Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and...

38
This article was downloaded by: [Thammasat University Libraries] On: 09 October 2014, At: 03:32 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Post-Soviet Affairs Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/rpsa20 Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses Alexi Gugushvili ab & Peter Kabachnik c a Bremen International Graduate School of Social Sciences (BIGSSS), University of Bremen, Wiener Straße/Celsiusstraße, D-28334 Bremen, Germany b Center for Social Sciences, Tbilisi, Georgia c College of Staten Island – The City University of New York (CUNY), New York, NY, USA Published online: 05 Aug 2014. To cite this article: Alexi Gugushvili & Peter Kabachnik (2014): Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses, Post-Soviet Affairs, DOI: 10.1080/1060586X.2014.940697 To link to this article: http://dx.doi.org/10.1080/1060586X.2014.940697 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

Transcript of Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and...

Page 1: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

This article was downloaded by: [Thammasat University Libraries]On: 09 October 2014, At: 03:32Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Post-Soviet AffairsPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/rpsa20

Stalin is dead, long live Stalin? Testingsocialization, structural, ideological,nationalist, and gender hypothesesAlexi Gugushviliab & Peter Kabachnikc

a Bremen International Graduate School of Social Sciences(BIGSSS), University of Bremen, Wiener Straße/Celsiusstraße,D-28334 Bremen, Germanyb Center for Social Sciences, Tbilisi, Georgiac College of Staten Island – The City University of New York(CUNY), New York, NY, USAPublished online: 05 Aug 2014.

To cite this article: Alexi Gugushvili & Peter Kabachnik (2014): Stalin is dead, long live Stalin?Testing socialization, structural, ideological, nationalist, and gender hypotheses, Post-SovietAffairs, DOI: 10.1080/1060586X.2014.940697

To link to this article: http://dx.doi.org/10.1080/1060586X.2014.940697

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

Page 2: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 3: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Stalin is dead, long live Stalin? Testing socialization, structural,ideological, nationalist, and gender hypotheses

Alexi Gugushvilia,b* and Peter Kabachnikc

aBremen International Graduate School of Social Sciences (BIGSSS), University ofBremen, Wiener Straße/Celsiusstraße, D-28334 Bremen, Germany; bCenter for SocialSciences, Tbilisi, Georgia; cCollege of Staten Island – The City University of New York

(CUNY), New York, NY, USA

(Received 19 May 2014; accepted 19 June 2014)

Recently, there has been a renewed focus on analyzing post-Soviet memory,including the rekindling of debate on contemporary perspectives of JosefStalin. Most notably, the publication of The Stalin Puzzle has helped bringattention to the persistence of positive accounts and admiration, along withambivalent and contested images, of the former dictator of the Soviet Union.Using survey data and multivariate statistical methods, we test five broadhypotheses – socialization, structural, ideological, nationalist, and gender – toascertain what factors might shape people’s attitudes toward Stalin in Georgia.Our analysis reveals that elderly, poor men from rural areas have the mostpositive associations of Stalin, whereas young, wealthier women from cities,those who are open to privatization, and perceive Russia as Georgia’s biggestthreat judge Stalin negatively. Counterintuitively, non-Georgian minoritiesshow higher esteem for Stalin than Georgians. We envision that the effects ofcohort replacement, economic development, and urbanization will decreasepositive perceptions of Stalin in years to come.

Keywords: Stalin; post-Soviet memory; Georgia; socialization; nationalism

Introduction

Scholars have identified the phenomena of nostalgia for the Soviet Union and the

Soviet era as a particularly salient feature of modern post-communist societies

(Boym 2001; Mendelson and Gerber 2005, 2006; Munro 2006; Popov 2008; Lee

2011; Khinkulova 2012; Sullivan 2013; Vegso 2013). Symbols of the Soviet era

manifest themselves in a variety of discourses, be it discussing the communist past

nostalgically, military parades that allude to the glory days of the Soviet empire

and its former superpower status, or the reproduction of the cult of the Great

Patriotic War.1 The symbols of the Soviet era also materialize in various places

like monuments (Forest and Johnson 2002), billboards and advertisements

(Danilova 2005; Kara-Murza 2013), or else in literature (Medvedev 2005;

Yasmann 2006; Marsh 2007) and television (Khinkulova 2012). In this article, we

q 2014 Taylor & Francis

*Corresponding author. Email: [email protected]

Post-Soviet Affairs, 2014

http://dx.doi.org/10.1080/1060586X.2014.940697

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 4: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

turn our attention to another such symbol – Josef Stalin – and the narratives and

opinions people have about him.

In the twenty-first century, Stalin is making his presence felt yet again. Surveys

have demonstrated that Stalin’s popularity remained high or has in fact risen in

recent years in the successor states to the Soviet Union (Mendelson and Gerber

2006). For instance, in Russia, Stalin repeatedly polls as one of the most celebrated

figures in Russian history (Slysz 2010), his image adorns a variety of products,

Volgograd has already reverted to its previous toponym – Stalingrad – 6 days a

year, and there are renewed discussions for a referendum that could put this name

back on the map (Taylor 2014). Meanwhile, new Stalin monuments have appeared

in several parts of Georgia, the Stalin Museum in Gori still idealizes the hometown

hero, and it is not unusual to hear toasts dedicated to the man of steel.

Beyond symbolic manifestations and anecdotal evidence, a recent study has

drawn attention to, and spurred considerable debate about, people’s attitudes toward

Josef Stalin throughout contemporary post-Soviet space (de Waal 2013a). Survey

results have revealed high levels of admiration for a person considered to be one of

the most notorious dictators in history. For some commentators, this became proof

that the people surveyed had failed the “Stalin test” (Mendelson andGerber 2006) or

else that these results indicated that there is a “Stalin puzzle” that needed solving

(deWaal 2013a). Questions undoubtedly arise: How to approach Stalin’s continued

presence into the twenty-first century, over 60 years after his death? How shall we

digest this information? How do we contend with this phenomenon? In this article,

we attempt to explain attitudes toward Stalin in his birthplace, Georgia, where

respondents to the aforementioned surveys demonstrated by far the highest support

for him. Therefore, in addition to the mainstream explanations of political attitudes

– socialization, social structural, ideological, and gender theories – we also look at

the ethno-nationalistic dimension of how Stalin is perceived in Georgia.

We utilize the survey data commissioned by the 2012 Carnegie survey and

administered by the Caucasus Research Resource Centers (CRRC 2012). Although

the descriptive results have been made public, in order to better situate the survey

responses, one needs to go beyond the surface results as captured in the charts

detailing the percentage of respondents answering each question (see de Waal

2013a, 7–11).While this is certainly a necessary and important starting point, to get

a sense ofwhat factors are informing people’s attitudes, onemust dig deeper into the

survey data. To this end, we use bivariate and multivariate statistical methods to

analyze the data, search for valid explanations of attitudes about Stalin, and draw

conclusions for existing and future studies.Our objectives in this article are twofold.

First, we wish to move past descriptive data in the search for determinants using

advanced statistical methods. Second, we test the applicability of the general

theories of public attitudes as well as nationalist sentiments that might shape

people’s attitudes toward Stalin in Georgia. This article is structured as follows: in

the following section, we provide an overview of the main issues relating to

contemporary understandings of Stalin in post-Soviet space and in Georgia. Next,

we detail our major hypotheses about why people hold the particular political

attitudes they do. Then, we provide an overview of our data and methods, followed

A. Gugushvili and P. Kabachnik2

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 5: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

by the bulk of the article, the analysis of the empirical data, along with our

predictions about what will happen with attitudes about Stalin in the future. We

close with a summary and discuss the possible implications of our findings.

Understanding Stalin today

Before elaborating the specific hypotheses of our study, we consider it essential to

describe what Stalin means today for people in the former Soviet Union. Many

scholars examine the issue of cultural memory in post-Soviet space, deeming it a

topic necessitating further research (Etkind 2013). These studies focus on how

Russia, in the first two decades after the collapse of the USSR, has not gone

through a process of de-Stalinization, failing to both properly reassess its past and

adequately commemorate and acknowledge the victims of the Soviet regime.

However, much of the research that does not only focus on this topic theoretically,

draws upon the experience of former dissidents, the intelligentsia, former

prisoners, and writers and artists (Etkind 2013). When everyday perspectives are

considered, or current understandings of the past explored, it is often done

anecdotally (Hochschild 1995; Smith 1996; Satter 2012; for a non-Russia

example, see Rosenberg 1996). By contrast, we systematically analyze survey data

about people’s attitudes towards Stalin in order to augment the rich, though often

methodologically less rigorous,2 qualitative data already available in the literature.

This serves two functions. First, it helps us better verify whether certain discourses

and claims being made about what opinions of Stalin signify are isolated themes or

are more broadly representative of the general public. Second, we can identify

what factors contribute to heightened levels of appreciation of Stalin or more

critical views about the Soviet dictator.

There has been some exploration of popular attitudes and analysis of survey

data on sensitive topics such as violence and conflicts in the North and South

Caucasus (Kolossov and Toal 2007; O’Loughlin, Toal, and Kolossov 2008; Bakke,

O’Loughlin, and Ward 2009; O’Loughlin, Kolossov, and Toal 2011; Toal and

Grono 2011; Toal and O’Loughlin 2013), as well as research analyzing popular

understandings of the Soviet era in Russia (Rose and Carnaghan 1995; Mendelson

and Gerber 2005, 2006, 2008; Sullivan 2013). Other related research has examined

a wide range of topics, frommeasuring the power of nostalgia for the Soviet Union

(Munro 2006; Popov 2008; Lee 2011; Sullivan 2013) to documenting the attitudes

of Russian youth toward the United States (Mendelson and Gerber 2008).

Within this literature, the analysis of attitudes towards Stalin has been

explored. However, few studies have consulted the broader literature on attitude

formation and persistence and employed rigorous statistical analysis to uncover

variables pertaining to people’s perceptions of Stalin (Reisinger et al. 1994; Munro

2006; Mendelson and Gerber 2008; Sullivan 2013). Furthermore, to date there has

been no attempt at multivariate analysis of survey data about attitudes toward

Stalin in Georgia. Bakradze (2013) provides the only analysis thus far of the

survey results in Georgia. He offers a general summary of the factors influencing

attitudes toward Stalin. Counterintuitively, he states that “a more precise

Post-Soviet Affairs 3

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 6: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

breakdown of Georgian attitudes by age and social group does not reveal

significant differences across society” (Bakradze 2013, 52). Age in particular is

often offered as a component that correlates with positive or negative feelings

toward Stalin, namely that older people are more likely to be proud of Stalin. He

goes on to say that education plays little role, residents of Tbilisi tend to be more

critical of Stalin, and attitudes toward Russia only provide a small level of

indication of support for Stalin. In this article, we will test these pronouncements

and other proffered explanations.

In the next section, we propose five broad hypotheses to test against the survey

data results. The development and selection of the five hypotheses – socialization,

structural, ideological, nationalist, and gender – was shaped by the Carnegie

report’s analyses (de Waal 2013a), the broader literature on attitudes about Stalin

in Russia (as discussed above), the prevailing “conventional wisdom” in Georgia,

and our own experiences speaking with people in Georgia about Stalin. The five

hypotheses encapsulate a variety of indices, including ethnicity, religiosity,

education level, age, gender, political and economic ideology, socioeconomic

status, and settlement type. These hypotheses will be considered and either

validated or challenged based on the results of various statistical methods.

Hypotheses

In order to explain attitudes toward Stalin in Georgia, we refer to major theories on

the micro-foundations of political attitudes – socialization, social structural, and

ideological perspectives. We also consider the role of Georgian nationalism in

explaining why respondents in this country are more likely to express approval of

Stalin than respondents in neighboring Armenia, Azerbaijan, and Russia (de Waal

2013a). In addition, we hypothesize that gender has specific implications in

relation to the Soviet dictator and that the outlined factors also manifest differently

within consecutive birth cohorts.

Socialization theory

We propose that socialization theory is the main framework for understanding

attitudes toward Stalin in Georgia. Each individual born in a society acquires

values and beliefs from the distinctive development of her or his birth cohort

through peer-group socialization, the content of formal education, and

idiosyncratic historical experiences such as war, revolution, or technological

change (Ryder 1965). Political attitudes are shaped at a relatively young age when

individuals are entering adulthood (Jennings andMarkus 1984). More specifically,

the impressionable years hypothesis implies that during late adolescence and early

adulthood, individuals are highly susceptible to attitude change and their

susceptibility significantly declines thereafter (Krosnick and Alwin 1989).

According to the aging-stability hypothesis, stable attitudes emerge soon after

early adult experiences, and differences between cohorts are quite durable in the

face of social change (Alwin and Krosnick 1991). Probably the main reason why

A. Gugushvili and P. Kabachnik4

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 7: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

attitudes adopted during childhood tend to persist over the life course is that

individuals in their adulthood are more likely to believe in the importance and

certainly of their convictions, while their perceived quantity of attitude-relevant

knowledge is greater than during their late adolescence or early adulthood (Visser

and Krosnick 1998).

Generational differences in social and structural conditions are one of the main

explanations why most social change progresses slowly. Many scholars of post-

socialist democratic transition have reached a consensus that a cohort replacement

over a long period of time would be necessary to create a civil society free of the

communist indoctrination (e.g., Finkel, Humphries, and Opp 2001; Pop-Eleches

and Tucker 2014). Individual studies in post-Soviet space also show that the

youngest generation is more favorably disposed toward popular participation

in decision-making, while the older generation exhibits admiration for a “strong

leader” (Hahn and Logvinenko 2008). Although the most appealing political

system for Russians, regardless of age, is a Soviet system, nearly half of

respondents born before the 1930s preferred an unreformed Soviet political regime,

whereas among those born after the 1960s that proportion was only 10% (Colton

and McFaul 2002). We know that cohorts born in the first half of the twentieth

century in Soviet Georgia were brought up in an environment whereby people were

indoctrinated to communist ideas and exposed to idealized representations of Stalin

(Kitaevich 2014), at least before Khrushchev’s “Secret Speech” and de-

Stalinization measures after 1956 (Kemp-Welch 1996).

The youngest generations are least influenced by socialization that glorified

Stalin and communist achievements in general, as they grew up during the less

authoritative 1960s and 1970s and in Mikhail Gorbachev’s perestroyka and

glasnost’ periods. In one of the few studies that analyzed multivariate framework

attitudes toward Stalin, Reisinger et al. (1994) inquired into how strongly

respondents agreed with the following statement: “These days, Stalin is not given

adequate credit for his accomplishments.” Their findings suggest that positive

attitudes toward Stalin indeed are highest for those born before 1925 and lowest

among individuals born after 1966. However, the same survey in Ukraine and

Lithuania suggests that the oldest cohorts have less positive perceptions of Stalin

than those born in 1926–1935. The most likely explanation for this trend is that

Lithuania was independent before World War II, and therefore, those born earlier

than 1926were not indoctrinated by the Soviet regime during their formative years,

whereas in Ukraine the oldest survey respondents might remember the famine of

1932–1933 and place blame for it on the late Soviet ruler (Bilinsky 1999). Since

Georgia was annexed by Soviet Russia in 1921, in our survey, we do not have

respondents with living memory of the first Georgian republic (1918–1921).

Following from the preceding review of the theoretical and empirical literature, we

therefore expect that the older cohorts will be most supportive of Stalin in Georgia.

Hypothesis 1: Positive attitudes toward Stalin are strongest among the oldest

birth cohorts and rapidly decline among those born in the second

half of the twentieth century.

Post-Soviet Affairs 5

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 8: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Social structural theory

Our first hypothesis assumes that individuals have different attitudes about Stalin

because of their generational affiliation and socialization. The levels of

indoctrination markedly differed in the first half of the twentieth century from the

period when Georgia became an independent country. However, it would be

naıve to completely explain attitudes of a certain individual toward the Soviet

system and Stalin based on her or his birth date (Ekman and Linde 2005). The

level of education is a good example to illustrate this point because it might have

implications for both socialization and social structural theories. If we assume

that an individual’s education has an effect on attitudes toward Stalin, then

expanding educational opportunities might be responsible for variance of

attitudes among different generations. Here, we are not talking about the content

of the curriculum at different historical times, which itself would be a

generational effect, but rather about the share of people with primary, secondary,

and tertiary educational attainment across time. Education is only one example,

other areas of social and economic change included urbanization, labor market

structure, material welfare, and other developments related to transition from a

more traditional, agrarian society to the more industrialized and urbanized

Georgian Soviet Socialist Republic (Rose and Carnaghan 1995). Social structural

theory assumes that values and beliefs can change throughout the life course,

and attitudes are shaped by the current structural conditions in which individuals

find themselves.

There is a scarcity of literature on how social and structural conditions are

associated with attitudes toward Stalin or other dictators, but perceptions of

democracy and authoritarianism seem to be related to a complex interplay of

institutional, cultural, and economic factors (Inglehart 1988). Leaving institutional

determinants aside, better educated individuals are expected to have more critical

attitudes toward strict norms and authoritative leaders because education equips

people with the skills that are necessary to understand political processes and to

communicate preferences to those who make decisions even in authoritative

states. Hahn (1991), analyzing attitudes in Russia just before the collapse of the

Soviet Union, concludes that educational attainment had the most important effect

on political attitudes. It seems that individuals with limited education are different

political actors than those with higher academic attainment. We should expect that

more educated individuals, controlling for their birth date, are less likely to

support Stalin because they are more aware of the mass executions and other

atrocities committed by the Soviet regime under Stalin’s leadership. A highly

educated public is less likely to tolerate the old Stalinist system’s heavy-handed

terror and controls (Bahry 1993).

Another important social structural explanation of attitudes toward Stalin can

be rural-urban differences. The increased complexity of social structure that

accompanies urban residence could be more difficult to accommodate within the

framework of an authoritarian state system. In urban areas, people are better

informed about actions implemented by governments. It was argued that in Russia

A. Gugushvili and P. Kabachnik6

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 9: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

during the early 1990s, a salient urban-rural split in the distribution of political

attitudes existed, and that rural residents were generally less supportive of political

reforms (Wegren 1994) and were more collectivist and egalitarian than urban

Russians (Reisinger, Miller, and Hesli 1995). Studies which analyze democratic

satisfaction in a number of post-socialist countries also show that in big cities and

national capitals people have been more critical of the quality of the existing

democracy than in small and medium-sized towns and rural communities

(Neundorf 2010; Pop-Eleches and Tucker 2014). When asked if “these days Stalin

is not given adequate credit for his accomplishments,” urban residents are less

likely to view Stalin favorably (Reisinger et al. 1994). The same can apply to

differences of support for Stalin between rural and urban settlements in Georgia. It

is expected that in rural areas Stalin is more popular than in the capital city and

other urban settlements. It has been shown that in consecutive national elections

since independence, Tbilisi residents were more likely to cast votes for opposition

parties and presidential candidates (George 2014).

In addition to education and urbanization, what might also shape attitudes

toward Stalin in Georgia is respondents’ socioeconomic status. It has been argued

that it was not the progress of modernization but its discontinuities that prompted

social and political change in the Soviet Union (Ruble 1987). For instance, higher

education had expanded far more rapidly than the existing occupational structure

could accommodate, which led to substantial underemployment of the educated

labor force. Similar trends were also observed in Georgia’s educational and

occupational expansion (Gugushvili 2013). In this article, we make an

assumption that Stalin is perceived as a symbol of the individual’s attitude

toward various aspects of the Soviet political and economic system (Cohen 1982).

In this regard, many of those who are dissatisfied with their current position in the

post-socialist socioeconomic hierarchy might evince positive attitudes toward

the Soviet system and Josef Stalin. For instance, using data for 16 post-socialist

societies, it was shown that an individual’s income is strongly and positively

related to support for a democratic political order, and it tends to have its

strongest effect if economic development is high (Andersen 2012). Studies on

nostalgia for the USSR also reveal that the poor are more likely to harbor positive

feelings about the Soviet past (Sullivan 2013). Munro (2006) shows that an

individual’s evaluation of the current economic system has a weak negative

influence on nostalgia, but more importantly the family economic situation in

the past has the strongest effect on attitudes toward the former system.

Individuals’ socioeconomic conditions could be even more important for attitudes

toward Stalin in Georgia, as the country experienced dramatic economic decline

in the 1990s and the poverty level has remained consistently high thereafter

(Gugushvili 2011).

Hypothesis 2: Attitudes toward Stalin are stratified based on an individual’s

current socioeconomic conditions – more educated, urban, and

wealthier individuals are against Stalin, while less educated,

rural, and poor individuals are pro-Stalin.

Post-Soviet Affairs 7

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 10: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Ideological preferences

The next theoretical approach assumes that the perception of the Soviet dictator is

linked to other sets of attitudes. Individuals view Stalin as part of a larger belief

system (Brooks and Bolzendahl 2004). People as a rule are not knowledgeable

about most issues at a given point of time, yet they are able to express preferences

when necessary or if asked to. Rational individuals sort their preferences by

answering two questions: Who am I? andWhat should I do? Their cultural identity

helps individuals to know what kind of preferences they are expected to have as

well as their desirability in comparison to others’ beliefs (Wildavsky 1987). We

can assume that democratic values in socialist and post-socialist societies stem

from a breakdown in confidence in the authoritarian regime. This trend was further

fueled by the diffusion of Western ideas about the democratic political system,

which was perceived as a powerful alternative political model (Gibson 1996). It is

likely that the residual authoritarian values and beliefs in post-socialist societies

stem from system-internal experiences, whereas the democratic elements stem

from system-external experiences (Roller 1994).

After the dissolution of the Soviet Union and the disillusionment with

communism, ideas of a freemarket economy and democratic political system began

to take hold. Like other Western practices, ideas, and items such as religion, Levis,

and Snickers, democracy became fashionable in post-socialist societies (Gibson

1996). In 1992, in one of the first nationally representative surveys conducted in

independent Georgia, 63% of respondents agreed that the creation of a market

economy largely free from state control was right for the country’s future. In the

same survey, more than half of the respondents expressed satisfaction with the

development of democracy (Reif and Cunningham 1993). In 2012, after 20 years of

post-socialist transition, most people in Georgia (83%) agree that democracy is

preferable to any other form of political system (CRRC 2012). The gradual spread

of ideas about the fundamental characteristics of the politico-economic system is

obvious in Stalin’s home country. The reason why we expect that these ideological

preferences will be reflected in views toward the Soviet dictator is that the

fundamental principles of a market economy and democracy are incompatible with

Stalinism. Individuals who perceive democracy and a market economy as the

preferred politico-economic system should not exhibit strong positive attitudes

toward Stalin and his policies.

Hypothesis 3: Support of an ideology that is not compatible with Stalinism –

democracy and a market economy – leads to a negative

perception of Stalin.

Nationalism theory

As outlined earlier in this article, one of the major findings of The Stalin Puzzle (de

Waal 2013a) is that attitudes toward the late Soviet dictator are significantly and

systematically more positive in Georgia than in Armenia, Azerbaijan, and Russia.

Beyond socialization, social structural, and ideological factors, we speculate that

A. Gugushvili and P. Kabachnik8

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 11: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Stalin’s ethnicity might be one of the major covariates that shapes his positive

perception (Zedania 2011). Many scholars have identified high levels of

nationalism in the country, even during the Soviet period (Jones 1992; Nodia

1995; Chinn and Kaiser 1996; Slider 1997; Kaufman 2001). It is also argued that

virulent ultranationalism dominated for a time in the 1990s (Suny 2006). There are

reasons to believe that many Georgians still identify with and are proud of their

ethnic kinsman – Ioseb Jugashvili (Bakradze 2013) – who is arguably the most

well-known Georgian in world history. Showing pride in Stalin, or displaying a

portrait of him in one’s home, could be a nationalist practice that reproduces

Georgian national identity. As nationalism scholars relate, narratives and symbols,

such as the image of Stalin and the stories told about him, help to embody

everyday performative nationalism (Bhabha 1990). Stalin’s role as a national

symbol can reproduce national identity both in explicit and ritualistic (Hobsbawm

1983) and banal ways (Billig 1995). If Stalin is understood as a particularly

Georgian symbol, and thus a key aspect of Georgia identity and history, then

attitudes toward Stalin can be seen as a measure of Georgian nationalism.

Furthermore, the unique role of Stalin in Georgian nationalism could be

affected by the 1956 demonstrations to mark the third anniversary of Stalin’s death

and consequently by the “9 March Massacre” unleashed by the Soviet military in

Tbilisi. There is evidence of strong nationalist sentiment present during these

protests in Georgia (Suny 1994). Though the degree to which Stalin was

imbricated in the nationalist fervor is difficult to quantify and is debated (Kozlov

1999; Blauvelt 2009), there was clearly a connection. It is argued that along with

the demands for the rehabilitation of Stalin, protestors raised the issues of self-

determination and the independence of Georgia. Some commentators also draw

parallels between the 1956 events and the 1989 independence movement massacre

for the formation of Georgian national identity (Vashakidze 2013). In spite of

Khrushchev’s de-Stalinization campaign, Stalin remained, perhaps more than in

other Soviet republics, a presence in domestic (photographs and portraits hanging

in homes and workplaces) and public space (monuments) in Georgia throughout

the Soviet period and still to this day. This indicates that Stalin remained a central

element of Georgian identity and the inability to limit his presence in the

landscape could be explained by Georgian nationalism.

Nationalism, especially its ethnic component, is often expressed through

religious nationalism and indeed one prominent feature of national identity in the

Georgian case is religion. Zedania (2012) highlights the role of religious

nationalism in the country and notes the central role of the Orthodox Church for

Georgian identity, as for many being Georgian “lies in ‘being Orthodox’” (125). It

is also argued that religious-nationalistic views and value orientations in Georgian

society prevail over civil and liberal ones (Chelidze 2014). Furthermore, there are

some direct links between religious nationalism and the Stalin cult in Georgia, as

many clergy are grateful for Stalin’s re-institution of the Church during World

War II. This narrative also implies that the death of Stalin in 1953 prompted the

new repressions of the Church, often ignoring the fact that Stalin’s regime was an

atheist dictatorship that was responsible for the destruction of numerous churches

Post-Soviet Affairs 9

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 12: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

in Georgia and elsewhere in the Soviet Union. Probably the most vivid

manifestation of the position of the Georgian Church is an interview with its

leader, Ilia II, who polls repeatedly indicate is the most trusted figure in Georgia,

who stated that “Stalin is an exceptional person, people like him are rarely born”

(Metskhvarishvili 2013).

However, an additional dimension to be considered here are alternative forms

of nationalism in Georgia, which particularly intensified after the Rose Revolution

(Vachridze 2012). These features of national identity emphasize civic nationalism

as well as anti-Russian rhetoric heard from the ruling elites. Indeed, the pride in

Stalin in Georgia is arguably linked to ethnic nationalism as opposed to its civic

form (Zedania 2011). On 25 June 2010, the monument to Stalin in his birthplace

that withstood the process of de-Stalinization in 1956 and Georgian independence

after 1989 was removed from the main square of the town of Gori. If people view

Stalin as more Soviet or Russian, rather than Georgian, or as someone who harmed

Georgian society, then their nationalism would manifest itself differently. If Stalin

is primarily equated with Russia, and Russia is seen as a colonizer both before and

during the Soviet period, then those people who perceive Russia to be Georgia’s

biggest threat will judge Stalin more harshly than those who do not.

Hypothesis 4: Ethnic and civil national identity, respectively, will positively

and negatively shape attitudes toward the Soviet dictator, as

Stalin is viewed by Georgians through the lens of nationalism.

Gender gap and within-cohort effects

One additional vector that might matter for attitudes toward Stalin is gender. There

is a scarcity of studies that show how males differ from females in their attitudes

toward authoritarian leaders, although the available scholarship highlights an

existing “gender gap” in a broad array of attitudes related to the use of force and

violence (Shapiro and Mahajan 1986). Smith (1984), by compiling six different

sources of survey data in the USA, shows that gender differences usually result

from males choosing pro-force options such as capital punishment, military

intervention, and weak gun control laws, while females either have no opinion or

select anti-force response choices. There is no conclusive agreement why females

differ in their political ideas, but researchers across disciplines have advanced

several possible interpretations. It has been suggested that women biologically

have stronger compassion, moral strength, and maternal instincts (Togeby 1994).

In the social psychology literature, these gender differences have been attributed

to social dominance orientation (SDO), which is linked to an individual’s gender.

SDO implies general preference for group inequality, and has been shown to be

more prevalent among men (Pratto, Stallworth, and Sidanius 1997).

We do not find any studies that investigate specifically the role of gender in

attitudes toward Stalin or other dictators in the post-Soviet space, but some

empirical evidence also supports the idea that males are more likely to be in favor

of the Soviet system and consequently Stalin and Stalinism. For instance, in one of

the rare studies of the gender gap in political attitudes, Carnaghan and Bahry

A. Gugushvili and P. Kabachnik10

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 13: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

(1990) show that females in the Soviet Union on average tended to oppose higher

spending on defense (although no gender-based differences were revealed in

support for the Soviet use of force). Munro (2006) showed that females in Russia

were less likely than males to express a desire to see the return of the communist

system. Males in Russia also have been shown to exhibit more anti-American

views than females in Russia (Mendelson and Gerber 2008). Furthermore, in

Georgia, the Caucasus Barometer (CRRC 2012) demonstrates that males are

marginally more likely to declare that in some circumstances authoritarianism is a

better political regime than democracy (11.0% vs. 8.8%). Therefore, gender can be

a significant covariate of attitudes toward Stalin.

Hypothesis 5: Controlling for other factors, males in Georgia are expected to

express more positive attitudes toward Stalin than females.

After reviewing what we expect from different birth cohorts, social and

structural conditions, ideological preferences, Georgian nationalism, and gender in

terms of attitudes toward Stalin, we also outline how these characteristics might

have different effects according to age. There are reasons to believe that the

hypothesized factors have different manifestations across different generations.

Perhaps the major achievement of Stalin, for which he is credited by many in

post-Soviet societies, is the victory in World War II. This might also imply that

males as direct participants in military activities would be closer to Stalin than

females, whose involvement in the war was less directly combat related. If so,

gender differences might be more intense among the older generations rather

than recent cohorts without living memory or first-hand accounts of the war from

their parents. We also expect that differences between social and structural

characteristics are more salient in post-Soviet times and therefore among the

youngest generations. Inequality levels in terms of rural-urban differences and

income distribution were much less pronounced before the dissolution of the

Soviet Union (e.g., see Fuchs and Demko 1979). The gap between urban and rural

lifestyles and socioeconomic conditions of poorer and wealthier individuals was

not markedly salient. In recent years, these differences are more intensively

manifested (World Bank 2009). In addition, we also speculate that ideological

convictions are more relevant for elderly individuals because in the era when Stalin

and the socialist system were idealized, having positive attitudes toward a free

market economy and democracy also would imply a negative perception of the

Soviet regime and Stalin.

Data and variables

In 2012, the Carnegie Endowment for International Peace commissioned the

Caucasus Research Resource Centers (CRRC 2012) to include several questions

designed to elicit people’s perceptions of Stalin in the South Caucasus. This was

the first time a nationally representative survey on this topic had been undertaken

in the birthplace of the Soviet dictator. The questions about Stalin were integrated

into the annual Caucasus Barometer survey, conducted every fall since 2006,

Post-Soviet Affairs 11

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 14: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

which allows for the testing of links between the perceptions of Stalin on the one

hand and individuals’ various demographic, economic, and attitudinal

characteristics. The Caucasus Barometer respondents answer face-to-face survey

questions, participation in the survey is voluntary and anonymous, and the results

are representative for the adult (18 þ) population of Georgia. In total, 2502

respondents in Georgia completed the survey in 2012.3 Below, we describe four

questions on attitudes toward Stalin that serve as the dependent variables in our

study, followed by an overview of independent variables that are used to explain

the perceptions of Stalin in Georgia.

Dependent variables

We employ four dependent variables from the survey that inquire into how

strongly respondents agree with the following statements: (1) “Stalin was a wise

leader who brought power and prosperity to the Soviet Union.” This question

intends to understand respondents’ attitudes toward Stalin as a political leader and

a manager who created one of the strongest states in world history. (2) “For all

Stalin’s mistakes and misdeeds, the most important thing is that under his

leadership the Soviet people won the Great Patriotic War.” This statement implies

that the victory in World War II can be considered more important than Stalin’s

misdeeds. (3) “Stalin was a cruel, inhuman tyrant, responsible for the deaths of

millions of innocent people.” The third question allows insight into whether

respondents know and acknowledge Stalin’s responsibility in the Soviet terror. (4)

“Our people will always have need of a leader like Stalin, who will come and

restore order.” The last statement relates Stalin’s legacy to the contemporary and

future needs of people to have a dictator who can remedy a dysfunctional political

and economic system. Answer options for each question consist of “completely

disagree,” “mostly disagree,” “mostly agree,” “completely agree,” “don’t know,”

and “refuse to answer.”

Interestingly, 4.4% of the survey respondents declared that they did not know

who Stalin was. Our analysis suggests that the individuals who are unaware of

Stalin are overrepresented among rural residents, ethnic Azerbaijanis, and cohorts

born after 1975, but surprisingly level of education is not related to awareness of

Stalin. Among those who have heard about Stalin, roughly 2% of respondents

refused to answer the questions, while respectively 13.9%, 11.9%, 20.6%, and

22.9% of respondents chose the option “don’t know” for the described four

statements. It is telling that more than one-tenth of respondents would not provide

their opinion about whether Stalin was a wise leader or whether the victory in

World War II is more important than his mistakes. But even more intriguing is that

about one in five respondents do not know if Stalin was responsible for the deaths

of millions of innocent people and whether people will always have need of a

leader like Stalin. This suggests that either the awareness of Stalin and Stalinism in

Georgia is low or that people are reticent to reveal their real attitudes toward

Stalin. The analysis of respondents who did not know what to answer suggests that

A. Gugushvili and P. Kabachnik12

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 15: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

females, non-Tbilisi residents, less educated, Azerbaijanis, and cohorts born after

1975 were more likely to choose this response.

In our empirical study, we leave aside those individuals who do not know who

Stalin is, who refuse to answer the questions asked, or do not know what to think

about the statements. The responses of the remaining individuals in our sample are

presented in Figure 1. It is clear that the majority of respondents in the CRRC

survey agree with the first two statements. Only less than 20% of respondents who

answered the question disagree that Stalin was a wise leader, bringing power and

prosperity to the Soviet Union, while 41.1% completely agree with this statement.

The consensus is even higher that under Stalin’s leadership the Soviet people won

the Great Patriotic War, which is more important than his shortcomings, as about

90% of respondents agree with the statement (49.5% completely agree and 40.0%

mostly agree in Figure 1(b)). More diverse are respondents’ answers about

Stalin’s responsibility for the deaths of millions of innocent people and the need

for a leader like Stalin. The Soviet leader is perceived as a cruel and inhuman

tyrant by about 65.9% of respondents (32.2% completely agree and 33.7% mostly

agree in Figure 1(c)), while 10.3% completely disagree with the statement. For the

last variable (Figure 1(d)), the majority of respondents disavow the idea that

people will always need a leader like Stalin (about 62%), but more than one-third

of respondents (37.4%) believe that it will always be necessary for such a leader to

come and restore order (14.0% completely agree and 23.4% mostly agree). We

can infer from the descriptive statistics that many individuals acknowledge Stalin

as a wise leader and that the defeat of Germany in World War II is more important

Figure 1. Descriptive statistics of the dependent variables. Notes: The answer option “donot know” is excluded from the sample.Source: Authors’ calculations based on data from Caucasus Barometer (CRRC 2012).

Post-Soviet Affairs 13

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 16: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

than his mistakes, but at the same time the same people think that Stalin was a

cruel, inhuman tyrant, and that people will never have need of a leader like Stalin.

The share of individuals with these simultaneously positive and negative attitudes

toward Stalin is about 26.6% in our sample.

Independent variables

To test our hypotheses we employ 14 independent variables. Following the work

of Breslauer (1978), Rose and Carnaghan (1995), and Munro (2006), we derive

seven distinct birth cohorts or generations: (1) World War II (born before 1935,

5.0% of the sample); (2) the Stalinist (1935–1944, 10.0%); (3) the post-Stalinist

(1945–1954, 13.4%); (4) the social contract (1955–1964, 19.1%); (5) the glasnost

(1965–1974, 17.2%); (6) the transition (1975–1984, 18.4%); and (7) the

independence (1985–1994, 17.0%, reference category) generations. The older

cohorts are hypothesized to be most supportive of Stalin in Georgia. We employ

several variables to test hypothesis 2 (H2). The years of education vary from 0 to

25, with mean value of 12.7 (SD ¼ 3.2). The elaborated hypothesis suggests that

the higher the respondent’s education, the lower his/her positive perception of

Stalin. Two variables that are related to an individual’s education are their

comprehension of foreign languages – Russian and English [from no basic

knowledge ¼ 1 to advanced knowledge ¼ 4, with mean values 2.9 (SD ¼ 0.89),

and 1.5 (SD ¼ 0.86), respectively]. It is assumed that a higher proficiency in either

language – but especially in English, which is much less prevalent in Georgia

(Blauvelt 2013) – would be associated with negative attitudes toward Stalin. For

settlement type [Tbilisi – 21.8%, other urban areas – 39.4% (reference category),

rural areas – 38.8%], we expect that Tbilisi residents hold the most critical views,

while rural residents will have the least negative perceptions of Stalin.

The social structural hypothesis is further tested with individuals’ socio-

economic conditions, which are operationalized in three variables: labor market

position, number of household’s material items, and respondents’ subjective

perception of their households’ socioeconomic status. The unemployed, who

constitute one-fourth of our sample (23.9%), are expected to express positive

attitudes about Stalin. In order to derive households’ objective socioeconomic

conditions, we use information on whether the respondents have a number of items

and services in their households.4 The seven items retained after factor analysis are

summed to derive a cumulative index with values ranging from 0 (no listed items

in the household) to 7 (all listed items are available in the household), with a mean

value of 2.7 (SD ¼ 1.95). For the perception of the subjective socioeconomic

situation, we use information from the survey question about respondent

households’ placement on a subjective income rung – from the lowest possible

economic position in society (1) to the highest possible position (10), with a mean

value of 4.2 (SD ¼ 1.7).

H3, on ideological preferences, is tested with two questions related to the

political and economic systems. Respondents are asked to choose from the

following three statements: “democracy is preferable to any other kind of

A. Gugushvili and P. Kabachnik14

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 17: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

government” (82.7%); “in some circumstances, a non-democratic government can

be preferable” (9.7%); and “for someone like me, it doesn’t matter what kind of

government we have” [7.7% (reference category)]. We expect that those who

prefer democracy would exhibit negative attitudes toward Stalin, while

authoritarian preferences would increase support for the Soviet dictator.

Respondents’ economic ideology is measured by their subjective placement on

the following scale – “government ownership of business and industry should be

increased” (1) to “private ownership of business and industry should be increased”

(10). This variable has mean value of 4.6 (SD ¼ 3.0). We expect that supporters

of government ownership will sympathize with Stalin, whereas the opposite

attitudes should be observed among supporters of private property and

privatization.

H4, on the links between Georgian nationalism and attitudes toward Stalin, is

tested with independent variables on respondents’ ethnicity, religiosity, and their

perception of the biggest enemy of Georgia. We expect that ethnic Georgians

(88.1%, reference category) would exhibit stronger support for Stalin than

Armenians (4.6%), Azerbaijanis (4.7%), and other ethnicities living in Georgia

(2.7%). Religiosity is measured based on respondents’ answers to how religious

they are from “not at all religious” (1) to “very religious” (10), with a mean value

of 6.3 (SD ¼ 2.3). It is difficult to hypothesize how religiosity could affect

attitudes toward Stalin because, on the one hand, religious people might oppose

communist ideology and the atheism associated with it, but, at the same time, due

to close links between ethnic nationalism and the Georgian Orthodox Church, we

expect that religiosity has a positive effect on attitudes toward Stalin. Respondents

were also asked which country is currently the biggest enemy of Georgia. We code

dummy variable 1 if respondents answer Russia (32.9%). Those who say Russia is

the biggest enemy are expected to manifest negative attitudes toward Stalin.

Lastly, we test H5 on the importance of gender with a dummy variable for males.

Similar to other surveys in Georgia, females are overrepresented in our sample

(62.5%), but we hypothesize that males are the ones who will express more

positive attitudes toward Stalin.

Results

Multivariate analysis

Our univariate analysis of variances tentatively suggests that generational

belonging, education, foreign language proficiency, the type of settlement,

objective and subjective socioeconomic status, attitudes towards private property,

and perception of Russia might indeed affect attitudes toward Stalin (please see

the results of the ANOVA tests in Table B1). The goal of the multivariate analysis

in Table 1 is to identify the net effect of specific factors by simultaneously

controlling for all independent variables in generalized ordered logit models.

Before proceeding with the description of findings, it is important to clarify how to

interpret the presented regression output. The first panel indicates coefficients of

covariates which differentiate between the answer “completely disagree” in the

Post-Soviet Affairs 15

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 18: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Table

1.

Covariatesofattitudes

towardStalinin

Georgia:log-oddsfrom

generalized

logisticregressions.

Model

1Model

2Model

3Model

4

“Stalinwas

awiseleader

whobroughtpower

and

prosperityto

theSoviet

Union”

“Themostim

portantthing

isthat

under

Stalin’s

leadership

theSoviet

people

wonWWII”

“Stalinwas

acruel,inhuman

tyrant,responsiblefor

thedeathsofmillionsof

innocentpeople”

“Ourpeople

willalways

haveneedofaleader

like

Stalin,whowillcomeand

restore

order”

Coef.(robuststd.err.)

Coef.(robuststd.err.)

Coef.(robuststd.err.)

Coef.(robuststd.err.)

1.Completely

disagreevs.mostly

disagree,mostly

agree,completely

agree

Cohort1925–1934

1.64(0.78)**

13.2

(1.38)***

22.26(0.52)***

2.23(0.54)***

Cohort1935–1944

1.60(0.66)***

20.96(1.28)

21.78(0.49)***

1.43(0.30)***

Cohort1945–1954

1.39(0.48)***

0.55(0.84)

21.04(0.49)**

1.08(0.26)***

Cohort1955–1964

0.77(0.36)**

20.70(0.78)

20.95(0.48)**

0.66(0.22)***

Cohort1965–1974

0.37(0.31)

0.17(0.66)

20.34(0.51)

0.48(0.22)**

Cohort1975–1984

0.38(0.33)

0.15(0.67)

0.02(0.53)

0.42(0.21)**

Years

ofeducation

20.06(0.04)

20.16(0.08)**

0.01(0.04)

20.02(0.03)

Proficiency

inRussian

20.40(0.18)**

20.49(0.35)

0.02(0.14)

20.21(0.09)**

Proficiency

inEnglish

0.02(0.14)

0.72(0.26)***

0.06(0.15)

20.00(0.08)

Tbilisiresidence

0.30(0.27)

1.18(0.59)**

20.32(0.29)

20.48(0.15)***

Ruralresidence

0.84(0.38)**

1.64(0.55)***

20.68(0.24)***

0.07(0.16)

Unem

ployed

0.21(0.26)

20.69(0.52)

20.16(0.26)

20.05(0.15)

Materialitem

s20.04(0.07)

0.11(0.11)

20.07(0.06)

20.02(0.04)

Subjectiveincomerung

20.07(0.10)

20.79(0.16)***

0.06(0.06)

20.00(0.04)

Dem

ocracyispreferable

1.06(0.43)**

0.52(1.13)

0.57(0.33)*

20.41(0.29)

Authoritarian

ispreferable

0.84(0.53)*

21.39(1.01)

0.64(0.46)*

20.14(0.35)

Privatization

20.05(0.05)

20.19(0.08)***

0.10(0.03)***

20.10(0.02)***

Arm

enianethnicity

0.70(0.87)

21.02(1.54)

2.42(1.05)**

0.92(0.36)**

Azerbaijaniethnicity

20.47(0.96)

0.00(0.00)

15.2

(0.44)***

218.1

(0.62)***

Other

ethnicity

20.34(0.74)

1.04(0.84)

20.45(0.48)

0.43(0.40)

Religiosity

0.06(0.07)

0.12(0.14)

0.03(0.05)

0.02(0.03)

Russia

isthebiggestenem

y20.67(0.23)***

20.67(0.38)*

0.24(0.21)

0.03(0.13)

Gender

(male¼

1)

0.85(0.25)***

0.63(0.41)

20.17(0.20)

0.26(0.13)*

Constant

2.88(0.88)***

9.34(1.96)***

1.85(0.88)**

1.85(0.56)***

A. Gugushvili and P. Kabachnik16

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 19: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

2.Completely

disagree,mostly

disagreevs.mostly

agree,completely

agree

Cohort1925–1934

1.90(0.54)***

0.64(0.50)

22.29(0.36)***

1.85(0.37)***

Cohort1935–1944

1.79(0.36)***

1.32(0.40)***

21.81(0.29)***

1.75(0.29)***

Cohort1945–1954

1.08(0.29)***

0.93(0.34)**

21.35(0.27)***

0.91(0.26)***

Cohort1955–1964

0.54(0.25)**

0.93(0.30)***

21.21(0.26)***

0.62(0.25)**

Cohort1965–1974

0.36(0.24)

0.44(0.28)

20.85(0.26)***

0.47(0.25)*

Cohort1975–1984

0.45(0.23)*

0.18(0.26)

20.70(0.26)***

0.22(0.25)

Years

ofeducation

20.04(0.03)

0.07(0.03)**

20.00(0.02)

20.02(0.02)

Proficiency

inRussian

20.15(0.11)

20.25(0.11)*

0.24(0.09)***

20.19(0.09)*

Proficiency

inEnglish

20.00(0.09)

0.11(0.11)

20.08(0.09)

20.05(0.10)

Tbilisiresidence

0.30(0.17)

20.19(0.22)

20.06(0.17)

0.10(0.17)

Ruralresidence

1.01(0.19)***

0.42(0.22)*

20.34(0.14)***

0.42(0.15)***

Unem

ployed

0.24(0.17)*

0.26(0.21)

0.04(0.16)

0.08(0.16)

Materialitem

s20.04(0.04)

0.03(0.05)

0.06(0.04)

20.08(0.04)*

Subjectiveincomerung

20.15(0.05)***

20.35(0.06)***

0.05(0.04)

20.08(0.04)**

Dem

ocracyispreferable

0.42(0.31)

20.78(0.44)*

20.06(0.26)

20.07(0.27)

Authoritarian

ispreferable

0.62(0.38)*

20.45(0.51)

20.53(0.31)*

0.30(0.33)

Privatization

20.10(0.03)***

0.00(0.03)

0.07(0.02)***

20.08(0.02)***

Arm

enianethnicity

20.45(0.31)

0.88(0.60)

0.06(0.30)

0.35(0.28)

Azerbaijaniethnicity

0.00(0.73)

0.00(0.00)

1.43(0.66)**

16.9

(0.43)***

Other

ethnicity

20.00(0.43)

20.12(0.49)

20.04(0.37)

0.42(0.37)

Religiosity

0.00(0.03)

20.03(0.04)

0.01(0.03)

20.00(0.03)

Russia

isthebiggestenem

y20.34(0.14)**

20.33(0.17)**

0.13(0.13)

20.20(0.14)

Gender

(male¼

1)

0.55(0.15)***

0.17(0.17)

20.30(0.13)**

0.24(0.13)*

Constant

2.25(0.57)***

3.40(0.74)***

0.59(0.51)

0.33(0.52)

3.Completely

disagree,mostly

disagree,

mostly

agreevs.completely

agree

Cohort1925–1934

1.17(0.32)***

0.86(0.32)***

22.25(0.48)***

1.42(0.46)***

Cohort1935–1944

1.43(0.26)***

1.16(0.25)***

21.35(0.29)***

1.33(0.43)***

Cohort1945–1954

1.10(0.25)***

0.74(0.23)***

21.09(0.25)***

0.61(0.40)*

Cohort1955–1964

0.57(0.24)**

0.59(0.22)***

20.74(0.22)***

0.12(0.40)

Cohort1965–1974

0.54(0.23)**

0.51(0.21)***

20.48(0.22)**

20.10(0.44)

(Continued

)

Post-Soviet Affairs 17

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 20: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Table

1–continued

Model

1Model

2Model

3Model

4

“Stalinwas

awiseleader

whobroughtpower

and

prosperityto

theSoviet

Union”

“Themostim

portantthing

isthat

under

Stalin’s

leadership

theSoviet

people

wonWWII”

“Stalinwas

acruel,inhuman

tyrant,responsiblefor

thedeathsofmillionsof

innocentpeople”

“Ourpeople

willalways

haveneedofaleader

like

Stalin,whowillcomeand

restore

order”

Coef.(robuststd.err.)

Coef.(robuststd.err.)

Coef.(robuststd.err.)

Coef.(robuststd.err.)

Cohort1975–1984

0.37(0.23)

0.19(0.21)

20.39(0.21)*

0.04(0.41)

Years

ofeducation

0.03(0.02)

0.05(0.02)**

0.01(0.02)

0.03(0.03)

Proficiency

inRussian

20.02(0.08)

20.04(0.08)

0.21(0.09)**

20.05(0.12)

Proficiency

inEnglish

20.07(0.08)

0.01(0.08)

20.10(0.08)

20.12(0.16)

Tbilisiresidence

0.18(0.15)

0.13(0.14)

0.33(0.15)**

0.42(0.27)

Ruralresidence

0.41(0.14)***

0.20(0.14)

20.10(0.15)

0.89(0.22)***

Unem

ployed

0.24(0.14)*

0.14(0.13)

0.01(0.15)

20.09(0.24)

Materialitem

s20.03(0.04)

20.04(0.03)

20.01(0.04)

20.07(0.06)

Subjectiveincomerung

20.08(0.04)**

20.12(0.04)***

0.09(0.04)**

20.12(0.06)**

Dem

ocracyispreferable

0.36(0.25)

0.16(0.23)

0.09(0.27)

0.30(0.38)

Authoritarian

ispreferable

0.18(0.31)

0.28(0.29)

20.60(0.35)

20.24(0.51)

Privatization

20.07(0.02)***

20.03(0.02)*

0.04(0.02)*

20.10(0.03)***

Arm

enianethnicity

20.29(0.29)

20.12(0.28)

20.41(0.33)

20.13(0.43)

Azerbaijaniethnicity

1.11(0.59)**

15.8

(0.33)***

20.79(0.61)

1.53(0.63)***

Other

ethnicity

20.36(0.38)

20.28(0.35)

20.32(0.36)

0.90(0.46)**

Religiosity

20.02(0.03)

20.07(0.03)***

20.03(0.03)

20.04(0.04)

Russia

isthebiggestenem

y20.32(0.12)**

20.19(0.12)

0.26(0.12)*

20.19(0.19)

Gender

(male¼

1)

0.14(0.12)

0.11(0.11)

20.18(0.12)

0.29(0.20)

Constant

20.92(0.48)**

20.18(0.46)

21.04(0.50)**

21.79(0.76)***

Statistics

N1497

1511

1390

1367

AIC

3412

3035

3492

3402

McFadden’s

pseudoR

20.09

0.09

0.08

0.11

Notes:***,**,and*denote

statisticalsignificance

atthe0.01,0.05,and0.10levels,respectively.

Source:

Authors’calculationsbased

ondatafrom

CaucasusBarometer

(CRRC2012).

A. Gugushvili and P. Kabachnik18

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 21: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

corresponding dependent variable on the one hand and the categories “mostly

disagree,” “mostly agree,” and “completely agree,” on the other. These

coefficients show the impact of independent variables in getting a respondent

up from the first step of the four-step ladder in our dependent variable. The second

panel of the table contrasts “completely disagree”/“mostly disagree” on the one

hand with “mostly agree”/“completely agree” on the other, differentiating

respondents on the third and fourth steps of the ladder from those on the first two

steps. The same logic applies to the third and the final panel of the table. The

positive sign of predictors across three panels would suggest that independent

variables have a positive association with attitudes toward Stalin; by contrast, if

coefficients have a negative sign, then independent variables are negatively

associated with perceptions of Stalin.5

Starting with the analysis of H1 on socialization in Table 1, respondents’

generational belonging appears to be the strongest explanation of support for

Stalin. When compared to those born in 1985–1994, all earlier-born cohorts have

more positive attitudes toward Stalin. Expectedly, World War II (born in 1925–

1934) and immediate post-war Stalinist (1935–1944) generations are Stalin’s

strongest supporters, while attitudes among respondents born later in time trend

downward. World War II generation is especially against the statement that Stalin

was a cruel, inhuman tyrant, responsible for the deaths of millions of innocent

people (b ¼ 22.12, p , 0.01, in panel 2). It is also interesting that individuals

born in 1975–1984 (b ¼ 20.68, p , 0.01) still have more positive ideas about

Stalin than the independent Georgia generation. In H2, we assumed that education

would be negatively associated with attitudes toward Stalin, but years of education

exhibit only weak, if any, links with our dependent variables. Furthermore, better

educated people counterintuitively tend to agree that the more important thing

than the crimes committed by Stalin is that under his leadership the Soviet people

won World War II. Respondents’ self-reported proficiency in Russian has a

negative association with attitudes toward Stalin. In all three panels of Table 1,

log-odd coefficients have negative values, which clearly suggest that people with

advanced knowledge of the Russian language are more likely to completely

disagree or mostly disagree that Stalin was a wise leader who brought power and

prosperity to the Soviet Union.

As expected, rural residents are much more likely to express positive attitudes

toward Stalin. Most obviously this can be seen in the second panel of the table,

which contrasts “completely disagree”/“mostly disagree” versus “mostly

agree”/“completely agree” answer options. Panel 3 of the regression output also

suggests that Tbilisi residents are more likely than others to completely agree with

the statement that Stalin was a cruel, inhuman tyrant, responsible for the deaths of

millions of innocent people. Contrary to H2, objective socioeconomic conditions

operationalized with a dummy variable for unemployment and the number of

household items exhibits virtually no association with attitudes toward Stalin. The

unemployed are more likely to perceive that Stalin brought power and prosperity

to the Soviet Union, but this association only holds at the 10% significance level.

The subjective perception of socioeconomic status of respondents’ families

Post-Soviet Affairs 19

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 22: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

appears to be more important for our dependent variables, which can be seen in all

coefficients of the second and third panels in Table 1 that have negative signs and

are statistically significant.

When testing H3, we do not find that perceptions of democracy and

authoritarianism are linked with attitudes toward the Soviet dictator; however,

those who believe that democracy is the preferable political system are more likely

to choose answer options “mostly disagree,” “mostly agree,” and “completely

agree” than “completely disagree” when asked if Stalin was a wise leader. On the

other hand, in line with our hypothesis, a respondent’s views toward privatization

appear to be the significant predictor of attitudes toward the Soviet dictator. Those

who think that private ownership of business and industry should be increased are

much more likely to disagree that Stalin brought power and prosperity to the

Soviet Union and that people will always have need of a leader like Stalin.

Respondents who think that government ownership of business and industry

should be increased are likely to disagree that Stalin was responsible for the deaths

of millions of innocent people.

H4, on nationalism, is tested with the ethnicity variable, in which Armenian

and Azerbaijani respondents are compared with Georgians in their attitudes

toward Stalin. Unexpectedly, we find that non-Georgian ethnic groups are more

likely to express positive attitudes toward Stalin than do Georgians. It is also

interesting that Azerbaijani residents of Georgia, on the one hand, think that Stalin

was a cruel, inhuman tyrant, responsible for the deaths of millions of innocent

people, but, on the other hand, they are more likely to agree that people will always

have need of a leader like Stalin, who will come and restore order. Azeri

respondents are also more likely to think that Stalin was a wise leader and that the

victory in World War II is more important than all his shortcomings. Although

religious people are less likely to choose the answer option “completely agree” in

Model 2, the effect is not statistically significant for any other panel or dependent

variable. In line with our hypothesis, the perception that Russia is the biggest

enemy of Georgia is negatively associated with the perception that Stalin was a

wise leader. These same people also disagree that under Stalin’s leadership the

Soviet people won World War II. Ending with the analysis of H5, on gender, the

first and the second panels of Table 1 suggest that males are more likely to express

positive attitudes toward Stain, which is indicated with positive and statistically

significant values of coefficients from generalized ordered logit regressions in

Models 1, 3, and 4.

Overall, the results partially or fully confirm our hypotheses and suggest that

representatives of older generations, rural residents, people with poor Russian

language skills, subjectively deprived individuals, ethnic Azerbaijanis, those who

think Russia is not Georgia’s biggest enemy, those who support government

ownership of business and industry, and males exhibit stronger support for Stalin.

To understand the scale of the effect of these variables, we utilize the aggregated

index of positive perceptions of Stalin. As we describe in Appendix A, 0 points on

this variable means that for the four questions about Stalin, respondents hold

extremely negative ideas, while those who score 12 points respond completely

A. Gugushvili and P. Kabachnik20

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 23: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

positively on all four questions. This form of the dependent variable allows us to

see clearly how each independent variable is associated with attitudes towards the

Soviet dictator. Figure 2 provides a clear illustration of the specific effect of

independent variables.

As hypothesized, age provides the strongest explanation of the cumulative

index of attitudes toward Stalin. Respondents born before 1934 and between 1935

and 1944 score 8.6 (SD ¼ 0.18) and 8.0 (SD ¼ 0.14) in our dependent variables,

while the youngest cohorts born in 1985–1994 score only 5.3 (SD ¼ 0.15) points

in our cumulative Stalin support index. In Figure 2(b), moving up by one level in

proficiency in Russian leads to a 0.35 lower score in the aggregated dependent

variable. Those who claim that they know Russian at an advanced level score 6.4

(SD ¼ 0.12) in the index of Stalin support. Salient differences are also observed

between individuals living in rural and urban settlements. If rural residents score

on average 7.1 (SD ¼ 0.12) points, residents both in Tbilisi and other urban areas

score 6.3 and 6.7, respectively (SD ¼ 0.14 and 0.07). Subjective perception of

respondents’ household socioeconomic status is also an important predictor of

attitudes toward Stalin, as shown in Figure 2(d). Moving up one rung of the ladder

on this scale is associated with a 0.38 point reduction in the dependent variable.

Those who classify their families in the lowest two income rungs score 7.3

(SD ¼ 0.16) points, while those perceiving themselves on the seventh or eighth

rung score only 5.7 (SD ¼ 0.27) points in the aggregated Stalin index.

For H3, those who completely support the idea that private ownership of

business and industry should be increased (scoring 9 or 10 on this indicator) are

Figure 2. Linear predictions of covariates of attitudes toward Stalin (0 ¼ oppose Stalin;12 ¼ support Stalin). Notes: Bars represent 95% confidence intervals. Coefficients showthe effect when controlling for all other variables.Source: Authors’ calculations based on data from Caucasus Barometer (CRRC 2012).

Post-Soviet Affairs 21

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 24: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

much less likely to express positive perceptions of Stalin (6.3, SD ¼ 0.12) than

those who think that government ownership of business and industry should be

increased (7.3, SD ¼ 0.15). Unlike the previous results for H4 on the ethnic

differences in terms of perceptions of Stalin in Georgia, for the cumulative index

of four dependent variables in Figure 2(f) we do not observe statistically

significant differences between Georgians, Armenians, Azerbaijanis, and other

ethnicities living in Georgia. However, non-Georgian ethnicities score higher in

their support of Stalin, but the low number of observations for them produces large

standard errors, which prevents us from making any definitive conclusions.

Believing that Russia is the biggest enemy of Georgia is also associated with our

dependent variable. Those who consider that Georgia’s northern neighbor is the

biggest foreign threat to the country score 6.4 points (SD ¼ 0.12), which is about

0.5 points lower than the score of those individuals who do not perceive Russia as

Georgia’s biggest enemy. Finally, in Figure 2(h) we once again confirm the effect

of gender, as males express 0.45 points higher support for Stalin than females.

Cleavages within generations

The multivariate analysis presented above suggests that birth year and

generational attachment are the strongest explanations of variance in our

dependent variables. However, in addition settlement type, subjective income

rung, belief in private ownership, the perception of Russia as the biggest enemy

of Georgia, and gender are the key explanatory variables that are linked with

attitudes toward Stalin in Georgia. It is not clear if any cleavages exist within the

defined generations. This is a substantively and statistically important question

because it will allow one to see what important covariates, if any, make some

elderly respondents exhibit negative attitudes toward Stalin when most of the age

peers in their demographic group judge Stalin positively. More importantly, this

analysis allows us to observe why some individuals in the youngest demographic

group are positive about Stalin, when they are expected to be indifferent or

express negative attitudes toward the late Soviet leader. To answer this question,

we interact the key explanatory factors with respondents’ birth years, starting

from the 1920s and ending in the 1990s. The interpretation of Figure 3 is

straightforward: The main lines in the diagrams show the marginal effects of

covariates, while dashed lines give 95% confidence intervals. The effect is

statistically significant when confidence intervals do not cross the horizontal zero

reference line.6 Although we are not able to follow the developments in time

because attitudes might change across an individual’s life course, the analysis in

this section is helpful in understanding what to expect in the long run in terms of

attitudes toward Stalin.

Figure 3(a) shows marginal effects of settlement in rural areas according to the

respondent’s birth year. Our main analysis suggests that much more positive

attitudes toward Stalin are held in rural areas than in Tbilisi and other urban

settlements, but interaction effects showed that the within-generational gap

between rural and urban residents intensifies, especially for those born in the 1960s

A. Gugushvili and P. Kabachnik22

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 25: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

or later. Among individuals born in 1935–1944, urban-rural differences are

virtually non-existent (as shown in the 95% confidence interval crossing the zero

reference line), but for the cohorts born after the 1970s, rural residents score on

average 1 point higher than urban residents in our cumulative index of attitudes

toward Stalin. The reason why the rural-urban divide is more intense among

the recent generations is not entirely clear, but we can speculate that differences in

access to electronicmedia, levels of social capital, and other areas of socioeconomic

development are much more pronounced between rural and urban youth than is the

case for rural and urban elderly.

Next, we analyze generational differences in terms of the subjective

socioeconomic status of respondents’ households. Figure 3(b) indicates that for

individuals born after the 1970s, moving up one subjective income rung is

associated with about 0.3 points decrease in support of Stalin in our cumulative

dependent variable. In general, we know that those individuals who classify

themselves at the highest rung in terms of socioeconomic status hold negative

perceptions of Stalin, but the differences are especially apparent among the

Figure 3. The marginal effect of settlement, subjective income rung, attitudes towardprivatization, non-Georgian ethnicity, the perception of Russia, and gender on cumulativeindex of perception of Stalin across respondents’ birth years. Notes: Interaction effectsderive from models with the same specification as in Table 1. Dashed lines give 95%confidence interval.Source: Authors’ calculations based on data from Caucasus Barometer (CRRC 2012).

Post-Soviet Affairs 23

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 26: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

youngest generations, who were born after the 1970s. In fact the group which is

subjectively richest in Georgian society exhibits the lowest support of Stalin among

all other demographic and socioeconomic groups (b ¼ 3.7, SD ¼ 0.38). Our

explanation for this trend is that the youngest generations, who feel socially and

economically excluded from society, might think of Stalin in association with

the old communist regime and the relative prosperity of the Soviet Union. This

finding suggests that socioeconomic status might be one of the few important

vectors that explain the differences in perception of Stalin among the youngest

residents in Georgia.

We analyze the interactions between the respondents’ birth year and their

ideologies and nationalist variables in terms of attitudes toward private ownership

of business and industry and perceived international threat. Interestingly the

interaction effect in Figure 3(e) suggests that attitudes about Russia do make a

difference for perceptions of Stalin among the elderly. More specifically, those

who were born before the 1970s and believe that Russia is the biggest enemy of

Georgia exhibit 0.5–1 point lower support for the Soviet dictator. For instance, in

the 1945–1954 generation, those who answer “Russia” when asked which country

is the biggest enemy of Georgia score 6.5 (SD ¼ 0.32) in the support of Stalin

index, while for those who feel otherwise this score reaches 7.9 points

(SD ¼ 0.23). The perception of Russia is only important in understanding

attitudes toward Stalin in the generations that exhibited the highest support for the

Soviet dictator. This category of respondents is probably ideologically close to

Soviet dissidents who opposed the Soviet regime and its most notorious leader.

This speculation is also supported by Figure 3(c), which shows the interaction

effect between birth year and attitudes about private property. Again, moving one

step up in support for private ownership of business and industry reduces positive

perceptions of Stalin by 0.1 to 0.2 points for cohorts born before 1970. It appears

that among the elderly ideological differences are the most salient factor that

differentiates the supporters and opponents of Stalin in Georgia.

To analyze within-generational cleavages in ethnicity, we combine all non-

Georgians in our sample as a distinct category.7 Figure 3(d) suggests that

differences between Georgians and other ethnicities in their attitudes toward Stalin

become statistically significant for those born after 1970. However, the few cases

of non-Georgians produce large standard errors and do not reveal the real effect of

the ethnic vector in perceptions of Stalin. However, if anything, Georgians exhibit

weaker support for Stalin, and this ethnic divide is likely to be higher for the recent

generations than among the elderly. Finally, the results in Figure 3(f) suggest that

the higher support of Stalin among men is primarily derived from individuals born

before 1970. Males and females of the two latest generations are not statistically

different from each other. For instance, among those born before 1934, males

express 1.77 (p , 0.01) points higher support for Stalin than did females, while

for the generation that grew up in an independent Georgia this difference is not

statistically significant. Stronger attachment of elderly men to Stalin can be related

to his military leadership during World War II, which becomes a less relevant

factor among the younger generations.

A. Gugushvili and P. Kabachnik24

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 27: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Predicting attitudes toward Stalin

After presenting across- and within-generational analysis of socialization, social

structural, ideological, nationalist, and gender hypotheses, we can tentatively

speculate upon how attitudes toward Stalin will evolve over time. Georgia’s fourth

president, Georgi Margvelashvili, stated that in the future Stalin will be perceived

as a great politician and military leader (Netgazeti 2013). We are unaware of the

kind of evidence on which his statement was based, but our findings allow us to

speculate about how public attitudes can evolve alongside changes in covariates

previously identified as important factors in explaining the perception of Stalin.

We assert with a high level of certainty that positive attitudes toward Stalin

should gradually decline in Georgia. First and foremost, this trend will be

determined by the reproduction of demographic cohorts. We know that

consecutive generations in twentieth-century Georgia substantially differ in

their childhood experiences, and attitudes toward Stalin adopted during their

formative years persist across the life course of these individuals. Because the

major cleavages in Stalin’s perception come from cohort differences, the

replacement of the older cohorts with the younger ones should lead to a sizable

decline in positive perceptions of Stalin. The results of our analysis suggest that

the most positive attitudes toward Stalin are held by individuals born before the

1950s. Considering the current average life expectancy, 73.3 years, and its slow

rise in Georgia, individuals born in 1920s–1940s (scoring more than eight points

in our cumulative dependent variable) will be replaced by younger cohorts

(scoring seven points in the same index) by the end of the 2020s. At the same time,

the relative share of the cohorts born in independent Georgia will increase, and

these are demographic cohorts that express the lowest sympathies toward Stalin

(about 5.5 in the aggregated Stalin support index) among all other groups with

various characteristics.

One of the most important factors of changing attitudes toward Stalin could be

economic development and further urbanization in Georgia. We have seen that

rural residents express significantly higher sympathies toward Stalin and therefore

the gradual reallocation of population toward urban areas will affect attitudes

toward the Soviet dictator. Although during the period 1985–2010, the level of

urbanization remained virtually unchanged or even declined, the United Nations

predicts urbanization levels in Georgia will rise to 66.4% by 2050, from 52.7% in

2010 (United Nations 2012). More importantly, our results also suggest that

subjective perception of socioeconomic status is negatively associated with

positive attitudes toward Stalin. The strongest determinant of subjective economic

perceptions is objective economic development (Evans and Kelley 2004). In

1995–2010, the Georgian economy grew by 6% on average (World Bank 2012),

and if this trend continues and Georgia becomes more prosperous,8 this may

impact attitudes toward the Soviet era and Stalin. The main channel through which

economic growth is expected to affect perceptions of Stalin is that people will less

likely be nostalgic for the relative prosperity and economic security of the Soviet

Union and its main architect – Josef Stalin. Time-series analysis in 2008–2013

Post-Soviet Affairs 25

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 28: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

suggests that the share of those households who define themselves as being at the

lowest (1–4) rungs on the 10-step ladder decreased from 66% to 47%, while the

share of the highest four rungs increased from 4% to 12% (CRRC 2013).

Some influence on attitudes toward Stalin will probably stem from the

geopolitical situation in Georgia and its foreign relations with Russia. The overall

perception of Stalin can deteriorate if animosity between these countries remains

high. However, this effect may be primarily due to those born before the 1980s

because the perception of Russia does not make much difference for the youngest

generations. Gradual movement toward establishing the rule of law, respect for

private property, and the dominance of private ownership which were observed in

the last decade (Gwartney, Lawson, and Hall 2013) could also undermine the

admiration of the Soviet leader, who was the architect of policies that are

incompatible with a free market economy. Nonetheless, the 2011–2013 Caucasus

Barometer analysis suggests that the share of those who think that private

ownership of business and industry should increase declined from 56% to 42%,

while the share of those who support government ownership increased from 23%

to 30%. We nonetheless expect that the joint effect of cohort replacement,

structural changes (primarily through economic growth and urbanization), and the

prevalence of liberal views on the economy and politics will decrease positive

perceptions of Stalin in Georgian society.

Conclusion

Examining Stalin is not (only) a historical exploration, as Stalin is not (only) a part

of the past, but “is still a figure of the present” (de Waal 2013b, 5). Ironically, after

more than half a century after the British press described Nikita Khrushchev

becoming the new Soviet leader as “Stalin Is Dead, Long Live Stalin” (The New

London Evening Day 1955), this metaphor is still relevant in reference to

contemporary public sentiments about Stalin in post-Soviet space in general and in

Russia and Georgia in particular. Stalin remains a dominant figure, symbol, and

presence in contemporary post-Soviet space. Despite suggestions otherwise

(Lipman 2013), one can still readily encounter reminders, representations, and

images of Stalin in many parts of the former Soviet Union, such as in Moscow or

Tbilisi. Stalin’s persistent popularity and his increasing presence in public space,

which can be encountered in a wide variety of formats – whether on a wine label,

in a biography, on a poster, or in a monument – has been dubbed by Eugenie

Ikhlov as the phenomenon of Stalinshchina, or Stalin nostalgia (Anders 2005).

Furthermore, this issue “goes to the heart of contemporary questions about

politics, the relationship between society and state, democratization, and

education in the former Soviet Union” (de Waal 2013b, 3).

This analysis has served to provide evidence from the survey data that both

validates and challenges typically invoked reasons that seek to explain people’s

admiration of Stalin, as well as introducing novel explanations that have

heretofore not been entertained. Not only do we contribute to the emerging

academic debate on the role of Stalin in contemporary post-Soviet sociopolitical

A. Gugushvili and P. Kabachnik26

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 29: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

life, but we also connect the broader theories on political attitudes with the

idiosyncratic characteristics of Stalin’s legacy in Georgia. Several hypotheses that

had been viewed as conventional wisdom, but previously only had anecdotal

evidence to support them, are now confirmed through sustained statistical

methods. This includes the idea that older generations and those who think that

they are poorer view Stalin more positively, and that pejorative associations

abound in those who consider Russia to be the biggest threat to Georgia and among

those who value private ownership. This study also rejects other well-trodden

assumptions, like education level and religiosity, and shows that these factors have

little to no effect on how people judge Stalin. We also are able to offer up an

intriguing twist in regards to ethnicity, as non-Georgian minorities in Georgia, if

anything, show even higher esteem for Stalin than do Georgians, although more

research needs to be conducted to better validate this finding due to the low

numbers of non-Georgian respondents in our sample.

Our analysis reveals that the continuing persistence of Stalin in Georgia is

mitigated by a number of factors. The eight main variables that influence attitudes

on Stalin are depicted in Figure 2. Elderly, poor men from rural areas have the

most positive associations with Stalin, while young, wealthier women from cities

judge Stalin negatively. Those with an advanced level of Russian language ability,

who perceive Russia as Georgia’s biggest threat, and who are open to privatization

also view Stalin more harshly. Overall, age is the most significant factor as an

explanatory variable for attitudes toward Stalin, with older generations having the

highest levels of support and the youngest age cohort having the lowest. The

analysis of cleavages within generations also reveals differences in covariates of

the perception of Stalin among the youngest and the oldest residents of Georgia. If

gender, the perception of Russia as the biggest threat for Georgia, and attitudes

toward private ownership of business and industry are the most important factors

that affect support for Stalin among the elderly, subjective income rung, rural-

urban inequalities, and ethnicity are the major explanations why some younger

people still maintain a positive image of Stalin. Our findings do indicate a clear

downward trend in the positive perception of Stalin, and as time progresses,

demographic shifts and economic development will eliminate those who were

socialized within Stalin’s personality cult or have nostalgia for Soviet economic

security, thus reducing a large portion of Stalin admirers from Georgian society.

Although we do not find strong links between attitudes toward democracy,

authoritarianism, and the perception of Stalin, thoughts about Ioseb Jugashvili still

can be proxies for the type of regime and political leader Georgian society is willing

to accept. Thus, what people think about the Soviet dictator can have political

consequences today and can affect the future development of the region. We must

remain cautious about the growing gap in pride in Stalin between the elderly and the

young. Various public awareness campaigns are unlikely to have a dramatic effect

on attitudes toward Stalin among those already socialized, but they might generate

the formation of more extreme perceptions of Stalin in the new generations. While

admiration for Stalin may be weaker among the youth, so too is the amount of

knowledge they have about Stalin. This latter point, coupled with higher levels of

Post-Soviet Affairs 27

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 30: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

ambiguity toward Stalin, leaves room for the possibility that the meaning of Stalin

can change once again. If de-Stalinization is not undertaken in Georgia in school

curricula and in the media, then it is entirely possible for Stalin to become a

renewed symbol of national pride, and be (re)invoked as a political tool, which can

shape a new generation and produce young Stalin admirers, just as studies of youth

in Russia have shown (Mendelson and Gerber 2005, 2006). This will be an

important issue to follow, and it remains to be seen how the current “Georgian

Dream” coalition and consecutive, hopefully democratically elected, governments

will deploy the image of Stalin for their own political agenda (Dixon Klump 2013).

Acknowledgements

This work was supported, in part, by a Dean’s Research Scholarship and a PSC-CUNYAward, jointly funded by The Professional Staff Congress and The City University of NewYork. Earlier versions of this article were presented at the “Europe in Georgia”International Conference (Center for Social Sciences, Tbilisi State University, Tbilisi,Georgia) and the Association for the Study of Nationalities World Convention (New York,NY, USA). We would like to thank Jesse Driscoll, who served as discussant on our ASNpanel, and everyone else who offered comments and questions. We would also like tothank George Breslauer and the anonymous reviewers for their helpful comments. Anyerrors or oversights are our own.

Notes

1. For example, see the special issue of The Soviet and Post-Soviet Review (Volume 38,2011) dedicated to the World War II cult in post-Soviet space.

2. We do not mean to imply that such data are not useful or that qualitative data ingeneral are less rigorous than quantitative data. However, qualitative data aretypically employed in this literature in an ad hoc, limited, and/or anecdotal fashion,rather than in a systematic way (with clearly documented methods and rationale forchoosing the survey sample, interview subjects, etc.).

3. For more details on the survey, see http://www.crrccenters.org.4. These items are: (1) color television; (2) digital photo camera; (3) automatic washing

machine; (4) DVD player; (5) refrigerator; (6) air conditioner; (7) car; (8) land linephone; (9) cell phone; (10) activated internet access from your personal cell phone;(11) personal computer, including laptop; and (12) Internet access from homecomputer or laptop.

5. The obvious exception is Model 3, in which respondents’ agreement with the surveyquestion reflects negative attitudes toward Stalin.

6. We do not report interactions of birth year with education, objective socioeconomicstatus, and religiosity because these variables proved to be insignificant in explainingattitudes toward Stalin across and within generations.

7. This is done because the low number of observations for non-Georgian ethnicity doesnot permit making statistically significant tests across consecutive birth cohorts.

8. In fact, 6% annual growth is sufficient to double the nominal GDP every 12 years.

References

Alwin, Duane F., and Jon A. Krosnick. 1991. “Aging, Cohorts, and the Stability ofSociopolitical Orientations over the Life Span.” American Journal of Sociology 97 (1):169–195. doi:10.1086/229744.

A. Gugushvili and P. Kabachnik28

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 31: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Anders, Jaroslaw. 2005. “Dead Souls.” The New Republic Online, February 14. http://www.newrepublic.com/article/dead-souls

Andersen, Robert. 2012. “Support for Democracy in Cross-National Perspective: TheDetrimental Effect of Economic Inequality.” Research in Social Stratification andMobility 30 (4): 389–402. http://dx.doi.org/10.1016/j.rssm.2012.04.002.

Bahry, Donna. 1993. “Society Transformed? Rethinking the Social Roots of Perestroika.”Slavic Review 52 (3): 512–554. doi:10.2307/2499722.

Bakke, Kristin M., John O’Loughlin, and Michael D. Ward. 2009. “Reconciliation inConflict-Affected Societies: Multilevel Modeling of Individual and Contextual Factorsin the North Caucasus of Russia.” Annals of the Association of American Geographers99 (5): 1012–1021. doi:10.1080/00045600903260622.

Bakradze, Lasha. 2013. “Georgia and Stalin: Still Living with the Great Son of the Nation.”In The Stalin Puzzle: Deciphering Post-Soviet Public Opinion, edited by Thomas deWaal, 47–54. Washington, DC: Carnegie Endowment for International Peace.

Bhabha, Homi. 1990. Nation and Narration. New York: Routledge.Bilinsky, Yaroslav. 1999. “Was the Ukrainian Famine of 1932–1933 Genocide?” Journal

of Genocide Research 1 (2): 147–156. doi:10.1080/14623529908413948.Billig, Michael. 1995. Banal Nationalism. London: Sage.Blauvelt, TimothyK. 2009. “Status Shift and EthnicMobilisation in theMarch 1956 Events

in Georgia.” Europe-Asia Studies 61 (4): 651–668. doi:10.1080/09668130902826246.Blauvelt, Timothy K. 2013. “Endurance of the Soviet Imperial Tongue: The Russian

Language in Contemporary Georgia.” Central Asian Survey 32 (2): 189–209. doi:10.1080/02634937.2013.771978.

Boehmke, Frederick J. 2008. “GRINTER: A Stata Utility to Graph Interaction EffectsAfter Regression Models (Version 1.5).” Iowa City: University of Iowa.

Boym, Svetlana. 2001. The Future of Nostalgia. New York: Basic Books.Brambor, Thomas, William Roberts Clark, and Matt Golder. 2006. “Understanding

Interaction Models: Improving Empirical Analyses.” Political Analysis 14 (1): 63–82.doi:10.1093/pan/mpi014.

Breslauer, George W. 1978. “On the Adaptability of Soviet Welfare-State Authoritarian-ism.” In Soviet Society and the Communist Party, edited by Karl W. Ryavec, 3–25.Amherst: University of Massachusetts Press.

Brooks, Clem, and Catherine Bolzendahl. 2004. “The Transformation of US Gender RoleAttitudes: Cohort Replacement, Social-Structural Change, and Ideological Learning.”Social Science Research 33 (1): 106–133. doi:10.1016/S0049-089X(03)00041-3.

Carnaghan, Ellen, and Donna Bahry. 1990. “Political Attitudes and the Gender Gap in theUSSR.” Comparative Politics 22 (4): 379–399. doi:10.2307/421970.

Chelidze, Ana. 2014. “Ethno-Nationalistic and Religious-Nationalistic Components ofIdentity in Post-Soviet Georgia.” Occasional Papers on Religion in Eastern Europe 34(2): 1–20.

Chinn, Jeff, and Robert Kaiser. 1996. Russians as the New Minority: Ethnicity andNationalism in the Soviet Successor States. Boulder: Westview.

Cohen, Stephen F. 1982. Rethinking the Soviet Experience: Politics and History Since1917. New York: Basic Books.

Colton, Timothy J., and Michael Mcfaul. 2002. “Are Russians Undemocratic?” Post-Soviet Affairs 18 (2): 91–121. doi:10.1080/1060586X.2002.10641515.

CRRC (Caucasus Research Resource Centers). 2012. “Caucasus Barometer.” Tbilisi:Caucasus Research Resource Centers. http://www.crrccenters.org/caucasusbarometer/

CRRC (Caucasus Research Resource Centers). 2013. “Caucasus Barometer Time-SeriesDataset Georgia.” Tbilisi: Caucasus Research Resource Centers. http://caucasusbarometer.org/en/cb-ge/codebook/

Post-Soviet Affairs 29

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 32: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Danilova, Maria. 2005. “Dictator Stalin Stirs Nostalgia as Russians Remember War.”Seattle Times, May 9. http://seattletimes.com/html/nationworld/2002268263_stalin09.html

de Waal, Thomas, ed. 2013a. “The Stalin Puzzle: Deciphering Post-Soviet PublicOpinion.” Washington, DC: Carnegie Endowment for International Peace.

de Waal, Thomas. 2013b. “Introduction.” In The Stalin Puzzle: Deciphering Post-SovietPublicOpinion, editedbyThomasdeWaal, 1–6.Washington,DC:CarnegieEndowment.

Dixon Klump, Sarah. 2013. “The Battle over Stalin Statues and the Ascent of the GeorgianDream Coalition.” The CESS Blog: Scholarly Sketches from Members of the CentralEurasian Studies Society, March 14. http://centraleurasia.org/thecessblog/2013/03/14/the-battle-over-stalins-statues-and-the-ascent-of-the-georgian-dream-coalition

Ekman, Joakim, and Jonas Linde. 2005. “Communist Nostalgia and the Consolidation ofDemocracy in Central and Eastern Europe.” Journal of Communist Studies andTransition Politics 21 (3): 354–374. doi:10.1080/13523270500183512.

Etkind, Alexander. 2013. Warped Mourning: Stories of the Undead in the Land of theUnburied. Stanford, CA: Stanford University Press.

Evans, M. D. R., and Jonathan Kelley. 2004. “Subjective Social Location: Data From 21Nations.” International Journal of Public Opinion Research 16 (1): 3–38. doi:10.1093/ijpor/16.1.3.

Finkel, S. E., S. Humphries, and K. -D. Opp. 2001. “Socialist Values and the Developmentof Democratic Support in the Former East Germany.” International Political ScienceReview 22 (4): 339–361. doi:10.1177/0192512101022004004.

Forest, Benjamin, and Juliet Johnson. 2002. “Unraveling the Threads of History: Soviet-EraMonuments and Post-Soviet National Identity in Moscow.” Annals of the Associationof American Geographers 92 (3): 524–547. doi:10.1111/1467-8306.00303.

Fuchs, Roland J., and George J. Demko. 1979. “Geographic Inequality Under Socialism.”Annals of the Association of American Geographers 69 (2): 304–318. doi:10.1111/j.1467-8306.1979.tb01259.x.

George, Julie A. 2014. “Can Hybrid Regimes Foster Constituencies? Ethnic Minorities inGeorgian Elections, 1992–2012.” Electoral Studies, no. March: 1–18, (in press).doi:10.1016/j.electstud.2014.02.007.

Gibson, James L. 1996. “A Mile Wide but an Inch Deep(?): The Structure of DemocraticCommitments in the Former USSR.” American Journal of Political Science 40 (2):396–420. doi:10.2307/2111630.

Gugushvili, Alexi. 2011. “Understanding Poverty in Georgia.” Caucasus AnalyticalDigest 34 (December 21): 15–18.

Gugushvili, Alexi. 2013. “Social Origin, Education and Occupation in Georgia.”Caucasus Social Science Review 1 (1): 1–26.

Gwartney, James, Robert Lawson, and Joshua Hall, eds. 2013. Economic Freedom of theWorld: 2013 Annual Report. Washington, DC: Fraser Institute. http://www.freetheworld.com/datasets_efw.html

Hahn, Jeffrey W. 1991. “Continuity and Change in Russian Political Culture.” BritishJournal of Political Science 21 (4): 393–421. doi:10.1017/S0007123400006232.

Hahn, Jeffrey W., and Igor Logvinenko. 2008. “Generational Differences in RussianAttitudes Towards Democracy and the Economy.” Europe-Asia Studies 60 (8):1345–1369. doi:10.1080/09668130802292168.

Hobsbawm, Eric. 1983. “Mass-Producing Tradition: Europe 1870–1914.” In TheInvention of Traditon, edited by Eric Hobsbawm, and Ternece Ranger, 263–308.Cambridge: Cambridge University Press.

Hochschild, Adam. 1995. The Unquiet Ghost: Russians Remember Stalin. New York:Penguin.

Inglehart, Ronald. 1988. “The Renaissance of Political Culture.” The American PoliticalScience Review 82 (4): 1203–1230. doi:10.2307/1961756.

A. Gugushvili and P. Kabachnik30

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 33: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Jennings, M. Kent, and Gregory B. Markus. 1984. “Partisan Orientations over the LongHaul: Results from the Three-Wave Political Socialization Panel Study.” TheAmerican Political Science Review 78 (4): 1000–1018. doi:10.2307/1955804.

Jones, Stephen. 1992. “Georgia: The Long Battle for Independence.” In Nationalism andthe Breakup of Empire: Russia and Its Periphery, edited by Miron Rezun, 73–96.Westport, CT: Praeger.

Kara-Murza, Vladimir. 2013. “Resurrecting Stalin – Again.” World Affairs, February 6.http://www.worldaffairsjournal.org/blog/vladimir-kara-murza/resurrecting-stalin-—-again

Kaufman, Stuart. 2001. Modern Hatreds: The Symbolic Politics of Ethnic War. Ithaca,NY: Cornell University Press.

Kemp-Welch, Tony. 1996. “Khrushchev’s ‘Secret Speech’ and Polish Politics: The Springof 1956.” Europe-Asia Studies 48 (2): 181–206. doi:10.1080/09668139608412343.

Khinkulova, Kateryna. 2012. “Hello, Lenin? Nostalgia on Post-Soviet Television inRussia and Ukraine.” Journal of European Television History and Culture 1 (2):94–104.

Kitaevich, Evgenia Jane. 2014. “History That Splinters: Education Reforms and MemoryPolitics in the Republic of Georgia.” Southeast European and Black Sea Studies 14 (2):319–338. doi:10.1080/14683857.2014.906089.

Kolossov, Vladimir, and Gerard Toal. 2007. “An Empire’s Fraying Edge? The NorthCaucasus Instability in Contemporary Russian Geopolitical Culture.” EurasianGeography and Economics 48 (2): 202–225. doi:10.2747/1538-7216.48.2.202.

Kozlov, Vladimir A. 1999.Massovyye besporyadki v SSSR pri Khrushcheve i Brezhneve –1953–1980 gody [Mass Disorder in the USSR Under Khrushchev and Brezhnev –1953–1980] Novosibirsk: Khronograf.

Krosnick, Jon A., and Duane F. Alwin. 1989. “Aging and Susceptibility to AttitudeChange.” Journal of Personality and Social Psychology 57 (3): 416–425. doi:10.1037/0022-3514.57.3.416.

Lee, Moonyoung. 2011. “Nostalgia as a Feature of ‘Glocalization’: Use of the Past in Post-Soviet Russia.” Post-Soviet Affairs 27 (2): 158–177. doi:10.2747/1060-586X.27.2.158.

Lipman, Maria. 2013. “Stalin Is Not Dead: A Legacy That Holds Back Russia.” In TheStalin Puzzle: Deciphering Post-Soviet Public Opinion, edited by Thomas de Waal,15–26. Washington, DC: Carnegie Endowment.

Marsh, Rosalind. 2007. Literature, History and Identity in Post-Soviet Russia, 1991–2006. Oxford: Peter Lang.

Medvedev, Roy. 2005. “Stalin Lives.” Project Syndicate, March 29. http://www.project-syndicate.org/commentary/stalin-lives

Mendelson, Sarah E., and Theodore P. Gerber. 2005. “Soviet Nostalgia: An Impediment toRussian Democratization.” The Washington Quarterly 29 (1): 83–96. doi:10.1162/016366005774859661.

Mendelson, Sarah E., and Theodore P. Gerber. 2006. “Failing the Stalin Test: Russiansand Their Dictator.” Foreign Affairs 85 (1): 2–8. doi:10.2307/20031837.

Mendelson, Sarah E., and Theodore P. Gerber. 2008. “Us and Them: Anti-AmericanViews of the Putin Generation.” The Washington Quarterly 31 (2): 131–150. doi:10.1162/wash.2008.31.2.131.

Metskhvarishvili, Maya. 2013. “Ilia II: I Love Russia Very Much, Stalin Was Religious.”Netgazeti, July 31. http://netgazeti.ge/GE/105/News/22214/

Munro, Neil. 2006. “Russia’s Persistent Communist Legacy: Nostalgia, Reaction, andReactionary Expectations.” Post-Soviet Affairs 22 (4): 289–313. doi:10.2747/1060-586X.22.4.289.

Post-Soviet Affairs 31

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 34: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Netgazeti. 2013. “Margvelashvili: Stalin Will Be Perceived in the Future as a GreatPolitician and Military Leader.” Netgazeti, September 18. http://www.netgazeti.ge/GE/105/opinion/23607/

Neundorf, Anja. 2010. “Democracy in Transition: A Micro Perspective on System Changein Post-Socialist Societies.” The Journal of Politics 72 (4): 1096–1108. doi:10.1017/S0022381610000551.

Nodia, Ghia. 1995. “Georgia’s Identity Crisis.” Journal of Democracy 6 (1): 104–116.doi:10.1353/jod.1995.0014. http://muse.jhu.edu/content/crossref/journals/journal_of_democracy/v006/6.1nodia.html

O’Loughlin, John, Vladimir Kolossov, and Gerard Toal. 2011. “Inside Abkhazia: Surveyof Attitudes in a De Facto State.” Post-Soviet Affairs 27 (1): 1–36. doi:10.2747/1060-586X.27.1.1.

O’Loughlin, John, Gerard Toal, and Vladimir Kolossov. 2008. “The LocalizedGeopolitics of Displacement and Return in Eastern Prigorodnyy Rayon, NorthOssetia.” Eurasian Geography and Economics 49 (6): 635–669. doi:10.2747/1539-7216.49.6.635.

Pop-Eleches, Grigore, and Joshua A. Tucker. 2014. “Communist Socialization and Post-Communist Economic and Political Attitudes.” Electoral Studies 33: 77–89. http://dx.doi.org/10.1016/j.electstud.2013.06.008.

Popov, N. P. 2008. “Nostalgia for Greatness – Russia in the Post-Soviet Space.”Sociological Research 47 (5): 36–51. doi:10.2753/SOR1061-0154470502.

Pratto, Felicia, Lisa M. Stallworth, and Jim Sidanius. 1997. “The Gender Gap: Differencesin Political Attitudes and Social Dominance Orientation.” British Journal of SocialPsychology 36 (1): 49–68. doi:10.1111/j.2044-8309.1997.tb01118.x.

Reif, K., and G. Cunningham. 1993. “Dataset: Central and Eastern Eurobarometer 3(Political Desintegration).” GESIS–Leibniz Institute for the Social Sciences, Cologne,Germany. http://zacat.gesis.org

Reisinger, William M., Arthur H. Miller, and Vicki L. Hesli. 1995. “Political Norms inRural Russia: Evidence from Public Attitudes.” Europe-Asia Studies 47 (6):1025–1042. doi:10.1080/09668139508412303.

Reisinger, William M., Arthur H. Miller, Vicki L. Hesli, and Kristen Hill Maher. 1994.“Political Values in Russia, Ukraine and Lithuania: Sources and Implications forDemocracy.” British Journal of Political Science 24 (2): 183–223. doi:10.1017/S0007123400009789.

Roller, Edeltraud. 1994. “Ideological Basis of the Market Economy: Attitudes TowardDistribution Principles and the Role of Government in Western and EasternGermany.” European Sociological Review 10 (2): 105–117.

Rose, Richard, and Ellen Carnaghan. 1995. “Generational Effects on Attitudes toCommunist Regimes: A Comparative Analysis.” Post-Soviet Affairs 11 (1): 28–56.doi:10.1080/1060586X.1995.10641393.

Rosenberg, Tina. 1996. The Haunted Land: Facing Europe’s Ghosts After Communism.New York: Vintage.

Ruble, Blair A. 1987. “The Social Dimensions of Perestroyka.” Soviet Economy 3 (2):171–183. doi:10.1080/08826994.1987.10641276.

Ryder, Norman B. 1965. “The Cohort as a Concept in the Study of Social Change.”American Sociological Review 30 (6): 843–861. doi:10.2307/2090964.

Satter, David. 2012. It Was a Long Time Ago and It Never Happened Anyway: Russia andthe Communist Past. New Haven, CT: Yale University Press.

Shapiro, Robert Y., and Harpreet Mahajan. 1986. “Gender Differences in PolicyPreferences: A Summary of Trends from the 1960s to the 1980s.” Public OpinionQuarterly 50 (1): 42–61. doi:10.1086/268958.

A. Gugushvili and P. Kabachnik32

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 35: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Slider, Darrell. 1997. “Democratization in Georgia.” In Conflict, Cleavage, and Change inCentral Asia and the Caucasus, edited by Karen Dawisha and Bruce Parrott, 156–198.Cambridge: Cambridge University Press.

Slysz, Gregory. 2010. “Crafting a Suitable Past.” Transitions Online, November 29. http://www.tol.org/client/article/21989-crafting-a-suitable-past.html

Smith, Tom W. 1984. “The Polls: Gender and Attitudes Toward Violence.” PublicOpinion Quarterly 48 (1): 384–396. doi:10.1086/268834.

Smith, Kathleen E. 1996. Remembering Stalin’s Victims: Popular Memory and the End ofthe USSR. Ithaca, NY: Cornell University Press.

Sullivan, Charles J. 2013. “Missing the Soviet Motherland: Nostalgia for the USSR inRussia Today.” March 20. http://www.globalinterests.org/wp-content/uploads/2013/03/Missing-the-Soviet-Motherland-Nostalgia-for-the-USSR-in-Russia-Today.pdf

Suny, Ronald Grigor. 1994. The Making of the Georgian Nation. Bloomington: IndianaUniversity Press.

Suny, Ronald Grigor. 2006. “Nationalism, Nation Making, and the Postcolonial State ofAsia, Africa, and Eurasia.” In After Independence, edited by Lowell W. Barrington,279–295. Ann Arbor: University of Michigan Press.

Taylor, Adam. 2014. “Calls for a Return to ‘Stalingrad’ Name Test the Limits of Putin’sSoviet Nostalgia.” The Washington Post, June 9. http://www.washingtonpost.com/blogs/worldviews/wp/2014/06/09/calls-for-a-return-to-stalingrad-name-test-the-limits-of-putins-soviet-nostalgia/

The New London Evening Day. 1955. “Long Live Stalin: Malenkov’s Fall Is Receivedwith ‘Forboding’ in Europe.” The New London Evening Day, February 8. http://news.google.com/newspapers?nid¼1915&dat¼19550208&id¼zREiAAAAIBAJ&sjid¼EXQFAAAAIBAJ&pg¼1772,1220962

Toal, Gerard, and Magdalena Frichova Grono. 2011. “After Ethnic Violence in theCaucasus: Attitudes of Local Abkhazians and Displaced Georgians in 2010.” EurasianGeography and Economics 52 (5): 655–678. doi:10.2747/1539-7216.52.5.655.

Toal, Gerard, and John O’Loughlin. 2013. “Inside South Ossetia: A Survey of Attitudes ina De Facto State.” Post-Soviet Affairs 29 (2): 136–172. doi:10.1080/1060586X.2013.780417.

Togeby, Lise. 1994. “The Gender Gap in Foreign Policy Attitudes.” Journal of PeaceResearch 31 (4): 375–392. doi:10.1177/0022343394031004002.

United Nations. 2012. World Population Prospects: The 2011 Revision. New York:United Nations.

Vachridze, Zaza. 2012. “Two Faces of Nationalism and Efforts to Establish GeorgianIdentity.” Identity Studies 4: 1–7, http://ojs.iliauni.edu.ge/index.php/identitystudies/article/view/47/35

Vashakidze, Nika. 2013. “The Role of 9 March 1956 in the Formation of GeorgianEthnopsychology.” Iveria, April 30. http://iveria.biz/18-1956-wlis-9-martis-roli-qartuli-etnofsiqikis-formirebasi.html (in Georgian).

Vegso, Roland. 2013. “Stalin’s Boots and the March of History (Post-CommunistMemories).” Cultural Critique 83 (Winter): 31–62, http://muse.jhu.edu/journals/cultural_critique/v083/83.vegs.html

Visser, Penny S., and Jon A. Krosnick. 1998. “Development of Attitude Strength over theLife Cycle: Surge and Decline.” Journal of Personality and Social Psychology 75 (6):1389–1410. doi:10.1037/0022-3514.75.6.1389.

Wegren, Stephen K. 1994. “Rural Reform and Political Culture in Russia.” Europe-AsiaStudies 46 (2): 215–241. doi:10.1080/09668139408412159.

Wildavsky, Aaron. 1987. “Choosing Preferences by Constructing Institutions: A CulturalTheory of Preference Formation.” American Political Science Review 81 (1): 3–22.doi:10.2307/1960776.

Post-Soviet Affairs 33

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 36: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Williams, Richard. 2006. “Generalized Ordered Logit/Partial Proportional Odds Modelsfor Ordinal Dependent Variables.” The Stata Journal 6 (1): 58–82, http://www.stata-journal.com/article.html?article¼st0097

World Bank. 2009. Georgia: Poverty Assessment. Washington, DC: World Bank, HumanDevelopment Sector Unit, South Caucasus Country Unit, Europe and Central AsiaRegion.

World Bank. 2012. World Development Indicators 2012. Washington, DC: World Bank.Yasmann, Victor. 2006. “Russia: Nostalgia for USSR Increases.” Radio Free Europe/

Radio Liberty, December 23. http://www.rferl.org/content/article/1073655.htmlZedania, Giga. 2011. “Societal Values in Georgia: Twenty Years Later.” In 20 Years after

the Collapse of Communism: Expectations, Achievements and Disillusions of 1989,edited by Nicolas Hayoz, Leszek Jesien, and Daniela Koleva, 253–272. Bern: PeterLang.

Zedania, Giga. 2012. “The Rise of Religious Nationalism in Georgia.” Identity Studies 3:120–128, http://www.identitystudies.ac.ge/index.php/IStudies/article/view/38

Appendix A. Overview of employed statistical methods

We employ two regression specifications in the multivariate analysis. In the ordinalform, the dependent variables permit the estimation of simple ordered logisticregressions, but the parallel regression assumption in these models is often violated. Inthis case, the coefficients that describe the relationship between, for instance,“completely disagree vs. mostly disagree, mostly agree, completely agree”’ should beidentical to the coefficients that describe the relationship between “completely disagree,mostly disagree vs. mostly agree, completely agree.” Indeed, the conducted Brant Test ofthe parallel regression assumption indicates that in each fitted model, the parallel-linesmodel assumption was violated for several independent variables such as gender, type ofsettlement, education, and ethnicity. Among several alternatives to the ordered logitmodel such as multinomial logit and linear probability models (for the created binaryanswer option), we employ the maximum-likelihood generalized ordered logit model,which relaxes the proportional odds assumption and allows the effects of the independentvariables to vary across ordered categories of the dependent variable (Williams 2006).Moreover, the output of generalized ordered logit regressions is relatively straightfor-ward to interpret, which is explained in detail in the Results section. We also calculaterobust standard errors.

Ordered variables on attitudes toward Stalini ¼ b0 i þ gGenerationi þ x Social

structural variablesi þ w3 Ideologyi þ uNationalismi þ lGenderi þ 1i:ð1Þ

In Model 1, b0i and 1i represent the intercept and error terms, respectively. Thedescribed ordered dependent variables and generalized ordered logit models are lesshelpful for identifying the scale of impact of independent variables on attitudes towardStalin. To address this problem, we derive the cumulative index of perception of the Sovietdictator by the survey respondents. This is done by combining answers on the dependentvariables from completely disagree ¼ 1 to completely agree ¼ 4 into a single index ofattitudes toward Stalin. We reversed answer options for question 3 because higher scoresindicated negative attitudes toward Stalin in contrast to the other dependent variables.Both the factor loading of these variables, all above 0.40 with an eigenvalue of Factor1 ¼ 1.58, and their scale reliability coefficient, 0.72, suggest that the derived index is arelevant aggregate measure of attitudes toward Stalin. In order to make the results easier tointerpret, we rescaled answer options to completely disagree ¼ 0 to completely agree ¼ 3.

A. Gugushvili and P. Kabachnik34

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 37: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Therefore, 0 points (4 £ 0) in the cumulative index means that for all four questions therespondent holds extremely negative ideas about Stalin (2.1% of the sample), while thosewho score 12 points (4 £ 3) have extremely positive attitudes toward the Soviet dictator(4.7% of the sample). Because the derived dependent variable varies from 0 to 12,calculating linear predictions from ordinary least squares (OLS) models presents itself asthe most convenient option.

Attitudes towards Stalini ¼ þbn Birth yeari þ bn Birth year £ Social structural

variablesþ bn Birth year £ Ideologyi þ bn Birth year £ Nationalismi

þ bn Birth year* Genderi:

ð2Þ

Finally, to understand how social structural, ideological, nationalism, and gendervariables differ in their effect on attitudes toward Stalin within consecutive generations,we investigate the effect of those factors across respondents’ birth cohort. We use the samespecifications as in Model 1, with the index of attitudes toward Stalin as the dependentvariable, but this time with the addition of interaction terms of generations and otherhypothesized independent variables that proved to be statistically significant in the earlierbivariate and multivariate analyses. Furthermore, we substitute the dummy variables forthe seven generations with a more precise year-of-birth variable that allows us to linearlyfollow change in attitudes along with individual years, conditioned by other independentvariables. We include in Model 2 all constitutive terms of interaction effects, whichincreases the size of standard errors and makes it less likely that the coefficient on theinteraction term will be significant (Brambor, Clark, and Golder 2006). Because theinterpretation of interaction terms in regression models is often misleading, we employ asimple figure that succinctly illustrates the marginal effect of hypothesized variables andthe corresponding standard errors across a substantively meaningful range of themodifying variable, respondents’ birth year, ranging from the 1920s to the 1990s. We usethe command GRINTER to visualize interaction effects and corresponding standard errors(Boehmke 2008).

Post-Soviet Affairs 35

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014

Page 38: Stalin is dead, long live Stalin? Testing socialization, structural, ideological, nationalist, and gender hypotheses

Table

B1.

Univariate

analysisofvariance

(ANOVA)testsforattitudes

aboutStalin.

Dependentvariables

Dependentvariable

1Dependentvariable

2Dependentvariable

3Dependentvariable

4

“Stalinwas

awise

leader

whobrought

power

andprosperity

totheSoviet

Union”

“Themostim

portant

thingisthat

under

Stalin’s

leadership

the

Sovietpeople

won

WWII”

“Stalinwas

acruel,

inhuman

tyrant,

responsible

forthe

deathsofmillionsof

innocentpeople”

“Ourpeople

will

alwayshaveneedofa

leader

likeStalin,who

willcomeandrestore

order”

Independentvariables

F-ratio

(h2)

F-ratio

(h2)

F-ratio

(h2)

F-ratio

(h2)

Birth

cohorts

25.4

(0.07)***

16.8

(0.05)***

29.0

(0.09)***

38.6

(0.11)***

Years

ofeducation

3.62(0.03)***

2.22(0.01)***

3.41(0.02)***

4.75(0.04)***

Proficiency

inRussian

9.96(0.01)***

6.93(0.01)***

10.9

(0.02)***

18.5

(0.03)***

Proficiency

inEnglish

35.1

(0.05)***

8.62(0.02)***

20.9

(0.03)***

42.1

(0.07)***

Settlem

ent

41.5

(0.03)***

21.5

(0.02)***

35.8

(0.04)***

53.5

(0.06)***

Unem

ployment

0.04(2

0.00)

0.88(2

0.00)

6.37(0.00)**

5.78(0.00)**

Number

ofhouseholdsitem

s16.1

(0.05)***

10.4

(0.03)***

11.4

(0.04)***

17.0

(0.06)***

Subjectiveincomerung

8.73(0.03)***

14.4

(0.05)***

5.62(0.02)***

6.95(0.03)***

Attitudes

towarddem

ocracy

0.84(2

0.00)

2.54(0.00)*

8.20(0.01)***

5.92(0.01)***

Attitudes

towardprivatization

6.77(0.01)***

7.57(0.00)***

5.32(0.01)***

9.21(0.02)***

Ethnicity

0.61(2

0.00)

13.1

(0.01)***

1.24(2

0.00)

10.6

(0.01)***

Religiosity

1.42(0.00)

6.59(0.01)***

2.88(0.01)***

3.11(0.01)***

Russia

isthebiggestenem

y21.9

(0.01)***

11.2

(0.01)***

19.0

(0.01)***

10.9

(0.01)***

Gender

7.04(0.00)***

1.23(0.00)

2.21(0.00)

3.34(0.00)*

Notes:***,**,and*denote

statisticalsignificance

atthe0.01,0.05,and0.10levels,respectively.

Source:

Authors’calculationsbased

ondatafrom

CaucasusBarometer

(CRRC2012).

Appendix

B

A. Gugushvili and P. Kabachnik36

Dow

nloa

ded

by [

Tha

mm

asat

Uni

vers

ity L

ibra

ries

] at

03:

32 0

9 O

ctob

er 2

014