the new russian nationalismand foremost around the flag – that is, impersonal, institutional
symbols of their state and nation – while authoritarian countries
governed (as the saying goes) ‘by men, not laws’ may be more
likely to feature rallying first and foremost around the leader per-
sonally, a kind of rallying likely to prove more contingent since it
would be less structurally grounded. Finally, we should also con-
sider that both the political system and nationalist identity effects
could be significantly affected by the nature and magnitude of the
rallying events per se. The annexation of Crimea could be viewed
as a tectonic geopolitical shift, as a turning point in the long
durée of Russia’s history, the regaining of territories lost due to
the Soviet Union’s collapse a quarter century earlier – something
perhaps akin to the beginning of the ‘gathering of Russian lands’
after casting off the two- and- a- half centuries’ rule of the Tatar-
Mongols in the late fifteenth century.
In any case, Russia’s leadership certainly appears to have
tapped into a pre- existing deep well of nationalist sentiment that
has translated into a large ‘rally- around- the- leader’ (more than
‘rally- around- the- flag’) effect, distracting people from problems
that had previously led them to hold more negative perceptions of
their political system, if not of the leader himself.^11 But, regard-
less of polity type, to distract is not to eliminate or solve. And
the evidence indicates that the distracting action is associated in
the eyes of a great many Russians with negative economic con-
sequences that are ‘too expensive’. As the Crimean annexation
moves further into history, major problems loom for the Kremlin,
problems that it has yet to find ways of resolving.
Notes
- This sample size, drawn randomly from Russia’s population of
about 143 million, would typically result in a margin of sampling
error of about 3.1 per cent with a 95 per cent confidence level,
assuming that responses to a question are fairly equally split across
response options. The sampling error effect would decline the more
responses are skewed toward one of the options – for example, with
this Romir NEORUSS sample, it would be as small as 2.5 per cent
if the percentage of responses on a question were split 80–20. See