Portal de Periódicos da CAPES

De Facto States in International Politics (1945–2011): A New Data Set

2014; Taylor & Francis; Volume: 40; Issue: 5 Linguagem: Inglês

10.1080/03050629.2014.915543

1547-7444

Adrian Florea,

Peacebuilding and International Security

AbstractSovereign states remain the primary units of analysis in conflict research. Yet, the empirical record suggests that the international system includes a wider range of actors whose behavior is relevant for conflict outcomes. This article introduces De Facto States in International Politics (1945–2011), a new data set dedicated to understanding the behavior of de facto states—separatist statelike entities such as Abkhazia. I begin by explaining why de facto states deserve attention. Further, I provide a definition of the de facto state that separates it from cognate phenomena. Thereafter, I offer an overview of the data set and illustrate its utility by demonstrating how it contributes to the literatures on war and state making, civil war, and rebel governance.KEYWORDS: civil warconflictfailed statespeacepeacekeeping/peacebuildingsecurity Notes1 Wimmer (Citation2013:85) discusses this selection issue at length.2 In other works, Roeder codes for the presence of some de facto states (Chapman and Roeder Citation2007) and autonomous territories (Roeder Citation2012).3 This is one of the main attributes in Caspersen's (Citation2012:11) definition of the "unrecognized state." A declaration of independence is also a key characteristic in Coggins's (Citation2011:454) definition of a secessionist movement.4 International legal sovereignty refers to recognition from a simple majority of UN Security Council permanent members plus recognition from a simple majority of UN members.5 The codebook includes a discussion of what de facto states are not.6 Abkhazia, Bougainville, Chechnya, Eritrea, Găgăuzia, Iraqi Kurdistan, Kosovo, Montenegro, Nagorno-Karabakh, Northern Cyprus, Republika Srpska, Krajina, Somaliland, South Ossetia, Tamil Eelam, Transnistria, Taiwan. All of these are included in the De Facto States data set except for Montenegro, which did not exert military control before independence in 2006.7 Abkhazia, Chechnya, Croatia, East Timor, Kosovo, Iraqi Kurdistan, Nagorno-Karabakh, Northern Cyprus, South Ossetia, Transnistria. All of these are included in the De Facto States data set except for Croatia, which functioned as a de facto state for one year only (1991).8 Abkhazia, Ajaria, Kosovo, Iraqi Kurdistan, Republika Srpska, South Ossetia, Nagorno-Karabakh, Northern Cyprus, Puntland, Somaliland, Southwestern Somalia, Transnistria. The De Facto States data set includes all of them except for Southwestern Somalia—a territory controlled by government-sanctioned Rahanweyn Resistance Army. Also excluded are two parent states (Bosnia-Herzegovina and Cyprus) that Sorens codes as de facto states.9 Abkhazia, Ajaria, Anjouan, Bangladesh, Biafra, Bougainville, Chechnya, Croatian Republic of Herzeg-Bosna, East Turkestan Republic, Eritrea, Găgăuzia, Hyderabad, Katanga, Kosovo, Krajina, Nagorno-Karabakh, Northern Cyprus, Republic of Mahabad, Republika Srpska, Somaliland, South Ossetia, South Sudan, Taiwan, Tamil Eelam, Transnistria, Western Bosnia, Western Sahara. All of these are included except for Bangladesh, Croatian Republic of Herzeg-Bosna, East Turkestan Republic, Hyderabad, Republic of Mahabad, and Western Bosnia (they do not meet the definitional criteria).10 I offer here only a brief description of how the data were coded. The codebook includes a chronology of events for each de facto state as well as details on coding and sources.11 The estimated mean takes into account censoring.12 Independence is one type of "disappearance" because it marks the transition to a different institutional status.13 Huang (Citation2012) uses some of these indicators to construct a rebel governance variable which she later connects to post–civil war democratization. However, Huang's unit of analysis is the conflict spell; hence, the coding for the state building variable captures the number of governance institutions at the end of the conflict (or the end of the observation period if the conflict was ongoing). De Facto States codes for the presence of governance institutions in each year.14 De Facto States also includes covariates on the parent state, such as military capacity, regime type, leadership change, ethnic diversity, or patterns of discrimination and repression against minorities.15 Since these preliminary results do not address issues of endogeneity or omitted variables, no causal interpretation can be attached to them. They are used here only for illustrative purposes.16 Each type of support is weighted equally.17 A faction can be a rebel group, political party, religious group, or civic group that operates within or outside the de facto state. A faction can demand reintegration into the parent state, limited or broad autonomy, no change in status, independence, (re)union with another state, or membership in a cross-national entity.18 It is worth noting that, counterintuitively, the Kaplan-Meier estimates (Figure 3) suggest that a low level of external military support does not necessarily raise the risk of de facto state disappearance. This is a pattern that deserves further exploration.FIGURE 3 External military support and duration of de facto separation.Display full sizeFIGURE 4 Rebel fragmentation and duration of de facto separation.Display full size