The Soviet Union had a tricky time dealing with issues of ethnicity and nationalism within its borders for two reasons: first, it inherited a multi-ethnic territory combined with a definition of the Russian nation that combined ethnic, linguistic, and geographic characteristics.  This is challenging on its own, but the USSR also espoused a pan-national communist ideology, where a person’s primary identity was not meant to be ethnic, but class-based.  So it was never really clear whether Soviet nationality policy was meant to protect or eliminate national groups.  But what is clear is the impact that the policy of institutionalizing national groups had – enhancing ethnic or national sentiment in general as well as its identification with specific geographic borders.

Multi-tiered Ethno-Federalism

The Soviet Union had a multi-tiered federal system, where some federal entities – including the most important, republic level – were associated with specific ethnic groups.  Each tier had different rights and independent institutions. 

• Union Republic – “titular republic” because they are named (titled) after ethno-national groups.  These are the 15 republics of Soviet Union that became states when it collapsed (Ukraine, Uzbekistan, etc).  With the exception of Russia, which did not have an independent communist party, these republics had a full set of independent institutions, including parliaments, party structures, bureaucracies, etc.
• Autonomous Republic – these are regions within union republics (here in orange) that received special status based on ethnic or geographic distinctiveness.  Autonomous republics included regions that later tried to break away, like Abkhazia in Georgia and Chechnya in Russia, as well as regions that did not, like Nakhchivan in Azerbaijan and Tatarstan in Russia.
• Oblasts and krais (there is no legal distinction between them) are administrative regions that have local governments but less independence than an autonomous republic.  Autonomous oblasts were similar, but while oblasts were just administrative divisions, autonomous oblasts were associated with ethno-national groups.  These were relatively rare (and several remain within Russia) since most were converted to Autonomous Republics, but South Ossetia and Nagorno-Karabakh were both autonomous oblasts.
• Underneath the oblasts and republics came okrugs.  As with oblasts, they were mostly just geographic borders, but several autonomous okrugs were created to represent indigenous populations in northern Russia. 

All of these subdivisions had local governments and party structures, the main distinction between them was legal authority to take independent action – whether they could legally adopt policies different than the administrative structure above them. 

Why so many levels?

What determined what territory a group got?  Mostly group size and geographic concentration – they were not intended to be proto-states, even though that is what they ended up being. 

In some cases, notably Nagorno-Karabakh, the status of being an oblast or republic was controversial.  As an autonomous oblast, Karabakh did not have the legal authority to secede from Azerbaijan and join Armenia (an autonomous republic would have been able to).  When Armenians in the Armenian SSR and Nagorno-Karabakh AO sought to unify within the Soviet Union in 1988, this sparked the ongoing war between Armenia and Azerbaijan.

Legacies

The main legacy of this policy was a general association of borders and ethnicity, but imperfectly so. As Valerie Bunce noted in Subversive Institutions, ethno-federalism both provided the formal structures upon which republics could build states and the justification for these states to use ethnicity to define their nations.  The combination of a nebulous definition of a Russian nationality and clearly institutionized definitions of other nationalities contributed to the border conflicts that arose with the end of the Soviet Union.