# Authoritarian Responsiveness Online: Social Media Strategy of Russian Regional Governance Centers
Replication materials for the paper "Authoritarian Responsiveness
Online: Social Media Strategy of Russian Regional Governance Centers"
## Abstract
*This article is dedicated to authoritarian responsiveness and feedback
mechanisms that autocratic regimes exploit to keep citizens loyal, using
the example of Russia. The study aims to describe how the Russian
government uses online feedback mechanisms in political communication. I
focus on the activities of the Regional Governance Centers, i.e.,
project management offices designed to establish communication between
authorities and citizens. I analyze the content of VK communities
operated by these project management offices, where they report how the
authorities solve citizens' problems initially indicated by citizens on
social media. The explorative text analysis shows that Regional
Governance Centers mostly report on how authorities process citizens'
complaints and solve problems indicated by them. They also promote live
Q&A sessions with regional and local authorities and advertise financial
support that different groups of citizens can get from the government.
Posts containing information about how to communicate directly with
officials, warnings about bad weather conditions, information about
financial assistance from the government, and self-promotion of these
Regional Governance Centers are the most popular in getting user
reactions.*
Here, I briefly describe the data, R scripts, and model outputs shown in
the paper's main text and online appendix.
## Data
`raw_vk_posts.RData` - contains VK posts of 84 channels operated by
Regional Governance Centers.
`vk_posts_without_reposts.RData` - contains VK posts of 84 channels
operated by Regional Governance Centers without reposts from other
communities.
## R_Scripts
This folder contains following scripts:
`General_Description_Text_Corpus.R` - for descriptive analysis of the VL
posts in the communities operated by Regional Governance Centers.
`Distribution_of_Posts.R` - a code to replicate Gini coefficient and
Figure 2 in the main text of the article.
`Text_Analysis.R` - for descriptive analysis of the VK posts from 84
communities operated by Regional Governance Centers (structural topic
modeling with a covariate indicating a region where a specific Regional
Governance Center operates and time of posting).
`Iterated_Computer_Assisted_Keyword.R` - for collecting words related to
mobilization and war in Ukraine.
`KeyATM.R` - keyword-assisted topic modelling to check the presence of
topic related to the war in Ukraine.
`Regression_Analysis.R` - to address the question on what aspects of the
activities of the Regional Governance Centers receive the greatest
response from VK users.
`vuong_test.R` - to choose the correct model, given the distribution of
the dependent variables (likes, reposts, comments).
## stm_outputs
The folder contains *csv* files with the diagnostic data for structural
topic modeling and outputs of the text analysis (pre-processed text, an
STM model with *k* = 13, etc.).
## regressions_outputs
The folder contains results of regression models to address Research
Question 2.
## Figures
The folder contains figures from structural topic models (visualization
of the results and diagnostics), distribution of posts overtime, etc.
## Miscellaneous
`Online Appendix.pdf` - contains additional analysis (STM diagnostics,
keyword-assisted topic modeling, regression analysis results, and
examples of posts).
