Generational effects of culture and digital media in former Soviet Republics

The study investigates how generational cohorts, cultural values, and media consumption relate to public opinion in three former Soviet republics (Belarus, Ukraine, Georgia). Motivated by debates on media influence (including disinformation) versus deep-seated cultural and generational effects, the authors compare responses to a recent international event (the January 6, 2021 U.S. Capitol riot) and to a long-term geopolitical preference (alignment with Russia vs. the European Union). They hypothesize that, consistent with Cultivation Theory and Selective Exposure Theory, broad geopolitical alignment will vary with year of birth, nationality, and other covariates reflecting formative socialization, whereas immediate attitudes to a recent event would depend more on media choices and less on individual covariates.

The paper situates its inquiry within theories of media effects and cultural change. Cultivation Theory posits long-term, cumulative media influence on worldviews through repeated exposure to discursive themes. Selective Exposure Theory suggests individuals preferentially consume attitude-consistent content, reinforcing existing beliefs and enabling homophily. The authors note social media’s dual role enabling both selective exposure and cultivation, and the increasing prevalence of coordinated manipulation (bots, trolls, sock-puppets). Broader cultural research shows national cultures differ and evolve intergenerationally, with values socialized in youth persisting over the life course. Historical narratives and national experiences can shape long-lived attitudes. The literature indicates that disentangling homophily and contagion from observational data is challenging, but patterns consistent with generational socialization versus short-term media effects can be probed. Regional background emphasizes Russia’s propaganda influence in FSRs and media environments differing across Belarus (state control), Ukraine (shifting media freedoms), and Georgia.

• Design: Representative surveys in Belarus (N=1014), Ukraine (N=2000), and Georgia (N=1000), fielded April–June 2021. Face-to-face interviews in Georgia and Ukraine; telephone interviews in Belarus due to unrest. Samples representative by age, gender, region, and settlement size; occupied territories (Donbas, Crimea) excluded in Ukraine. Small focus groups complemented surveys.
• Key questions and constructed variables:
  • Ephemeral event: Perception of how the U.S. Capitol riot affects U.S. global standing with options: stronger / no effect / weaker / don’t know. Constructed ordinal C: weaker=0; no effect=0.5; stronger=1. Analyses also compare “no effect” proportions.
  • Geopolitical alignment: Preference between improving relations with Russia vs. joining the EU. Constructed ordinal F: Russia=0; uncertain=0.5; EU=1.
• Media measures (respondents named up to three major sources):
  • Domestic mass media (Dm), Russian mass media (Rm), Domestic digital media (Dd), Russian digital media (Rd), Social media (S). Digital news websites/blogs were separated from social media; due to minimal differences, digital and social were combined for analysis.
  • Composite indices per respondent:
    • Digital vs. mass media use E: E = Rd + Dd + S − Rm − Dm.
    • Russian vs. domestic media use Q: Q = Rd + Rm − Dd − Dm.
  • Trust variants: Using “most trustworthy sources” question to compute Et and Qt analogous to E and Q.
• Cultural context: Incorporated national-level cultural factors (secularism vs religiosity; cosmopolitanism) from World/European Values Surveys, aggregated by decade of birth.
• Statistical analysis: 36 multivariate logistic regressions with binary outcomes (C or F), run separately by country with varying covariate sets:
  • Predict C from Ec, Qc; add F; add demographics (gender, urban/rural, wealth, education). Repeat with Et, Qt.
  • Predict F from Ec, Qc; add C; add demographics. Repeat with Et, Qt.
  • Marginal effects: For non-linear models, average marginal effects and their variances computed using the R package margins. Supplementary materials provide full model details.

• Generational effects on geopolitical alignment:
  • Year of birth predicts favoring EU over Russia: younger cohorts more pro-EU across pooled data (n=4000), Pearson r=0.113 (95% CI 0.082–0.144), p<1e-11. Country-specific patterns differ: in Belarus, a leveling among older cohorts suggests Soviet-era effects (born before 1970s).
• Media use patterns:
  • Younger respondents rely more on digital media than traditional mass media.
  • Greater reliance on digital media correlates with more pro-EU attitudes, strongest in Belarus; however, media technology choice is age-related and not necessarily causal.
  • Reliance on Russian media correlates with more pro-Russia orientations, particularly in Ukraine and Georgia. Year of birth shows little effect on Russian vs. domestic media use.
• Education and wealth:
  • Higher education and wealth correlate with more pro-EU views; strongest overall correlation is education level with pro-EU (r≈0.16, p<0.00001).
• Reactions to the U.S. Capitol riot (ephemeral event):
  • Minimal association with year of birth (n=4000, r=0.055, p=0.007) and no clear effects of education or wealth. Majorities in each FSR answered “No effect” or “Don’t know”. Among those with opinions, Belarusians were most pessimistic and Georgians most optimistic about U.S. standing post-riot.
  • “No effect” responses varied by media source: domestic media users were more likely than Russian media users to say “no effect”; digital media users somewhat more likely than mass media users to perceive an effect.
  • Pro-EU attitudes associate with expecting the U.S. to be stronger after the riot, indicating a broader pro-Western disposition across all three countries.
• Cultural values:
  • Across countries, more religious populations are also more pro-EU; within Belarus and Ukraine, younger cohorts are more secular and more pro-EU. Cosmopolitanism shows relatively less effect.
• Country differences in media influence on alignment:
  • Ukraine and Georgia: trust/consumption of Russian media predict pro-Russia alignment.
  • Belarus: trust/consumption of Russian media showed no detectable effect; trust/consumption of digital media predicted pro-EU alignment.
• Overall: Generational cohort and nationality are more consistent predictors of geopolitical orientation than media reliance; media influences are inconsistent and stronger for long-term alignment than for short-lived event perceptions.

Findings support both Cultivation Theory and Selective Exposure: deep cultural values and generational socialization strongly shape geopolitical orientation (EU vs Russia) in FSRs, while short-term reactions to distant events (U.S. Capitol riot) show weak links to media and demographics. Media’s role is context-dependent: Russian media trust/consumption aligns with pro-Russia preferences in Ukraine and Georgia, but not in Belarus, where digital media align with pro-EU views. This suggests selective exposure into culturally congruent media ecosystems and cultivation of long-run preferences within those ecosystems. Education and wealth modestly align with pro-EU views, but the dominant structuring effects are nationality and cohort. The results imply “long-wave” cultural dynamics, relatively independent of “short-wave” news cycles, and highlight how national histories, religion, language, and institutional trajectories shape orientations, with media acting within these broader cultural frameworks rather than determining them outright.

The study demonstrates clear generational effects on geopolitical attitudes in Belarus, Ukraine, and Georgia: younger, post-Soviet cohorts are more pro-EU and more digitally oriented. While media choice has some predictive power, especially regarding long-term alignment (e.g., Russian media linking to pro-Russia alignment in Ukraine/Georgia), it is inconsistent and weaker than cohort and national context. The most parsimonious explanation is homophily by birth year and nationality: digital media usage patterns reflect generational participation rather than driving geopolitical attitudes. Future research should investigate motivations for media source selection and examine how socio-political and cultural contexts shape selective exposure and cultivation over time, enabling stronger causal inference on media effects versus cultural generational dynamics.

• Observational design limits causal inference; homophily and influence are difficult to disentangle.
• Limited number of survey questions constrains parsing specific mechanisms (e.g., distinct propaganda narratives, nuanced attitudes).
• Different survey modes (phone in Belarus vs face-to-face in Ukraine/Georgia) may introduce mode effects.
• Media source categories were non-mutually exclusive; some subgroups (e.g., heavy combined digital/social users) had small sample sizes with large standard errors.
• Geographic exclusions (occupied territories) limit generalizability within Ukraine.
• Cultural factor comparisons rely on external aggregated WVS/EVS data by decade of birth, not the same respondents.