Replication package for: "Polarise and Rule: Independent Media in Autocracy and the Role of Social Media"
Authors/Creators
-
1.
Universitat Pompeu Fabra
-
2.
Institut de Política Econòmica i Governança
- 3. Barcelona Institute of Political Economy and Governance
-
4.
Barcelona School of Economics
-
5.
Ghent University
-
6.
University of Oxford
-
7.
University of Freiburg
Description
Replication package for: "Polarise and Rule: Independent Media in Autocracy and the Role of Social Media" (forthcoming in The Economic Journal)
The replication folder contains all data and code files for reproducing the tables and figures in the original article. The file “code_master.do” runs the individual do-files for both experiments (“code_city_experiment.do” and “code_individual_experiment.do”). Individual do-files include variable descriptions and code annotations for each step of the analysis. The results are stored in the “replication_output” folder (the preferred root to the folder should be manually adjusted in the do-file). Table R2 at the end of this document lists all the output files and links them to the respective table/figure in the text. The datasets are also provided in the CSV format for non-Stata users.
Additional information on the experimental designs, the survey instructions, and the questionnaires is enclosed in the two separate PDF files: “questionnaires_instructions.pdf” and “experiment_design.pdf”.
The letter with the IRB approval is included as a PDF file (“IRB_approval.pdf”) .
All code was created on Windows 10 using Stata/SE 18, and the packages required to run the Stata code are the following: outreg, outreg2, mmat2tex, grc1leg, unique, palettes, colrspace, ritest. (If replication is performed with an earlier version of Stata, we highly recommend removing the color scheme setting ‘set scheme stcolor’; also, please note that the user-written pscore command has to be updated to the latest version, or it may cause issues generating the match variable in earlier versions of Stata).
The user will have to change the central directory in “code_master.do”. The code takes about 3 hours to run.
Abstract
How do independent media affect the support of the regime in an autocracy? We carried out two complementary field experiments in Russia at the city and individual levels, randomising access to the country’s only independent TV channel before the 2016 parliamentary elections. In both experiments, we find that independent media foster polarisation. They increase turnout and pro-government votes among regime supporters. However, this effect only applies to voters who rely on news from social media. Among consumers of traditional media, the independent channel uniformly decreases regime support. Our results highlight how social media can condition the effect of traditional media.