Is Communism to Blame for Political Disenchantment in Post-Communist Countries? Cohort Analysis of Adults' Political Attitudes

Abstract

In this article, we apply a new, original technique of cohort analysis to test empirically whether political disenchantment in the post-communist countries of Central and Eastern Europe can be linked to the previous political culture. On the basis of International Social Survey Programme 1996 and 2006 data we find a surprisingly similar and unique cohort effect in all analysed post-communist countries, reflecting persistent generational differences in perceived political competence (interest and understanding of political processes). However, the communist legacy does not seem to be important for explaining low self-efficacy or distrust in political authorities and their responsiveness to citizens' demands.

Notes

¹ We define political efficacy following Angus Campbell et al. as ‘the feeling that an individual political action does have, or can have an impact upon the political process’ (Campbell et al. 1954, p. 187).

² See the study by Alwin and Krosnick (1991).

³ One could argue that it would have been better to analyse actual political participation rather than attitudes. However, modes of participation have changed dramatically due to increasing use of digital technologies and online social media, making an analysis of the dynamics of participation extremely difficult and prone to misleading results. Our approach captures more general political dispositions.

⁴ The only exception here is Hungary. In this country middle-aged people were overrepresented in the 1996 sample, and the available weights were based on age groups, thus giving the answers of this cohort less ‘weight’. As we are specifically analysing answers by age of respondents, using such weights might skew the data (the answers would be biased towards other age groups) and distort the differences between years (the measured age and cohort effects would be inaccurate).

⁵ Different individuals within cohorts might be socialised differently, depending on their education, sex, position in society, political views, individual experiences etc. (Sapiro 2004; Torney-Purta 2004; Loewenberg et al. 2010). Cohort analysis, as noted by Sears (1983, p. 89) ‘can yield information only about the stability of a particular cohort's aggregate distribution of opinions, not about stability within individuals’. However, if the post-communist legacy hypothesis is true, the cultural impact should be observable in the society as a whole.

⁶ See, for example, the findings of Carnaghan (1996) and Neundorf (2010).

⁷ The wartime generation (born before 1925) was represented in our surveys by a small number of respondents, so it was merged with the post-war generation (born 1926–1940) in a category labelled ‘war and post-war generation’. This merging seems to be empirically justified (Mishler & Rose 2007).

⁸ Generational differences in political attitudes in different post-communist countries are presented in Appendix A, in the supplementary material for this article which can be found hosted alongside the article on tandfonline.com at: http://dx.doi.org/10.1080/09668136.2014.927641.

⁹ The new method and the way it disentangles cohort, life-cycle and period effects is described in more detail in Appendix E, in the supplementary material for this article which can be found hosted alongside the article on tandfonline.com at: http://dx.doi.org/10.1080/09668136.2014.927641.

¹⁰ Initially ‘I think most people are better informed about politics and government than I am’ was also considered in this factor, but based on relatively low factor loading, weak correlation with the other two questions and Crombach's Alpha, it was decided to exclude it.

¹¹ Due to a translation error, this indicator cannot be used for the 1996 wave of Latvian and Russian data.

¹² See the dynamics of attitudes in Appendix B, in the supplementary material for this article which can be found hosted alongside the article on tandfonline.com at: http://dx.doi.org/10.1080/09668136.2014.927641. The different dynamics of both confidence in authorities and their responsiveness among post-communist countries suggests that these perceptions depend largely on the performance of incumbents.

¹³ See Appendix C in the supplementary material for this article which can be found hosted alongside the article on tandfonline.com at: http://dx.doi.org/10.1080/09668136.2014.927641.

¹⁴ Some other variables such as individuals' economic position could also be relevant as control variables. However, no comparable income data were available in the two ISSP surveys.

¹⁵ In all regression models the reference category is the Reform generation. In order to test the differences between all cohorts, we used analysis of covariance (ANCOVA), including the same control variables and, in essence, replicating the models shown in Table 1.

¹⁶ The choice of a smoothing parameter is described in Appendix D in the supplementary material for this article which can be found hosted alongside the article on tandfonline.com at: http://dx.doi.org/10.1080/09668136.2014.927641.

¹⁷ See the full description of the method in Appendix E in the supplementary material for this article which can be found hosted alongside the article on tandfonline.com at: http://dx.doi.org/10.1080/09668136.2014.927641.

¹⁸ Since the graphs of country groups are constructed simply by merging all cases from which the individual graphs are constructed, the greater number of cases and similarities between countries lead to an increase of the R² as compared to individual countries.

¹⁹ The non-parametric method we use does not allow us to control for other covariates in the same way as the OLS regression. Thus, the solution we applied resembles the one used by Robert Putnam (1995). We divided the reported years of schooling into three groups: up to 10 years; 11–12 years; 13 years or more. The answers of each age group were weighted according to the average proportion of people with a certain level of education in this group and in the sample in general.

²⁰ Only in Great Britain, New Zealand (significance level 0.001), Norway and the United States (significance level 0.05) do we also observe at least some cohort effect, but the pattern is very different from post-communist countries. See Appendix G in the supplementary material for this article which can be found hosted alongside the article on tandfonline.com at: http://dx.doi.org/10.1080/09668136.2014.927641.

²¹ The Great Novena was a ten-year celebration dedicated to the millennium of the Polish adoption of Christianity in 966. It awakened romantic Polish nationalism and resulted in a confrontation between the Polish Church and the communist authorities. The massive mobilisation around the Church not only enhanced social capital but also provided training in social movements' methods and civic consciousness (Osa 2003).

²² The Czech government refrained from implementing the most dramatic changes in the economic and social sphere, consequently avoiding (or at least delaying) the initial ‘shock’. Thus, compared to other countries adolescents in the Czech Republic were the least cynical about the economic changes of the 1990s (Macek et al. 1998).

²³ Still, an important consideration that has to be taken into account in this type of survey is the time lag between events thought to have had a lasting impact on attitudes, and the time of the survey. According to the literature, the cohort effect can fade with time, as individuals adjust their cognitive schemes to the new reality. It is possible that we would have found at least some path dependency if we had looked for it sooner after the events presumed to create the ‘Soviet mentality’ took place.

At the time of writing this paper, Inta Mieriņa was Assistant Professor at the Institute of Sociology, University of Warsaw, working on an ERC Starting Grant Project ‘Public Goods through Private Eyes: Exploring Citizens' Attitudes towards Public Goods and the State in East-Central Europe’.