Independent media and political engagement in Kazakhstan

How to cite

OBrien, Dina. 2026. "Independent media and political engagement in Kazakhstan." AEA RCT Registry. AEARCTR-0016924. https://doi.org/10.1257/rct.16924-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Dina OBrien*

* Primary investigator

Registry ID
AEARCTR-0016924
DOI
https://doi.org/10.1257/rct.16924-2.0
First registered
Last updated
Status
On going
Trial dates
2026-03-02 – 2026-09-30
Countries
Kazakhstan (Central Asia)
JEL codes
C93, D72, L82, L86, L88
Secondary IDs
GI-GR-3082 J-PAL
Registry record
http://www.socialscienceregistry.org/trials/16924

Abstract

Independent media is widely regarded as a cornerstone for democratization, but citizens may respond in diverging ways. This study investigates how locally sourced, uncensored media information influences beliefs, attitudes, and political engagement within an autocratic context. In collaboration with an independent media organization in Kazakhstan, I conduct a field experiment in which individuals are randomly assigned to receive non-mainstream political information via a fact-checking service. The intervention lowers the cost of accessing independent information by delivering daily fact-checked messages via a dedicated channel on a popular social media platform. Evidence from the pilot study indicates that exposure to independent information leads individuals to update their beliefs about the country’s limited democratic practices and to express greater support for protesters. It increases polarization in their views of both the regime and the opposition, while reducing polarization regarding information control. The pilot results also reveal that, on average, exposure to independent information reduces political engagement, particularly among individuals with low prior support for democracy. The full study expands the sample size and further investigates the mechanisms of the effects of independent media exposure on political beliefs, attitudes and behavior, offering broader insights into the determinants of political engagement in environments where citizens face severe constraints on political activity, leaving protests as one of the few available outlets.

Intervention

The field experiment includes an intervention that isolates the effects of independent information in a setting with substantial amounts of disinformation. There are two treatment arms and one control arm. The control group is exposed to the general news, and the treatment groups receive the same information as the control group, plus independent media information. The treatment groups are split into two to investigate mechanisms. The treatment group's messages cover topics related to domestic politics, governance, accountability, and freedoms.

Experimental design

A sample of 1,200 respondents is randomly chosen from the general population aged 18 and above in Kazakhstan who are not yet subscribed to the Factcheck service. Randomization is stratified on two variables: i) the main language of communication (Kazakh or Russian), and ii) preferences for democracy. Respondents are divided into one control and two treatment groups. The participants
are randomized into social media channels based on their chosen language and assigned arm.

Design details

Randomization method
Stratified randomization
Randomization unit
Individual
Clusters
0
Observations
1200 individuals
Arms
800 in treatment and 400 in control
Primary outcomes
Media consumption, beliefs (about the state of democracy, economic outlook, importance of independent media), attitudes (toward government, protesters, opposition), and political behaviours.
Secondary outcomes
Attitudinal polarization
Sponsors
Name: J-PAL GI Url: https://www.povertyactionlab.org/initiative/governance-initiative-gi
Partners
Name: International Centre for Journalism "MediaNet" Partner_type: ngo Url: https://medianet.ngo/factcheck-kz

Cited by

No citation data available for this registration yet.

Keywords