identify as a Yugoslav than those who do not participate. The odds of identifying as a Yugoslav are approximately 60 percent higher for Party members than for non-members. To evaluate change across time in Yugoslav self-identification, we compared respondents in the 1985 Croatian sample to respondents in the 1989 Croatian sample. In 1985, Croatians were about 12 percent more likely to self-identify as Yugoslavs than in 1989, but this coefficient is not statistically significant and therefore may not be a reliable estimate. Differences between republics in Yugoslav self-identification are indicated by the contrasts between Croatia and Bosnia and between Croatia and Serbia in 1989. Bosnians were 35 percent more likely to self-identify as Yugoslavs than were Croatians, and this difference is statistically significant at the .05 level. Serbians were only 35 percent as likely to identify as Yugoslav as Croatians in 1989, and this difference is significant at the .001 level. The effect of Croatian parentage on self-identification as a Yugoslav is significant at the .001 level. However, since Table 3 is based on the combined sample across the 1985 and 1989 surveys and across Croatia, Bosnia, and Serbia, this coefficient cannot be used to evaluate the hypothesis of defensive Yugoslavism which suggests that minority status in a republic will encourage self-identification as a Yugoslav. The appropriate tests of this hypothesis is presented later in Table 4, where the model is evaluated separately for each republic and time period. It is possible that the various factors in our model of Yugoslav self-identification have different effects across time or across republics rather than uniform effects as assumed for the model evaluated in Table 3. Possible differences in the model across time and republics are evaluated in Table 4. This model is also required in order to test the hypothesis that minority groups within republics have a greater likelihood of identifying as Yugoslavs. The model fits each of the four subsamples well and FAILED SOURCES OF A COMMON IDENTITY IN 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 Croatia 1989 Survey Independent Variable 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* i.oo#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/Minority Status Croat parentage .29*** .30** 1.22*" 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 interaction effects 18.2 (reference group) 42.9*** 34.2*** (d.f. = 10) *p < .05 **p < .01 ***p < .001 (two-tailed r-tests) V < .05 t1> < .01 tt1> < .001 (two-tailed ŕ-tests indicating a significant difference when compared to Croatia in 1989) coefficient is not significantly different from 1.00, but is significantly different from Croatia in 1989. is statistically significant in each case; the contribution to chi-square (d.f. = 10) ranges from 581.4 in Croatia in 1985 to 256.1 in Serbia in 1989. We used the 1989 survey in Croatia as a baseline for evaluating temporal and regional differences. We computed a Chow test to evaluate changes over time by selecting only the two Croatian subsamples and evaluating the contribution of a full set of interactions that allow the regression coefficients to differ in 1985 and 1989. These interactions reduce the chi-square (d.f. = 10) by 18.2, which is not statistically significant at the .05 level. The only interaction that is individually significant is for participation in work organizations, which increases self-identification as a Yugoslav in Croatia in 1985, but not in 1989. This contrast suggests that between 1985 and 1989 partici- pation in work organizations became a less effective foundation for Yugoslav identification. This interpretation must be made with some caution, however, given that the overall contrast between 1985 and 1989 is not significant. In contrast to the negative findings for temporal differences, the Chow tests for regional differences between Croatia and Bosnia in 1989 and between Croatia and Serbia in 1989 are statistically significant at the .001 level (with d.f. = 10, x2 = 42.9 and 34.2, respectively). Several interactions are individually significant. Urban residence has a larger positive effect on Yugoslav self-identification in Bosnia, a less economically developed region, than in Croatia. In Croatia in 1989, each level of increasing urbanism increases the odds of self-identifying as a Yugoslav by a factor of 1.17 (which is not significantly different from 1.00). 94 AMERICAN SOCIOLOGICAL REVIEW In Bosnia, each level of increasing urbanism increases the odds of self-identifying as a Yugoslav by a factor of 2.21. In Serbia, the urban residence effect is statistically significant, but it is not significantly different from the effect in Croatia. A second difference in the models involves regional differences in the effects of Communist Party membership. Party membership increases Yugoslav self-identification most in Croatia in 1989, less in Bosnia, and in Serbia it has no effect at all. The difference between Croatia, where Party membership more than doubles the likelihood of identifying as a Yugoslav, and Serbia, where it has no effect at all, is statistically significant at the .05 level. The third significant interaction involves the negative effect of homogenous Croatian parentage on Yugoslav self-identification in Croatia, where Croatians are the majority, and its positive effects in Bosnia and Serbia, where Croatians are a minority. The higher level of Yugoslav self-identification associated with Croatian parentage is most dramatic in Serbia, where Croatians are a small minority. In Serbia, having Croatian parentage increases the likelihood of identifying as a Yugoslav by over 15 times. Respondents with Serb parentage were also more likely to identify as Yugoslavs if they lived as a minority in Croatia rather than as a member of the majority in Serbia, but this contrast is not statistically significant. DISCUSSION Identifying as a Yugoslav can occur through several possible routes—through modernization, political participation, demographic factors, and majority/minority status. Our analysis provides evidence for all four bases of Yugoslav identification. Modernization effects are evident in the influence of urban residence in all three republics examined. Urbanism is one of the strongest and most consistent determinants of Yugoslav identification. This finding is consistent with the long-standing recognition that city living is associated with cosmopolitan attitudes (Wirth 1956). The declining influence of urbanism in Croatia between 1985 and 1989 may reflect the growth of Croatian nationalism during this period in the cultural and political capital of Zagreb. The lack of influence of education on self-identification as a Yugoslav, however, does not support modernization theory. Increased educational attainment had been thought to be an important factor in helping to generate an increasingly cosmopolitan outlook and to facilitate the emergence of a common identification as Yugoslavs. That education did not increase Yugoslav self-identification may attest to republic-level control of curriculum content and the teaching of separate national histories in each republic (see Jelavich 1983). Any increase of cosmopolitan attitudes associated with higher education appears to have been cancelled out by the attraction of national identities as a basis for social identity and political mobilization. The control of newspapers by local and republic-level Party organizations may also be responsible for the limited influence of reading the news on Yugoslav self-identification. The limited effect of education and access to the mass media in encouraging Yugoslav identification thus suggests the success of the regional Parties' efforts to represent nationalist interests and to condemn Yugoslavism. The image of partisan Yugoslavism based on political participation in Communist Party, workplace, or community political structures is supported by the finding that Party members were more likely to identify as Yugoslavs. Similarly, in at least some instances, participation in workplace or community organizations also increased the likelihood of identifying as a Yugoslav. Despite the Party's commitment to a multinational society and the growing nationalist agenda of republic-level Party organizations, those directly involved in the Party, at least outside of Serbia, were more likely to subscribe to the view that Communist Party members should be above the provincial attachments of the past and that their personal loyalties to the concept of a unified Yugoslavia should provide an example for others to follow. The significance of age in the analysis supports the notion of demographic Yugoslavism. Young people were more likely to reject national identities in preference for identifying as a Yugoslav. In this manner they sought symbolic entry into the milieu of modernism associated with the broader European culture. Also notable is the powerful effect of nationally-mixed parentage, which reduced allegiance to national identities and allowed an opening for identification as a Yugoslav. Other demographic factors such as migration and increased contact with other nationalities may also have FAILED SOURCES OF A COMMON IDENTITY IN YUGOSLAVIA 95 been encapsulated in the effects of nationally-mixed parentage. The concept of Yugoslav identity as a defensive strategy for minority nationalities is supported by higher rates of Yugoslav identification among those with Croat parents in Bosnia and, most dramatically, Serbia. A similar pattern, though not statistically significant, exists for Serbs, with Yugoslav identification being more likely for Serbs living in Croatia as a minority group than for Serbs living in Serbia as a majority group. The difference between Croats and Serbs in this regard provides additional support for the hypothesis of defensive Yugoslavism for minority nationalities. Serbs were numerically, politically, and militarily dominant in the former Yugoslavia, and it appears that they felt less need to subscribe to Yugoslavism, whether they lived inside or outside Serbia. Croats, by contrast, when living outside Croatia, were likely to take on the more neutral and defensive posture of identifying as Yugoslavs. The higher aggregate level of Yugoslav identification in ethnically-mixed Bosnia than in more homogenous Croatia or Serbia lends further support to the image of Yugoslav identification as a defensive strategy for minorities. CONCLUSIONS Only a relatively small proportion of people in the former Yugoslavia ever expressed the social identity of being Yugoslavs. Many of the social and political underpinnings of an emergent shared identify, however, were in place and operating to increase the likelihood of Yugoslav identification. These included powerful influences for urban residence, intermarriage, and political participation, even participation in a Communist Party apparatus that itself was ambivalent about promoting a Yugoslav identity. It will remain unknown if these forces, given sufficient time, would ever have created a strong collective identity for the citizens of the former Yugoslavia as common members of a unified nation-state. While this failure to establish a shared identity among the people of this region cannot be said to explain the disintegration of Yugoslavia, it is apparent that a shared identity was not much in evidence as a mediating mechanism sustaining Yugoslavia through difficult transitions or slowing its disintegration into warring national camps. The civil and military conflicts now underway in the former Yugoslavia and threatening to erupt in other parts of Eastern Europe and the former Soviet Union unmistakably exhibit elements based on nationalist fears, resentments, and hostilities. Emerging leaders of increasingly independent republics and states in these areas identify themselves with particular nationalities and claim to represent the interests of all those similarly identified, both within and outside currently defined boundaries. The end of Soviet hegemony and the eclipse of communist parties throughout Eastern Europe have provided new reminders of the strength of such nationalist feelings and claims. The Yugoslav experience suggests that the path to a shared identity, either through political persuasion or through industrialization and development, is neither simple nor assured. In Yugoslavia, strong Communist Party leadership and the transformation of the economy from an agricultural to an industrial base were expected to erode the traditional bases for national differences. When worker self-management was introduced and the role of the LCY at the federal level diminished, however, economic nationalism and the creation of bastions of regional political strength were made possible. Rather than explicitly attacking nationalism as a source of divisiveness and instability, Yugoslavia's leaders, especially at the republic level, endorsed nationalism as a right within the context of equality among the nations of Yugoslavia. The policies of Yugoslavia's leaders, at least as far as the national question is concerned, seemed both expedient and prudent. By emphasizing economic development, workplace democracy, economic and gender equality, tolerance for national differences, and equal legal rights of all citizens, Tito and his colleagues assumed that time was on their side and that the crises the state would inevitably face could be cast in other than nationalist terms. For the former-Yugoslavia, history has proven these assumptions wrong. The playing out of conflicting nationalist forces in the 1990s in other parts of Eastern Europe and in the former Soviet Union provides a profound challenge to the political and social will of the people of these areas. And understanding these forces provides a profound challenge to sociological theory. 96 AMERICAN SOCIOLOGICAL REVIEW DusKO Sekuuc is Associate Professor of Sociology and Professor of International Studies at George Mason University. He served as Editor in Chief of Revija za Sociologiju (Sociological Review), the journal of the Croatian Sociological Association, and President of Zagreb University. His main fields of interest are social stratification, nationalism, economic sociology, and the problems of transformation of East European Societies. Recent publications include "Dissolution of Yugoslavia: International Consequences" (in NATO: The Challenge of Change, edited by J. Simon, National Defense University, 1993) and ''Nationalism Versus Democracy: Legacies of Marxism " f International Journal of Politics, Culture and Society, vol. 6(1), 1992, pp. 113-32). Currently he is finishing a book titled Nationalism and the Dissolution of Yugoslavia. Garth Massey is Professor of Sociology at the University of Wyoming. He is a Fulbright Research Fellow at Godollo University in Hungary for the 1993-1994 academic year studying changing social structure in rural Hungary. His research interests include social change in socialist and formerly socialist societies, labor, and international political economy. His collaborations with Sekulic and Hodson include "Political Affiliation and Social Mobility in Socialist Yugoslavia" (in Research in Social Stratification and Mobility, vol. 11, edited by R. Althauser and M. Wallace), and "National Tolerance in the Former Yugoslavia " f American Journal of Sociology, forthcoming in May 1994). Randy Hodson is Professor of Sociology at Indiana University. His main research interests are in social stratification and the sociology of work. He is currently studying ethnic relations and changing patterns of social organization in the former Yugoslavia. He is also using survey techniques and the descriptions provided in workplace ethnographies to study the maintenance and defense of dignity in the workplace. His recent publications include "Is Worker Solidarity Undermined by Autonomy and Participation?" (American Sociological Review, vol. 58, 1993, pp. 398-416) and with Teresa A. Sullivan, The Social Organization of Work, (Wadsworth, 2nd ed., forthcoming 1995). REFERENCES Alba, Richard D. 1990. Ethnic Identity: The Transformation of White America. New Haven, CT: Yale University. Archdeacon, Thomas J. 1983. Becoming American: An Ethnic History. New York: Free Press. Banac, Ivo. 1984. The National Question in Yugoslavia. Ithaca: Cornell University . ------------. 1992. "The Fearful Asymmetry of War: The Causes and Consequences of Yugoslav Demise." Daedalus 121(2): 1-22. Belanger, Sarah and Maurice Pinard. 1991. "Ethnic Movements and the Competition Model: Some Missing Links." American Sociological Review 56:446-57. Bilandzic, Dusan. 1985. Historija Socijalisticke Federatívne Republike Jugoslavije (History of the Socialist Federal Republic of Yugoslavia). Zagreb, Yugoslavia: Skolska Knjiga. Bozic, Ivan, Sima Cirkovic, Milorad Ekmecic, and Vladimir Dedijer. 1973. Istorija Jugoslavije (History of Yugoslavia). Beograd, Yugoslavia: Prosveta. Breuilly, John. 1982. Nationalism and the State. New York: St. Martin's. Burg, Steven L. 1983. Conflict and Cohesion in Socialist Yugoslavia. Princeton, NJ: Princeton University. Chirot, Daniel. 1988. Social Change in the Modern Era. New York: Harcourt, Brace, Jovanovich. Cohen, Lenard. 1982. "Balkan Consociationalism: Ethnic Representation and Ethnic Distance in Yugoslav Elites." Pp. 29-62 in At the Brink of War and Peace: The Tito-Stalin Split in a Historic Perspective, edited by W. S. Vucinich. New York: Brooklyn College. Cohen, Lenard and Paul Warwick. 1983. Political Cohesion in a Fragile Mosaic. Boulder, CO: Westview. Collins, Randall and David V. Waller. 1992. "The Geopolitics of Ethnic Mobilization: Some Theoretical Projections for the Soviet Bloc." Paper presented at a conference sponsored by George Mason and George Washington Universities on "The Legacies of the Collapse of Marxism," 22-25 Mar., Washington, DC Connor, Walker. 1984. The National Question in Marxist-Leninist Theory and Strategy. Princeton, NJ: Princeton University. Cviic, Christopher. 1990. "The Backgrounds and Implications of the Domestic Scene in Yugoslavia." Pp. 89-119 in Problems of Balkan Security, edited by P. Shoup. Washington, DC: Wilson Center. Davis, Horance. 1978. Toward a Marxist Theory of Nationalism. New York: Monthly Review. Denitch, Bogdan. 1976. Legitimation of a Revolution. New Haven, CT: Yale University. —---------. 1991. Limits and Possibilities. Minneapolis, MN: University of Minnesota. Deutsch, Karl W. 1969. Nationalism and Its Alternatives. New York: Alfred A. Knopf. Devetak, Silvo. 1988. The Equality of Nations and Nationalities in Yugoslavia: Successes and Dilemmas. Vienna, Austria: Wilhelm Braumuller. Djilas, Aleksa. 1991. The Contested Country: Yugoslav Unity and the Communist Revolution, 1919-1953. Cambridge, MA: Harvard University. Edelman, Murray. 1971. Politics as Symbolic Action. New York: Academic Press. FAILED SOURCES OF A COMMON IDENTITY IN YUGOSLAVIA 97 Enloe, Cynthia H. 1973. Ethnic Conflict and Political Development. Boston, MA: Little, Brown and Company. Gertz, Clifford. 1973. The Interpretation of Cultures. New York: Basic Books. Hechter, Michael. 1976. "Ethnicity and Industrialization: On the Proliferation of the Cultural Division of Labor." Ethnicity 3:214-24. Hodson, Randy, Dusko Sekulic, and Garth Massey. Forthcoming. "National Tolerance in the Former Yugoslavia." American Journal of Sociology (May 1994). Horowitz, Donald L. 1985. Ethnic Groups in Conflict. Berkeley, CA: University of California. Isaacs, Harold R. 1975. Idols of the Tribe. New York: Harper and Row. Jelavich, Charles. 1983. "Serbian Textbooks: Toward Greater Serbia or Yugoslavia?" Slavic Review 42:601-19. Kardelj, Edvard. 1960. Razvoj Slovenackog Nacionalnog Pitanja (Development of the Slovenian National Question). Beograd, Yugoslavia: Kultura. Kedouire, Elie. 1960. Nationalism. London: Hutchinson. Kohn, Hans. 1961. The Idea of Nationalism. New York: MacMillan. Lederer, Ivo J. 1969. "Nationalism and the Yugoslavs." Pp. 396-438 in Nationalism in Eastern Europe, edited by P. F. Sugar and I. J. Lederer. Seattle, WA: University of Washington. Lydall, Harold. 1989. Yugoslavia in Crisis. Oxford, England: Clarendon. Massey, Garth, Randy Hodson, and Dusko Sekulic. 1992. "Political Affiliation and Social Mobility in Socialist Yugoslavia." Pp. 233-58 in Research in Social Stratification and Mobility, vol. 11, edited by R. Althauser and M. Wallace. Greenwich, CT: JAI. Nielsen, Francois. 1985. "Toward a Theory of Ethnic Solidarity in Modern Societies." American Sociological Review 50:133-49. Olzak, Susan. 1992. The Dynamics of Ethnic Competition and Conflict. Stanford, CA: Stanford University. Petro vie, Ruza. 1983. "The National Composition of the Population." Yugoslav Survey 24 (3 Aug.):21-34. Ragin, Charles C. 1979. "Ethnic Political Mobilization." American Sociological Review 44:619-35. Ramet, Pedro. 1984. Nationalism and Federalism in Yugoslavia, 1963-1983. Bloomington, IN: Indiana University. Rokkan, Stein. 1975. "Dimensions of State Forma- tion and Nation Building." Pp. 562-600 in The Formation of Nation States in Western Europe, edited by C. Tilly. Princeton, NJ: Princeton University. Rusinow, Dennison. 1977. The Yugoslav Experiment, 1948-1977. Berkeley, CA: University of California. ------------. 1985. "Nationalities Policy and the 'National Question.'" Pp. 131-65 in Yugoslavia in the 1980s, edited by P. Ramet. Boulder, CO: Westview. Sekulic, Dusko. 1990. "Recruitment for Elite Positions." Sociologija 32:105-25. Seroka, Jim and Rados Smiljkovic. 1986. Political Organizations in Socialist Yugoslavia. Durham, NC: Duke University. Shoup, Paul S. and George W. Hoffman. 1990. Problems of Balkan Security. Washington, DC: Wilson Center. Smith, Anthony D. 1986. The Ethnic Origins of Nations. Oxford, England: Basil Blackwell. Stanovcic, Vojislav. 1988. "History and Status of Ethnic Conflicts." Pp. 23-40 in Yugoslavia: A Fractured Federalism, edited by D. Rusinow. Washington, DC: Wilson Center. Terzuolo, Eric R. 1982. "Soviet-Yugoslav Conflict and the Origins of Yugoslavia's Self-Management System." Pp. 195-218 in At the Brink of War and Peace: The Tito-Stalin Split in a Historic Perspective, edited by W. S. Vucinich. New York: Brooklyn College. Tilly, Charles. 1975. The Formation of National States in Western Europe. Princeton, NJ: Princeton University. Tomasevic, Jozo. 1975. The Chetniks. Stanford, CA: Stanford University. Tome, Gregor. 1988. "Classes, Party Elites and Ethnic Groups." Pp. 58-77 in Yugoslavia: A Fractured Federalism, edited by D. Rusinow. Washington, DC: Wilson Center. Tudjman, Franjo. 1981. Nationalism in Contemporary Europe. Boulder, CO: East European Monographs, University of Colorado. Voirst, Milton. 1991. "The Yugoslav Idea." New Yorker (18 Mar.):58-79. Vuskovic, Boris. 1982. "Tko su Jugoslaveni?" (Who Are the Yugoslavs?). Nase Terne 26:1701-23. Waters, Mary C. 1990. Ethnic Options: Choosing Identities in America. Berkeley, CA: University of California. Wirth, Louis. 1956. The Ghetto. Chicago, IL: University of Chicago. Zwick, Peter. 1983. National Communism. Boulder, CO: Westview.