Critical social media and political engagement in authoritarian regimes: the role of state media fairness perceptions

ABSTRACT

This study investigates how exposure to critical social media content influences political engagement in authoritarian regimes, focusing on Kazakhstan. Addressing a gap in existing research, this paper examines the moderating role of state media fairness perceptions. Using representative survey data (N = 6,800), the analyses confirm that critical social media exposure is positively linked with political engagement, including state-sanctioned activities like party contact and legislative engagement. However, this effect varies significantly based on individuals’ perceptions of state media fairness. These results suggest that critical content can foster broader political participation and highlight the nuanced role of state media in shaping political behavior in closed states.

KEYWORDS:

• Authoritarian resilience
• political participation
• critical social media
• Kazakhstan
• state media

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19331681.2024.2382695.

Notes

1. Survey was conducted in November 2021 by Research Institute “Public Opinion,” Astana, Kazakhstan. The data utilized in this study was collected by the institute and was provided to me for research purposes. The dataset was de-identified and anonymized prior to my access, ensuring the privacy and confidentiality of the participants. Given the anonymous and de-identified nature of the data, ethical approval was not sought or obtained for the analysis presented in this manuscript. The manuscript adheres to ethical guidelines for research involving human data.

2. Kazakhstan’s Bureau of National Statistics website: https://new.stat.gov.kz/

3. While this survey question was intended and presented as ordinal, the original design of its response categories is flawed. Yet, this should not present significant issues for the overall results as the author’s backstage implementation of a multinomial analysis demonstrated (see a regression table of the multinomial logit model and the predicted probability interaction plot in the appendix).

4. In a list experiment later in the interview, respondents were randomly assigned to two groups: one group received a list of 5 political actions to assess their willingness in case of rights violations, and the other group received a list of 6 actions, including protests. The analysis showed no significant difference between the two groups, increasing confidence in respondents’ truthfulness about their intent to participate in protests.

5. Age, Education, and Income ordinal categories were converted to numeric and rescaled to range from 0 to 1 for the analysis. Gender was made binary numeric with Female assigned 1 and Male 0. See the appendix for detailed distributions.

6. This process involved generating 30 datasets where missing data were replaced with draws from the posterior distribution (50 iterations per dataset), based on the observed values and assuming a normal distribution (Little & Rubin, 2020). The observed values used included all variables from the analyses. Instead of combining the results from these models and applying Rubin’s Rules to adjust for any deflation in standard errors, the average value across each imputed dataset was taken to replace the missing cases. With a high number of replicate datasets (30), the distribution of imputed values should approximate a normal distribution, resulting in an unbiased estimate for the mean imputation value of any missing case.

7. The regression table was generated using stargazer R package by (Hlavac, 2022)

8. For the details on how the penalized model was implemented, refer to the appendix.

Additional information

Notes on contributors

Amanzhol Bekmagambetov

Amanzhol Bekmagambetov is an Assistant Professor at the Department of Political Science and International Relations at KIMEP University, Almaty, Kazakhstan. His research focuses on the political effects of digital, traditional, and hybrid media use in authoritarian regimes, with a particular emphasis on political attitudes and both conventional and unconventional forms of political participation.