Experimental measurement of polarization in attitudes towards war

The main experimental tool is an interaction between the respondent and a hypothetical peer, on the one hand, who can transfer the money to a hypothetical recipient in need. The respondent and hypothetical peer play the roles of two dictators who make independent transfer decisions to a recipient in need. Experimental manipulation consists in varying attitudes towards a potentially polarizing event (ongoing war) of the parties involved, which attitudes are known to the participant.

We test multiple hypotheses using 2x2x2 factorial design by varying across subsamples of the survey:
- attitude of the hypothetical alternative dictator to the war in Ukraine (labeled Special Military Operation (SMO), in accordance with the Russian legislation as of the end of 2025)
- attitude of the hypothetical recipient towards the SMO
- nature of hypothetical monetary transfer (lending or donating the money)
All these scenarios are described in one paragraph, formulated in Russian as follows:

"Suppose you and your colleague, not supporting/supporting SMO, have independently received unexpected bonuses of 100K Rubles, which is known to both of you. At the same time, it turned out that a fellow of yours who supports/does not support SMO, needs at least 100K for a medical emergency. This person asks both you and your colleague to lend for 1 year/donate him whatever you can, for this person has no one else to ask. What maximal amount you would be willing to donate him, not knowing how much your colleague will donate/lend?"

In additional question, we solicit normative expectations of the respondent about transfer/donation of the alternative dictator:

"How do you think, what maximal amount should your colleague lent for 1 year/donate to this person?"

In all cases, we assume that the recipient knows her/his own attitude towards SMO, which are solicited using a number of questions in the same survey before the experimental intervention.

2026-01-10

2026-01-31

Transfer decisions of the respondent under alternative scenarios are the main measure. Normative expectations about transfer of alternative dictator is a treatment-specific covariate.

Reported willingness to transfer the money to the hypothetical recipient under alternative experimental scenarios are taken as revealed attitudes towards the SMO. Differences in these transfers across scenarios controlling for expectations about alternative dictator's transfer and other individual responses and characteristics measure the attitudes towards person(s) sharing one's opinions, and hence of the strength of their position towards SMO.

Online survey of a randomly selected sample of 2500 individual respondents in internationally recognized regions of Russia. Survey conducted by a partner company Intra Research, using randomized procedures stratified by age, gender, and region, none interlocked soft quotas (in the range 0.8-1.2), with a minimum age range of 10 years.

Each individual participant will receive exactly one of 8 (= 2x2x2) descriptions of the main experimental situation, and shall provide an answer as a single number - maximum amount (s)he is willing to transfer to the hypothetical recipient under this scenario.

Randomization done by Intra research in office by a computer

Randomization at a level of individual respondents over all Russia database https://intraresearch.com/online-panels/europe/russia

No

2500 individuals

2500 individuals

approximately 315 respondents per treatment

Anticipated effect size of the main treatment variable (transfer of the respondent to hypothetical recipient) is expected to be about 0.4 = 20/50, where 20 (out of 100) is the expected difference in mean transfers across treatment, and 50 is normalized standard deviation. At alpha = 0.01 and power = 0.8, with out planned sample size we expect to have MDE of about 12 (out of 100).