Add 1st level category for state-like entities?

[philip-schrodt]

Should we add a first-level code (comparable to the IMG -- international militarized group -- code to deal with areas that are relatively stable and "state-like" but not formally states: this would include

• the former Yugoslavian entities during periods before they were states

• areas that eventually become East Timor, Eritrea and South Sudan

• current situations in West Bank, Gaza and the Hezbollah-controlled areas of southern Lebanon

• Greek and Turkish sectors of Cyprus

• ISIS areas for at least a couple years

• maybe the two halves of Libya for the last couple of years?

• same for Yemen if the current stalemate persists

• definitely Iraqi Kurdistan (and potentially extensions into Syria)

• various state-like sub-sets of Somalia

• Western Sahara before it was absorbed by Morocco

This is a decidedly non-trivial group and, e.g. in the KEDS project we never really came up with a coherent way of dealing with former Yugoslavia, though I think in some later variant in the Penn State work we had something like a "contested territories" code. "NST" for "non-state territories" maybe? Or "SLT" for "state-like territory" (neither of these codes are in ISO-3166-alpha-3) -- implication here is that the area has relatively stable borders within which there is an entity exercising state-like control (as opposed to situations like Boko Haram and LRA where they are mostly just terrorizing people). This will be fuzzy -- but hey, that's why actor dictionaries need to be open source -- but may still be better than trying to pretend a recognized state is controlling the territory when they aren't.

[ahalterman]

I think this is worth thinking more about about. Hong Kong came up as an example of a related phenomenon: semi-autonomous regions within defined states.