backing the ussr 2.0about twice as high as that of the non- Russians. Hence, using a
random number generator in the Statistical Package for the Social
Sciences (SPSS), about half of ethnic Russian respondents in the
Krasnodar sample were excluded. This improved the correspond-
ence between the two sub- samples on the distributional proper-
ties of socio- demographic indicators. The independent samples
t- test showed that age, sex, education level and household income
among respondents, as well as the size and location of the sam-
pling units – in fact, all socio- demographic control measures
available – were about the same between the sub- samples, and
corresponded closely with Russian averages. The results remained
constant with equal variances assumed and not assumed, indicat-
ing that the difference in sub- sample size had no significant effect
on sub- sample means.^11
The design was quasi- experimental. Ethnic identification was
the independent variable, generated by sub- sample selection. In
the language of experimental research, the ethnic non- Russian
sub- sample was the ‘treatment group’ and the ethnic Russian sub-
sample the ‘control group’. Additional control variables – held
constant as the t- tests showed – were socio- demographic charac-
teristics of respondents and sampling units. The five dependent
variables were territorial state identity preferences, ethnic and
civic pride, voting preferences, economic valuations and respon-
siveness to Putin’s statements. To rule out the selective measures
bias, the analysis included all survey questions under these topics.
Territorial state identity preference was measured by the ques-
tion asked in all four NEORUSS polls: ‘In the course of history,
the borders of states sometimes change. Where do you think the
borders of Russia should be – where they are now, but without
the Muslim republics of the North Caucasus (response value =
1); exactly where they are now (value = 2); where they are now
but with the addition of the former Soviet Slavic republics of
Ukraine and Belarus (value = 3); or where the borders of the
former Soviet Union were (value = 4)?’ In addition, the third and
the fourth responses were recoded as dummy variables (0 = not
selected, 1 = selected) measuring support separately for the Slavic
Union and USSR 2.0. The reference to ‘Slavic’ in the first item is
a hypothetical proxy measure of predominantly ethnic identity