Introduction
The concept of age has been prominent in social science research (Glenn, Reference Glenn1974; O'Rand and Krecker, Reference O'Rand and Krecker1990; Riley, Reference Riley1987). A large body of literature has investigated how age affects public opinion, including party identification (Abramson, Reference Abramson1979; Converse, Reference Converse1976), voting (Blais et al., Reference Blais, Gidengil and Nevitte2004; Goerres, Reference Goerres2007; Waas, Reference Waas2007), and democratic support (Finifter and Mickiewicz, Reference Finifter and Mickiewicz1992; McFaul, Reference McFaul2003; Rose, Mishler and Haerpfer, Reference Rose, Mishler and Haerpfer1998). There is, however, little consensus about the mechanisms through which age shapes political dispositions. Within this body of literature, the distinction between life cycle, generational and period effects has drawn considerable attention. One strand of research emphasizes attitudinal and behavioural adjustments associated with the passage through various life stages (Braungart and Braungart, Reference Braungart and Braungart1974; Jarvikoski, Reference Jarvikoski1993; Watts, Reference Watts1999). Another line of inquiry posits that generational replacement accounts for value change (Inglehart, Reference Inglehart1977; Jennings, Reference Jennings1987; Tessler et al., Reference Tessler, Konold and Reif2004). Additionally, others argue that dramatic external shocks, including wars, epidemics and revolutions, affect individuals irrespective of age or generational membership (Beck and Jennings, Reference Beck and Jennings1979). Though empirical research on this subject spans several decades, the debate still continues. A recent wave of regime change in Central and Eastern Europe affords a fresh opportunity for re-evaluating the importance of age against the backdrop of dramatic social transformations.
This paper represents an attempt to assess the impact of life cycle, generational and period effects on protest potential in post-communist Russia by analyzing data from three waves of the World Values Survey. Over the course of Boris Yeltsin's ten-year presidency, the survey was administered three times—in 1990, 1995, and 1999—offering a snapshot of public opinion at various stages of the transition period. The substantive focus on protest activity is appropriate not only because this form of political participation is emblematic of a remarkable departure from the habitual patterns of mass mobilization under communism, but also because direct political action constitutes an important part of democratic politics.
There are at least three reasons why the concept of age merits further systematic attention in the post-communist world. First, attending most previous research in advanced industrial democracies has been the stability of the political regime and the relative mildness of economic shocks. In the absence of macro-level traumatic events, it is reasonable to expect that generational and period effects will be hard to find in mature democracies. Post-communist citizens, in contrast, have lived through immense political, economic and social transitions that might have exacerbated age-related cleavages in society. A cross-time analysis of public opinion data from post-communist states will broaden our knowledge of life cycle and generational effects in times of social turmoil.
Another reason for directing more academic attention toward the age-graded study of public opinion in Eastern Europe lies in connecting two bodies of literature that emphasize either short-term or long-term effects on political attitudes. Over the past decade, scores of post-communist studies sought to explain patterns of political behaviour by concentrating on such short-term factors as subjective well-being, evaluations of incumbent performance or issue opinions (for example, Barnes and Simon, Reference Barnes and Simon1998; Colton, Reference Colton2000; Kluegel and Mason, Reference Kluegel and Mason1999). Though age is routinely included in statistical models on political participation, the tripartite distinction between life cycle, generational, and period effects is rarely made. This paper argues that the integration of political socialization and developmental psychology arguments in empirical studies of protest potential will aid our understanding of political attitudes in the region.
Finally, a close inspection of generational patterns has far-reaching implications for regime change in countries that fall somewhere between democracy and dictatorship. Sovietologists have long searched for a sign of intergenerational rifts within the ranks of the Communist Party as a precursor of possible policy shifts (Bahry, Reference Bahry and Millar1987; Beissinger, Reference Beissinger1986; Lodge, Reference Lodge1968). Likewise, early studies of the post-Soviet citizenry sought to identify a clear break with communism (for example, Gibson et al., Reference Gibson, Duch and Tedin1992; Hahn, Reference Hahn1991). By gauging the magnitude of generational differences, analysts may gain valuable insights into the dispersion of democratic ideas across population groups.
The remainder of the paper proceeds in the following manner. The first section presents hypotheses derived from major theoretical approaches to the study of age. Next, the paper outlines methodology employed in this study and discusses the findings. The paper concludes by drawing out implications of age-related differences for political stability in Russia.
Theories of Life Cycle, Generational, and Period Effects
Life Cycle Effects
The underlying assumption of life cycle theories is that each life stage is associated with a distinct set of biological needs, cognitive functions, and psychosocial characteristics (O'Rand and Krecker, Reference O'Rand and Krecker1990). The development of cognitive skills during adolescence equips individuals with the ability to grasp abstract ideas and exercise critical thinking skills (Adelson and O'Neil, Reference Adelson and O'Neil1966). Concomitantly, quest for identity influences adolescents' social behaviour (Erickson, Reference Erickson1959). By the middle age, individuals tend to sharpen their cognitive skills and assume a variety of social roles linked to marriage, parenthood, and career advancement (Willis, Reference Willis1999). Aging gradually leads to a decline in physical strength. A common false belief is that senior citizens also experience a marked loss of cognitive skills. Empirical research, however, confirms that intellectual skills derived from accumulated experience increase with age (Fischer et al., Reference Fischer, Yan, Stewart, Valsiner and Connolly2003).
The salience of age-specific developmental characteristics motivated social scientists to put forward life cycle interpretations of political behaviour. Evidence abounds that youth, with its propensity to rebel against traditional norms and challenge parental authority, has been a major carrier of revolutionary ideas (Braungart, Reference Braungart, Dragastin and Elder1975). For instance, Watts's (Reference Watts1999) analysis of the link between protest and age in Germany finds an inverted U-relationship: there is a steady rise in protest potential until the mid-20s and a monotonic decline afterwards. Participation in mainstream politics, on the contrary, is skewed to the middle aged (Leighley, Reference Leighley1995). National election results from 22 European countries, for example, indicate that young people are less likely to participate than older voters (Fieldhouse, Tranmer, and Russell, Reference Fieldhouse, Tranmer and Russell2007).
At the same time, recent demographic changes in advanced industrial democracies suggest that senior citizens may remain engaged in politics upon retirement. Campbell, for example, shows how Social Security and Medicare programs in the United States have created a constituency of citizens willing to defend their welfare privileges through political action (Reference Campbell2003). As citizens in advanced industrial democracies live longer and enjoy a wider range of opportunities for self-realization, the literature has responded to the social change by revising conventional expectations about age-specific behaviour. One of the arguments in this line of research is that individual's position in the labour market structure, rather than biological age, is a strong indicator of a life cycle phase (Kohli, Reference Kohli, Riley, Kahn and Foner1994).
Still, biological age is a defining feature of a life phase in contemporary Russia. First, the demographic situation in the former Soviet republic stands in stark contrast with population trends in North America and Western Europe. While advanced industrial democracies have experienced an increase in longevity, the average life expectancy in Russia dropped from 69 years in 1990 to 65 years in 2000, with women living longer than men.Footnote 1 By 1995, for example, male life expectancy has sunk to 57 (Specter, Reference Specter1995). In comparison, male life expectancy in Canada has increased from 60 in 1931 to 75 in 1991 (Denton and Spencer, Reference Denton and Spencer2000). Second, dismal economic performance under Yeltsin's presidency has narrowed down employment opportunities of senior citizens. Compared to their Western counterparts, Russian pensioners possess scant resources for challenging age-specific norms and improving their quality of life. Ample empirical evidence documents abysmal living standards of Russia's senior citizens who struggled to survive on meagre pensions in the 1990s.Footnote 2
Electoral studies provide insights into the relationship between age and political behaviour in post-communist societies. Research shows that young people are less likely to vote than older age groups (Colton, Reference Colton2000). According to White, Rose, and McAllister, for example, nonvoters in Russia's 1993 parliamentary election were “twice as likely to be under 30 as over 55” (Reference White, Rose and McAllister1997: 121). Meanwhile, it is widely held that young people are more prone to protest because they have less access to positions of power and fewer commitments associated with adult roles (Jarvikoski, Reference Jarvikoski1993: 82; Marsh, Reference Marsh1974: 124). Consistent with this view, this study hypothesizes that younger respondents are likely to exhibit a higher level of protest potential than older age groups.
Generational Effects
A key argument of the generational theory is that pre-adult socialization exerts enduring effects on political attitudes (for example, Reshon, Reference Reshon1977; Sears and Valentino, Reference Sears and Valentino1997). Proponents of this approach postulate that adolescence is a pivotal period for the development of political thinking (Jennings and Niemi, Reference Jennings and Niemi1981; Torney-Purta et al., Reference Torney-Purta, Lehmann, Oswald and Schultz2001). Influenced by Mannheim's seminal work, this strand of research defines each political generation in terms of its “common location in the social and historical process” during the formative years (Reference Mannheim and Kleckemeti1952: 291). Yet the manifestation of generational differences in the political sphere is quite rare due to the irregular occurrence of tumultuous events (Tessler et al., Reference Tessler, Konold and Reif2004). Opinion polls from post-communist Russia support the claim that dramatic social change exacerbates cross-generational differences. According to the 1998 poll, for example, 62 per cent of Russia's respondents aged 50 or older stated that social reforms had sharpened generational confrontations in the country.Footnote 3
Since Soviet times, generational turnover has been a vexing issue in studies of Russian society. The ascendancy of a new political elite in the post-Stalin period has spawned an academic debate over the unity of the Communist party and the stability of the political system (Beissinger, Reference Beissinger1986; Lodge, Reference Lodge1968). Likewise, the introduction of glasnost, a new information policy, and perestroika, economic restructuring, has revived hopes about the emergence of the young generation capable of dismantling the communist regime (Finifter and Mickiewicz, Reference Finifter and Mickiewicz1992; Gibson, Duch, and Tedin, Reference Gibson, Duch and Tedin1992).
The fall of the Berlin Wall and the dissolution of the Soviet Union opened up fresh opportunities for examining the extent of generational cleavages in the region. Utilizing 1994 survey data from 11 post-communist states, Rose and Carnaghan find significant generational effects on the level of approval of the communist regime (Reference Rose and Carnaghan1995). Similarly, Mishler and Rose demonstrate that Russia's older generations report higher levels of support for the communist regime (Reference Mishler and Rose2006).
Consistent with earlier research, the study distinguishes political generations based upon four periods in Soviet history: Stalin, Khrushchev, Brezhnev, and Gorbachev.Footnote 4 As the name of each historic period suggests, the political leadership of the Communist party had a strong bearing on life experiences of ordinary citizens in the Soviet Union.Footnote 5 The rampant use of political violence in the form of massive purges and forced collectivization is likely to have made members of the Stalin generation, born 1907–1935, quite wary of expressing dissent in the public arena. Nikita Khrushchev's subsequent denunciation of the cult of personality coupled with a rehabilitation of political prisoners and a “thaw” in cultural sphere leads us to believe that citizens born 1936–1946 are prone to protest more than the preceding generation.
Meanwhile, political dispositions of the Khrushchev generation are likely to bear close resemblance to those of the successive generation, born 1947–1965. Notwithstanding divergent leadership styles, Khrushchev and Brezhnev were committed to salvaging the communist system through the build-up of its economic might and the expansion of welfare services. For example, thousands of Soviet households have benefited from Khrushchev's housing policy that increased the availability of separate apartments as a substitute to communal, dormitory-like residences (kommunalki). By the same token, Brezhnev's “stability of cadres” policy fostered widespread expectations about lifelong employment. This streak of relative economic prosperity and unprecedented welfare security has nurtured complacency and political passivity.Footnote 6 Hence, the study hypothesizes that members of Khrushchev and Brezhnev generations will be more reluctant to engage in protest activity than members of the last Soviet generation.
This inquiry begins with assumption that the Gorbachev generation, born 1966–1973, will exhibit the highest level of protest potential. Individuals socialized during this period have witnessed a wave of political liberalization that has accelerated the collapse of communism and the dissolution of the Soviet Union. The new information policy spawned not only media criticism of inept factory managers, but also a close scrutiny of the moribund political system. By the same token, an outbreak of nationalist mobilization erupted in the regions, delivering another blow to the legitimacy of the communist state. It is reasonable to assume that the last cohort of the Soviet citizenry will be most prone to political action against the authorities.
Period Effects
Like generational theories, this theoretical approach is preoccupied with the role of historic events in influencing political attitudes. Yet, rather than assuming the importance of formative years, the period effects argument posits that macro-level traumatic events affect each individual regardless of one's membership in a political generation or life cycle phase. The emphasis on period effects is evident in American studies of mass mobilization during the Vietnam War. Beck and Jennings (Reference Beck and Jennings1979), for example, find a liberal participation bias both among young respondents and their parents in the late 1960s and early 1970s.
In Central and Eastern Europe, the late 1980s and early 1990s represent a period of extraordinary political, economic, and social transitions (White, Reference White1991). A wave of ethnic mobilization pushing for the collapse of communism brought thousands of people into the street in the late 1980s (Beissinger, Reference Beissinger2002; Gorenburg, Reference Gorenburg2003). The disintegration of the Soviet Union, however, failed to create an institutional context conducive to political and civic engagement of ordinary citizens in Russia. Instead, spurred by the increasing freedom from compulsory political participation and expanding opportunities for the pursuit of private solutions to individual problems, “disengagement became a respectable position” in the post-communist state (Roberts and Jung, Reference Roberts and Jung1995: 165).
According to numerous accounts, citizens identify the right to decide for oneself whether or not to engage in politics as a major personal gain of living in the post-Soviet period studies (Alexander, Reference Alexander2000: 150–57; White et al., Reference White, Rose and McAllister1997: 132–33). The Communist system coerced individuals into compulsory and unanimous voting. According to official statistics, voter turnout reached 99 per cent in the former Soviet Union. The so-called “compliant activism” also involved membership in the Communist party (Letki, Reference Letki2004). Since the collapse of communism, the intensity of political engagement has subsided. Levels of voter turnout have significantly dropped in East-Central Europe (Kostadinova, Reference Kostadinova2003). Moreover, citizens are now reluctant to join political parties. In their study of Russian electoral behaviour during the early 1990s, White, Rose and McAllister find “the absence of commitment to a party by almost three-quarters of the electorate” (Reference White, Rose and McAllister1997: 137). In addition, post-communist citizens refrain from joining voluntary organizations (Howard, Reference Howard, Badescu and Uslaner2003).
In line with the disengagement argument, the present study expects to find lower levels of protest in Russia in the second half of the 1990s, compared to the travails of 1990. Throughout the 1990s, President Yeltsin governed the country. From the historian's perspective, Yeltsin's tenure may be treated as a single period. Yet, from the political scientist's standpoint, it is imperative to distinguish between the first years of transition from communism marked by political and economic liberalization and the last years of Yeltsin's presidency sprinkled with non-democratic measures and opaque business ventures. The consolidation of political and economic power coupled with public acquiescence to the status quo is likely to have deflated protest potential in Russia.
Data and Methods
Case Selection
This study treats Russia as a “critical case” that holds “a strategic importance in relation to the general problem” (Flyvbjerg, Reference Flyvbjerg, Seale, Gobo, Gubrium and Silverman2004: 425). The case of Russia can clearly confirm or falsify propositions about generational cleavages in the post-communist region. Of all the former Soviet republics, Russia endured the lengthiest period under communism. The Baltic states, for example, were coerced into joining the Soviet Union during the Second World War and, thus, had one generation less living under the Soviet system. Another reason why other former Soviet republics are excluded from this analysis has to do with cross-national variations in political openness. For instance, Belarus under President Lukashenka developed into a more repressive political regime than Yeltsin's Russia.
Data Source
The present study uses data from the World Values Survey administered in Russia three times over the past decade.Footnote 7 The first wave of survey research was carried out in October–November 1990, less than a year before the August 1991 coup and Gorbachev's removal from power. The second wave of data collection occurred in November 1995–January 1996, almost half a year prior to Yeltsin's re-election for a second term in office. Finally, the third wave of survey research was administered in April–June 1999, during the last months of Yeltsin's presidency. A total of 6,501 respondents aged 18–90 participated in the survey.
Dependent Variable
Protest Potential. The survey gauges protest potential by asking respondents to indicate the likelihood of their participation in five types of protest activity: (1) signing a petition, (2) joining a peaceful rally, (3) joining a boycott, (4) joining a strike and (5) occupying an administrative building. Each indicator is measured on a three-point scale with the following response categories: never, might do, and have done. Using these five indicators, the additive index of protest potential is computed.Footnote 8 The value of Cronbach's alpha is greater than 0.7 implying a high degree of internal consistency.Footnote 9 Since the contrast between protesters and non-protesters is of primary interest to this study, the resultant index is recoded into a dichotomous variable, with 0 representing low protest potential and 1 representing high protest potential.Footnote 10
Independent Variables
Life Cycle. Biological age is used as a conventional way to distinguish between three life phases: youth, middle age, and older adulthood. Those aged 18–30 are coded as young. Respondents aged 31–55 belong to the middle age category.Footnote 11 Finally, respondents aged 56 and more are coded as the elderly. This coding of age categories seeks to reflect local perceptions of the old age. According to the 2005 survey of Russians, one-fifth of respondents indicate that old age (starost') begins at the age of 50–54, and 47 per cent of the surveyed raise the threshold to 55–64.Footnote 12
Political Generations. Using the structure of Russia's educational system as a guideline for drawing temporal boundaries of the so-called formative years, this study treats middle and late adolescence (14–23) as crucial for the development of political attitudes. Prior to making a transition to a vocational college or high school, Russian adolescents enter the final year of compulsory schooling at the age of 14. Those who obtain a university degree usually enter the job market by the age of 23.
Table 1 summarizes the coding scheme employed to group respondents into political generations. Each respondent is assigned to a particular generation based upon the position in a historic period for at least five first years of the formative ten-year span. For example, an individual born in 1968 turned 14 in 1982 and spent the first two formative years under Brezhnev's rule. Yet it would have been misleading to classify this person as a member of the Brezhnev generation, since he spent a larger part of his pre-adult years during the Gorbachev period. Consequently, this person is assigned to the Gorbachev generation. For the purpose of logistic regression, four dummy variables are constructed representing each political generation.
Table 1 Classification of Respondents by Political Generations

Periods. The logistic regression model includes dummy variables representing two time points (1995 and 1999), with the year 1990 as the reference category. Throughout the decade, according to Freedom House rankings, Russia was “partly free,” scoring 4.5 in 1990, 3.5 in 1995 and again 4.5 in 1999 on a scale from one to seven.Footnote 13 By the end of Yeltsin's presidency, a whiff of political liberalization was gone, signalling a rollback of democratic reforms. In addition, poverty rates were on the rise in late 1990s, jumping from 11.5 in 1990 to 26.2 in 1995 and 34.2 in 1999 (Shorrocks and Kolesnikov, Reference Shorrocks and Kolenikov2001). A combination of political and socioeconomic factors characteristic of each time point (1990, 1995, and 1999) is likely to produce cross-time variations in protest potential.
Control Variables
Four additional variables are used as controls: income, gender, political interest and trust in government. Prior research shows that socioeconomic status is positively associated with unconventional political participation (Barnes and Kaase, Reference Barnes and Kaase1979). As an indicator of socioeconomic status, income is here measured on a ten-point scale, with greater values implying higher income. Respondents are grouped into three categories representing low, medium, and high levels of income. The high level of income is then treated as a reference category. Another control variable is gender (male=1). Empirical evidence suggests that being female negatively affects the odds of unconventional political participation (Burns, Reference Burns, Katznelson and Milner2002). Based upon the survey of 504 Moscow oblast residents, Tedin and Yap (Reference Tedin and Yap1993), for example, conclude that Russian women are less politicized and less likely to become involved in protest activity than men.
This study also hypothesizes that interest in politics provides a strong motivation for citizens to become engaged in political activity. The variable “political interest” distinguishes between those who report low levels of political interest (“not at all interested” or “not interested”) and those who express high levels of interest in politics (“very interested” or “somewhat interested”). In a similar vein, trust in government is coded as a dichotomous variable distinguishing between low (“none at all” or “not very much”) and high (“quite a lot” or “a great deal”) levels of trust. Low levels of trust in government are likely to breed discontent and propel political action (Gamson, Reference Gamson1968; Milbrath, Reference Milbrath1965).
Data Analysis Methods
Binary logistic regression analysis is performed to estimate the odds of protest potential with the help of age-related variables. A principal advantage of using logistic regression is that it does not assume a linear relationship between the dependent and the independent variables (Hosmer and Lemeshow, Reference Hosmer and Lemeshow2000). The general form of a logistic regression equation is:
where P is the probability of the event, exp is the exponential function written as exp(x) or e x (e is the base of the natural logarithm and is approximately equal to 2.72); b 0 is the intercept; b 1 is the coefficient for the explanatory variable; x 1 is the value of explanatory variable. Specifically, the logistic regression equation used in this study is as follows:
A glaring methodological conundrum in longitudinal studies of political attitudes derives from the linear dependency of age, generation and year (Magnusson, Bergmann, and Rudinger, Reference Magnusson, Bergmann and Rudinger1991; Mason and Fienberg, Reference Mason and Fienberg1985). To reduce multicollinearity, the present study excludes from a model several dummy variables measuring membership in a political generation. Each regression model includes dummy variables for two life cycle phases, two historic periods, and a single political generation. Admittedly, the survey period is rather short to unequivocally separate generational effects. Age and belonging to a political generation are likely to be strongly correlated. Given these limitations, the results need to be considered with caution.
Results
This study begins with identifying a cross-time pattern of protest potential. The initial analysis also distinguishes between types of protest activity. The results presented in Table 2 confirm Barnes and Kaase's argument regarding a hierarchy of unconventional political participation (Reference Barnes and Kaase1979). As intensity and resource demand of protest activity increase, protest potential declines. For example, 30 per cent of respondents signed a petition in 1990, whereas ten times fewer individuals went on strike. Similarly, one-third of respondents attended a rally in 1990, and only 4 per cent of the surveyed joined a boycott. The same trend is observed in mid- and late 1990s. According to the results from the 1995 survey, the number of participants in boycott action was tenfold less than that of demonstrators. Across the ten-year period, less than 1 per cent of respondents claimed they had occupied an administrative building in an attempt to articulate their grievances.
Table 2 Protest Potential by Year

Note: Column entries are percentages.
* p < .05;
** p < .01;
*** p < .001.
Cramer's V is a correlation coefficient that measures the strength of the relationship between categorical variables. Cramer's V ranges from 0 (no relationship) to 1 (perfect association).
Source: World Values Survey.
Furthermore, the results clearly show that the extent of protest potential has markedly decreased since the early years of transition from communism. The reported participation in a rally has declined by 11 per cent from 1990 to 1995. A drop in the number of petition signatories has been even more dramatic, sinking from 29.6 per cent in 1990 to 10.8 per cent in 1995. Subsequently, the survey data capture hardly any attitudinal change between 1995 and 1999. The same proportion of respondents (76 per cent) during the last two waves of the survey claimed that they would never consider joining a boycott. Likewise, 83 per cent of respondents were reluctant to join a strike action. A cursory look at the dynamics of protest potential in the 1990s suggests that period effects are at play here, with the peak of protest activity during the early phase of transition and a sharp decline afterwards.
Additional cross-tab analysis reaffirms the significance of period effects. The results show a systematic decline in protest potential from 1990 onwards, controlling for the location of the respondent in the life cycle. Moreover, differences in behavioural preferences of the young and the middle-aged shrink in the second half of the 1990s. For example, 68 per cent of young and 58 per cent of the middle-aged reported “high” protest potential in 1990, compared to 32 per cent in both age groups in 1999.Footnote 14 The results also reveal a drop in protest potential across generations.
Next, this study examines a bivariate relationship between protest potential and political generations. The findings highlight, in particular, attitudinal differences between representatives of the Stalin and Gorbachev generations. As shown in Table 3, the Stalin generation is least inclined to engage in political action. Only 11 per cent of respondents from this generation signed a petition and merely 1.6 per cent of them joined a boycott, while participation rates were much higher among the other generational groups. Forty per cent of respondents from the Gorbachev generation, for example, reported that they might participate in a rally, compared to 25 per cent from the Stalin generation. In addition, the findings suggest that members of the Stalin generation shy away, in particular, from radical forms of political action: an overwhelming majority of respondents born 1907–1935 would never consider joining a boycott or a strike. The Gorbachev generation, in contrast, is most open to the idea of engagement in these types of protest activity. Sixty-two per cent of respondents from the Gorbachev generation, compared to 84 per cent from the Stalin generation, would refrain from joining a boycott. Belonging to the Gorbachev generation is, however, strongly correlated with being young, and belonging to the Stalin generation is strongly correlated with being old.
Table 3 Protest Potential by Generation

Note: Column entries are percentages.
* p < .05;
** p < .01;
*** p < .001.
Source: World Values Survey.
This study further explores the interplay between life cycle and generational effects by analyzing the distribution of protest potential by life cycle controlling for membership in a political generation. As shown in Table 4, the protest potential of respondents changes in sync with their transition from one life stage to another. The level of protest potential declines among those members of the Brezhnev and Gorbachev generations who move from the young to the middle age category. Similarly, the level of protest potential decreases when members of the Stalin and Khrushchev generations advance into the old age. But the level of protest potential is not uniform across all generations at the same stage of their life cycle. As youth, the Brezhnev generation exhibits higher levels of protest potential than the Gorbachev generation. Likewise, members of the Khrushchev generation in the middle age category are more likely to engage in protest activity than their counterparts in the Gorbachev generation. It is unclear, however, from cross-tab analysis how a combination of age-related variables affects protest potential controlling for socio-demographic characteristics and such political attitudes as trust in government and interest in politics.
Table 4 Protest Potential by Lifecycle Controlling for Generation

Note: Column entries are percentages combining the response categories “might” and “have done.”
* p < .05;
** p < .01;
*** p < .001.
Source: World Values Survey.
Table 5 presents the results of logistic regression analysis using protest potential as the dependent variable. The results confirm the salience of life cycle effects on protest potential. Across models, the regression coefficient for “old” is negative and statistically significant (except model 1) indicating that older people are less prone to protest than youth in Yeltsin's Russia. The results for model 3, for example, suggest that the probability of protest potential increases by 23 per cent for young people, compared to the elderly.Footnote 15 As expected, the results also show that individuals protested at a lower rate in 1995 and 1999 than in 1990.
Table 5 Results of Binary Logistic Regression for Protest Potential

Note: Cell entries are logit coefficients b with standard errors in brackets and the odds ratio exp (b). Significance levels:
* p < .05;
** p < .01;
*** p < .001.
a Middle age is the reference category;
b 1990 is the reference category;
c High level of income is the reference category.
What is most interesting, however, is that the analysis finds statistically significant generational effects. Controlling for life cycle effects, membership in the Stalin generation reduces the probability of engagement in protest activity. For example, the statistics for model 1 suggest that an older male respondent of the Stalin generation, with a low level of income, a high level of interest in politics and a low level of trust in government had .32 probability of having “high” protest potential in 1999.Footnote 16 According to the results for model 2, the respondent representative of the Khrushchev generation with similar characteristics (old, male, a low level of income, a high level of interest in politics and a low level of trust in government) had .41 probability of having “high” protest potential in 1999. Contrary to the original hypothesis, belonging to the Gorbachev generation has a diminishing effect on the probability of protest potential.
One of the striking findings presented in Table 5 is that belonging to the Khrushchev and Brezhnev generations increases the probability of protest potential. By invoking the concept of uncertainty, the study proposes an explanation for these findings. Scores of journalistic accounts document how post-communist citizenry have become caught in a stress spiral in the aftermath of mounting uncertainties, ranging from job insecurity and wage arrears to food shortages and price hikes (Wines, Reference Wines2000). The study suggests that individuals who grew up during the tumultuous period of the late 1980s might have developed a better capacity for coping with uncertainty than those raised during an era of socioeconomic stability. Given a stark contrast between the economic equilibrium of the 1960–1970s and a swirl of dramatic changes in the 1990s, it is plausible to assume that members of the Khrushchev and Brezhnev generations exhibited a willingness to protest in a futile attempt to restore what they have lost in the course of the transition process.
The other individual-level variables, except income, perform as hypothesized and achieve conventional levels of statistical significance. The analysis finds that women are less likely to become involved in protest than men. By the same token, trust in government reduces the probability of engagement in unconventional political participation. Interest in politics, in contrast, emerges as the strongest predictor of protest potential implying that those who took a keen interest in public affairs formed the core of protesters in Yeltsin's Russia.
Overall, the findings clearly demonstrate that three age-related explanations of political attitudes need to be treated as complementary, rather than self-exclusionary. The results suggest that life cycle effects have the strongest impact on protest potential. But the analysis also shows that generational and period effects are important in accounting for variations in unconventional political participation.
Conclusion
Scholars have long debated the salience of age in shaping political attitudes, but there is a dearth of empirical studies estimating life cycle, generational, and period effects in developing countries and post-communist societies. The purpose of this paper has been to fill this gap by investigating the impact of age-related variables on protest potential in Yeltsin's Russia. Taken together, the analysis furnishes evidence in favour of age-related effects on the propensity to protest. In line with life cycle theory, the study finds that protest potential declines with age. Moreover, the results suggest that generational differences in political attitudes persist in the post-Soviet period despite the traumatic experience of the regime change. At the same time, the study provides support for the hypothesis that a period of social turmoil increases the odds of engagement in political action regardless of one's age and generational membership. The paper concludes that each of the rival explanations contributes to a more nuanced understanding of protest potential. A life cycle theory of political attitudes, however, has the strongest explanatory power.
Another important finding to be gleaned from this study is that generations socialized during periods of relative socioeconomic stability in the Soviet Union are most prone to protest against the status quo under Yeltsin's presidency. The 1990s were notorious for skyrocketing inflation, surging unemployment rate, and plummeting living standards. This study speculates that citizens socialized during the Khrushchev and Brezhnev's eras might have perceived the disorder associated with the shift from a planned economy to a market one as a startling deviation from the tranquil days of their youth. The resultant psychological stress might have aggravated individuals to the extent that they were willing to become engaged in protest activity.
One implication of these findings is that multiple approaches to the conceptualization of age-related differences need to be integrated in empirical analyses to provide a fuller understanding of political attitudes. A longitudinal examination of attitudinal and behavioural patterns is particularly useful in transition societies. Given cross-time differences in the scope, pace, and outcome of political and socioeconomic reforms in the post-communist world, a survey at a certain time point may obscure the dynamics of public response to societal changes.
These findings also have implications for political stability in post-communist Russia. One implication is that the importance of protest as a mechanism for channelling citizens' demands tends to decline with the advance of the transition processes. This shift away from unconventional political participation, however, may be contingent upon the performance of the incumbent government. Should trust in government drop to an all-time low, there might be a resurgence of political action. Another implication is that generational differences persist, notwithstanding a plethora of political, economic and social transformations in the post-Soviet era. Therefore, scholars need to pay greater attention to ways through which individuals develop political orientations and social norms supportive of the regime.
Acknowledgments
An earlier version of this paper was presented at the annual meeting of the Midwest Political Science Association, Chicago, 2006. I thank Ellen Carnaghan, Achim Goerres, Neil Nevitte and the journal's anonymous reviewers for helpful comments, and Crystal Cordell for the French-language translation of the abstract. I am also grateful to the World Values Survey Research Group and the Inter-University Consortium for Political and Social Research for generously providing access to the dataset. The author is solely responsible for all analyses and interpretation.
Appendix
Appendix. Questionnaire Wording and Coding of Variables
