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.
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.