The Causes of Separatist Violence, by Danielle Rae


Introduction ("Why Should We Care?")

Civil war accounts for the vast majority of armed conflicts in the post-1945 world (Fearon and Laitin, 2003, pp. 75), and half of those have been fought over separatism (Griffiths and Wasser, 2019, pp. 1311). This makes understanding violent separatism a key topic in political science and conflict studies. Political scientists therefore seek to understand why some separatist movements erupt into violence while others remain peaceful, and over the course of this post I hope to illuminate some of the dimensions of this research, as well as some of the challenges and methods of political science in general.


What is separatism?

In political science, the meanings of key terms are frequently debated, with researchers defining those terms at the start of their projects and having to justify the boundaries they draw around the phenomena they study. While there are some widely-agreed examples of separatism -Eritrea, South Sudan, and Northern Ireland- other situations are contested. A common dispute is whether decolonization movements such as the Algerian independence war should be classified as a form of separatism or as their own phenomenon, with Griffiths and Wasser (2019) arguing for their inclusion against other scholars who would classify them as antioccupation movements (pp. 1313-1314). In general however, separatist movements are movements which aim at seceding a region from its parent state, and achieve that when the parent state no longer exercises control over the region and it has been recognized as an independent state by the international community (Griffiths and Wasser, 2019, pp. 1315). In some cases like Northern Ireland a separatist movement may seek instead to merge with a neighbouring state separatists view as their “true home.” Separatism is related to claims of national identity, which often form the basis for separatist mobilization (Cunningham, 2013, pp. 297), and struggles for self-determination, which range from demands for local autonomy within a state to demands for full independence, which crosses the line into separatism (Cunningham, 2013, pp. 292).


How violent is violent?

Similarly, what constitutes a “violent movement” or a “nonviolent movement” is up for debate: is it any movement in which rioting and clashes have occurred between supporters and opponents, or does it have to escalate into full-blown civil war? Most political scientists take an intermediate approach, classifying only movements which use violence as a major political tool or those which cause a certain number of battle-deaths as violent. Fearon and Laitin (2003), studying the outbreak of civil wars, limited their cases to those which had caused 1000 battle-deaths or more (pp. 76). Meanwhile in Griffiths and Wasser's work (2019), which also covered lower-level conflicts, a separatist movement was defined as violent if it met “a threshold of 25 battle-related deaths” (pp. 1315). What this means in concrete terms is that movements such as Quebec separatism are included as examples of violent movements in articles like Griffiths and Wasser's (2019, pp. 1330) while escaping Fearon and Laitin's. This means that, in articles like Griffith and Wasser's, aspects of the Quebec separatist movement are included, at least potentially, within their statistical analyses of the possible causes of separatist violence, while they are omitted from Fearon and Laitin's analysis. This shows how what one sees as the causes of separatist violence can depend on what one considers significant separatist violence.


What causes separatist violence?

Implicit in the question of what makes some separatist movements become violent is a false dichotomy: that some movements attempt to accomplish independence entirely through violence while others refrain. Cunningham (2013) cautions that this kind of binary thinking can obscure the ways in which violent and nonviolent methods may feed into each other, with many movements using a combination of methods (pp. 293). Dunning (2011), using Hindu nationalists as an example, showed that violent riots could be used for electoral advantage, combining violent and institutional forms of politics (pp. 332-333). It is therefore important to remember that political movements, including secessionist ones, can use violence alongside non-violent protest and institutional politics (elections, policymaking, court cases, and so on). Griffiths and Wasser (2019) found that 28 out of the 136 movements they studied combined violence, nonviolent civil resistance and institutional means (pp. 1324).

Whether they have thought of violent and nonviolent politics as binary opposites or as complementary tools, political scientists have suggested many different explanations for why political movements turn to violence. During the Cold War, scholars posited economic inequality as the primary factor in causing civil war, separatist or otherwise (Fearon and Laitin, 2003, pp. 78). In the immediate post-Cold-War period, events like the Yugoslav Wars saw new theories gain traction which centred culture, claiming that ethnically diverse states were prone to eruptions of separatist or nationalist violence (Fearon and Laitin, 2003, pp. 78). Controversially, Samuel Huntington claimed that Islam, as a religion, sparked violent conflict due to “demographic and cultural features” (Fearon and Laitin, 2003, pp. 85) including a prevalence of young males, who make ideal targets for radicalization and recruitment into rebel armies (pp. 86). Other culture-centric explanations abound, especially in the popular press where “ancient hatreds” are frequently used as a catch-all explanation for civil conflict, notably in the Yugoslav wars (Majstorovic, 1997, pp. 171).

These cultural explanations are now contested by more recent scholars, who broadly argue that structural factors are more important for determining whether political movements will turn to violence. Structural factors refer to circumstances surrounding separatist movements which make certain political methods seem more or less likely to succeed, more or less costly, or more or less profitable. These scholars, broadly “structuralists,” tend to agree that the more a movement is locked out of institutional channels of realizing its demands, and the weaker or more ineffective those institutional channels are, the greater the incentive to turn to violence (Griffiths and Wasser, 2019, pp. 1312, Cunningham, 2013, pp. 301, Dunning, 2011, 331). Structuralists also posit a certain kind of state which incentivizes violence: weak states with pervasive poverty, instability, and which borders a state the separatists see as “kin” (Cunningham, 2013, pp. 301, Fearon and Laitin, 2003, pp. 88). Cunningham (2013) differs from Fearon and Laitin (2003) in that the latter sees larger populations as a risk factor, while the former does not, with both using statistical analyses to support their hypotheses. Geographical features also play a role: terrain which is easy to hide in, as well as oil and diamond deposits, seem to increase the chances of political violence: the former it becomes easier to avoid detection by the military and the latter helps keep an insurgency supplied even when controlling only a “small enclave” (Fearon and Laitin, 2003, pp. 87). Rebutting Huntington, Laitin found no significant relationship between Islam and civil conflict when the presence of oil was accounted for (pp. 85-86).

In addition to the characteristics of the state, the characteristics of the movement have a major impact on whether it turns to violence. While Fearon and Laitin (2003) found no correlation between nationalist sentiment and the chance of civil war, Cunningham finds that independence struggles are more likely than others to result in violence (pp. 301). Since independence poses a direct threat to the integrity of the state, it should come as little surprise that states act harshly with secessionists in particular, even compared to anti-regime and anti-occupation movements (Griffiths and Wasser, 2019, pp. 1313). This is particularly true the more closely-integrated a seceding region is: movements in dual-states such as Czechoslovakia create less fear that the state will unravel completely if independence is granted to one region (pp. 1317). Cunningham (2013) also notes that the more fragmented a movement is, the more likely they are to turn to violence, since “When groups have multiple internal factions making demands, they have difficulty bargaining effectively... [making] civil war more likely.” (pp. 296, 300).

This last finding underscores a point of nuance in the structuralist position: they do not suggest a one-to-one correlation between the ability to win a civil war and the likelihood of a movement attempting one. A fragmented movement would likely have a harder time winning a civil war than a united front, but they are still more likely to attempt one because they have an even harder time achieving their goals through institutional means. Similarly, Dunning (2011) notes a debate between structuralist scholars: whether a position of near-parity of numbers or of great disparity between majority and minority groups is more likely to result in the use of violence (pp. 330-331). Since numbers make it easier for a movement to succeed both in war and in elections, different scholars have come down on either side.



There is far more that could be said about separatist violence, but hopefully this illuminates the dimensions of historical and present-day debates around the topic in political science. While I have found the structuralist account more persuasive in its overall methods, which tend more towards rigorous observation than the culturalist explanations, even within this broad framework there is vigorous debate over the causes of separatist violence and civil violence in general.




Cunningham, K.G. (2013). Understanding Strategic Choice: The determinants of civil war and nonviolent campaign in self-determination disputes. Journal of Peace Research, 53(3), pp. 291-304. doi: : 10.1177/0022343313475467


Dunning, T. (2011). Fighting and Voting: Violent Conflict and Electoral Politics. Journal of Conflict Resolution, 55(3), pp. 327-339. doi: 10.1177/0022002711400861


Fearon, J.D. And Laitin, D.D. (2003). Ethnicity, Insurgency, and Civil War. American Political Science Review, 97(1), pp. 75-90.


Griffiths, R.D. And Wasser, L.M. (2019). Does Violent Secessionism Work? Journal of Conflict Resolution, 63(5), pp. 1310-1336. doi: 10.1177/0022002718783032


Majstorovic, S. (1997). Ancient Hatreds or Elite Manipulation? Memory and Politics in the Former Yugoslavia. World Affairs, 159(4), pp. 170-182.