Conservative Values and Authoritarian Support

This study includes, along with a series of direct survey questions, four survey-embedded experiments (priming, conjoint).

This study includes, along with a series of direct survey questions posed to samples of adults in the Russian Federation designed to be representative of the country as a whole and of six of its regions individually, a series of survey-embedded experiments (randomized controlled trials, RCTs) designed to explore the role of conservative values in authoritarian politics.

1. The first RCT is three-arm priming experiment. A control group gets no prime at all. The first treatment group is primed to reflect on what Russian President Vladimir Putin has said about how the state should treat members of the LGBTQ+ community. The second treatment group is primed to reflect on what Putin says about what the relationship should be between the state and the Russian Orthodox Church.

2. The second RCT is three-arm priming experiment. A control group gets no prime at all. The first treatment group is primed to reflect on what representatives of the United Russia Party have said about how the state should treat members of the LGBTQ+ community. The second treatment group is primed to reflect on what these representatives says about what the relationship should be between the state and the Russian Orthodox Church.

3. The third RCT is a conjoint experiment. All respondents are told about a hypothetical district parliamentary candidate who either does or does not possess the following qualities: support for legislation supporting religious believers, experience bringing material benefits to their home town, support for Putin, female gender, a long-term homosexual relationship, and support for restricting LGBTQ+ speech. The value on these items attributed to the candidate is independently randomly assigned.

4. The third RCT is a conjoint experiment to be conducted in a Russian region with a substantial traditionally Muslim population. All respondents are told about a hypothetical district parliamentary candidate who either does or does not possess the following qualities: belonging to different religious categories, experience bringing material benefits to their home town, support for Putin, female gender, a long-term homosexual relationship, and being of one of the region’s three largest ethnic groups. The value on these items attributed to the candidate is independently randomly assigned.

2021-04-26

2021-05-14

1. The outcome variable is a four-point scale of approving Putin’s work. Since what is of primary interest is the valence difference between support and non-support, this will be converted into a binary variable with all who lean supportive coded 1 and all others coded 0.

2. The outcome variable is a four-point scale of approving United Russia’s work. Since what is of primary interest is the valence difference between support and non-support, this will be converted into a binary variable with all who lean supportive coded 1 and all others coded 0.

3. The outcome variable is an 11-point scale of how likely the respondent would be to vote for the candidate.

4. The outcome variable is an 11-point scale of how likely the respondent would be to vote for the candidate

1. The outcome of primary interest is the difference in the average responses produced by the control subsample and the two treatment groups.

2. The outcome of primary interest is the difference in the average responses produced by the control subsample and the two treatment groups.

3. The outcomes of primary interest will be the average marginal effects of the different treatment variables on the outcome variables. Also of primary interest is whether the effects of the treatments reflecting conservative values change systematically depending on whether the hypothetical candidate in question is a Putin supporter/opponent or is someone with (un)proven experience delivering material benefits.

4. The outcomes of primary interest will be the average marginal effects of the different treatment variables on the outcome variables. Also of primary interest is whether the effects of the treatments reflecting conservative values change systematically depending on whether the hypothetical candidate in question is a Putin supporter/opponent or is someone with (un)proven experience delivering material benefits.

The survey agency VCIOM selects phone numbers from a national database at random by a computer within primary sampling units that are identified to produce a nationally representative sample of the adult population, following VCIOM’s standard sample design. Analogous techniques are used to generate samples representative of each of six Russian regions. The VCIOM interviewer then calls the selected phone numbers, choosing one member of the household. Interviews will be conducted by phone.

Randomization of treatment assignments will be accomplished by VCIOM using the randomization function in the software SurveyStudio.

Individual

No

0

1. 1600 adult residents of the Russian Federation. 2. 1800 adult residents of three regions of the Russian Federation. 3. 1800 adult residents of three regions of the Russian Federation. 4. 600 adult residents of one region of the Russian Federation.

1. Approximately 533 respondents per treatment arm.

2. Approximately 600 respondents per treatment arm.

3. Approximately 900 respondents per treatment category.

4. Approximately 300 respondents per treatment category where treatment variables are binary; about 200 when treatment variables have three values.

Analysis+Plan+-+LegitRuss+Values+20210415.docx