Controlled Confusion: Manipulation of Public Attribution of Responsibilities in Decentralized Autocracies
Last registered on November 19, 2018

Pre-Trial

Trial Information
General Information
Title
Controlled Confusion: Manipulation of Public Attribution of Responsibilities in Decentralized Autocracies
RCT ID
AEARCTR-0003578
Initial registration date
November 19, 2018
Last updated
November 19, 2018 11:09 PM EST
Location(s)
Primary Investigator
Affiliation
Columbia University
Other Primary Investigator(s)
Additional Trial Information
Status
Completed
Start date
2018-08-01
End date
2018-09-14
Secondary IDs
Abstract
Correct attribution of responsibility for economic outcomes is one of the key assumptions underlying citizens’ ability to hold politicians accountable: It allows citizens to use punishment and reward strategies to discipline politicians and to prevent them from introducing policies that contravene the preferences of the electoral majority. This project seeks to empirically assess whether (potentially biased) source of information can affect public perception’s of re- sponsibility in non-democratic setting. That is to answer whether government which controls large part of media in autocracies, can effectively shift public attribution of responsibility and therefore allocation of blame and credit for public policy outcomes. To tackle this question I propose a panel factorial survey experiment, which exposes random subsets of residents of the largest municipality in Russia (the city of Novosibirsk) to the pre-recorded local TV news reports covering outcomes of and/or allocation of responsibility for local health care (HC) policy. The experiment allows me to assess (1) whether the framing of responsibilities affects citizens’ attri- bution of blame and credit for particular policy outcomes to different tiers of government, and (2) whether the perceptions of the bias of the local media, knowledge of the issues covered and prior beliefs about local policy outcomes mediate the effects of such framing.
External Link(s)
Registration Citation
Citation
Syunyaev, Georgiy. 2018. "Controlled Confusion: Manipulation of Public Attribution of Responsibilities in Decentralized Autocracies." AEA RCT Registry. November 19. https://www.socialscienceregistry.org/trials/3578/history/37566
Experimental Details
Interventions
Intervention(s)
The intervention includes showing a random groups of adult residents of the city of Novosibirsk, enrolled in the study, two short video reports prepared by local online media outlet, \emph{Tayga.info}. These video reports are administered prior to the endline measurement either via online web access, or in person on a tablet. In total there will be four equally sized non-overlapping experimental groups in the study (see \Cref{fig:sample} for the sample treatment assignment structure), where each group was administered different combination of video reports. Each respondent is exposed to two video reports, approximately 60 seconds each.
Intervention Start Date
2018-08-16
Intervention End Date
2018-09-14
Primary Outcomes
Primary Outcomes (end points)
1. Evaluation of Public Health Care Provision
2. Responsibility allocation for public health care provision between three levels of government: Municipal, regional and federal
3. Evaluation of objectiveness of the local media outlet
4. Overall evaluation of performance of all three levels of government
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
I conduct a panel 2×2 factorial randomized controlled trial among citizens of the city of Novosibirsk, Russia. To test the predictions about the effects of media reports, I designed the intervention that includes showing a random groups of adult residents of the city of Novosibirsk, enrolled in the study, two short video reports prepared by local online media outlet, Tayga.info. These video reports are administered prior to the endline measurement either via online web access, or in person on a tablet. In total there will be four equally sized non-overlapping experimental groups in the study, where each group was administered different combination of video reports. Each respondent is exposed to two video reports, approximately 60 seconds each.
Experimental Design Details
Randomization Method
The assignment to one of the four experimental groups was conducted using block complete random assignment. The blocks of size 4 were constructed using optimal greedy algorithm on Mahalanobis distances by a number of baseline characteristics: Gender, age group, media independence evaluation, knowledge of allocation of responsibility for public health care provision, support for acting governor, and whether respondent agreed to offline/online mode of endline survey. I employ blockTools package in [R] to implement the block random assignment.
Randomization Unit
individual level randomization by block
Was the treatment clustered?
No
Experiment Characteristics
Sample size: planned number of clusters
1526 residents of the city of Novosibirsk from 18 to 64 y.o. who agreed to participate in the endline survey
Sample size: planned number of observations
1526 residents of the city of Novosibirsk from 18 to 64 y.o. who agreed to participate in the endline survey
Sample size (or number of clusters) by treatment arms
Every individual was assigned to one of the four experimental groups with equal probability resulting in 356/356/357/357 split.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Given the panel structure of the intervention, I use the baseline data to identify the MDE for the study. Using DeclareDesign package in [R] I find that for all main predicted effects I have at least 0.8 power to detect the effect size as little as 25% of standard deviation of corresponding baseline variable.
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
Columbia University Morningside IRB
IRB Approval Date
2018-08-16
IRB Approval Number
IRB-AAAR9146
Post-Trial
Post Trial Information
Study Withdrawal
Intervention
Is the intervention completed?
Yes
Intervention Completion Date
September 14, 2018, 12:00 AM +00:00
Is data collection complete?
Yes
Data Collection Completion Date
September 14, 2018, 12:00 AM +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
Was attrition correlated with treatment status?
Final Sample Size: Total Number of Observations
Final Sample Size (or Number of Clusters) by Treatment Arms
Data Publication
Data Publication
Is public data available?
No
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers