Social Cleavages and Civil War Onset

Abstract

This paper addresses the false dichotomy of ‘greed–grievance’ regarding the onset of civil wars, providing a theory that specifies the interaction of economic and identity-based factors within a two-stage causal process. The sociological concept of overlapping versus cross-cutting social cleavages is applied to explain the origins of the potential for ethnic civil war, as the overlap of ethnic cleavages with material cleavages fuels an in-group/out-group ideology. Such situations can, however, exist for decades without the onset of war, thus it is necessary to delineate this question from the necessarily separate, temporal question of the timing of the outbreak of such wars as a second stage. Catalysts such as economic decline and adverse regime change provide the variables that, in the presence of a social structure of overlapping social cleavages, exacerbate existing intergroup tensions to the point of organized violence, or are mitigated by the lack of major cleavages or the presence of cross-cutting cleavages. Evidence from cases is provided, followed by a discussion of the implications of this theory for further research.

Acknowledgements

This research was begun under the initial guidance of the late Donald Rothchild; the author remains eternally grateful. A special thanks to Caroline Hartzell, as well as the editors and two reviewers for their helpful suggestions in revising this paper. Additional thanks to Steve Burgess, David Laitin and Miroslav Nincic for their helpful feedback on earlier versions of this paper.

Notes

1. This analysis, as with most on the subject, adheres closely to the definitional standards in the Correlates of War project, defining civil war as intrastate violence challenging a sovereign state, involving the state as party to the conflict, and in a reciprocal violent exchange so as to distinguish civil wars from massacres. Most utilize the 1,000-deaths threshold, though scholars vary in how many annual deaths and in how much reciprocal violence must be present (Sambanis, 2001; Fearon & Laitin, 2003). This is not to dispute the possibility that this same theory may be applicable to smaller and communal conflicts, a question that would necessitate a separate analysis. For a detailed analysis of the challenges of definitional specifics, see Sambanis (2004).

2. 'Sufficient’ is used here not in an absolute sense but in a conjunctural sense; see George and Bennett (2005, p. 26).

3. Though FL code a civil war in Haiti in the early 1990s, this is invalid. Other leading data sets (e.g. COW and PITF) do not code a civil war, nor do country experts consider it as such. Even FL's case narrative for Haiti acknowledges that it was ‘more like a coup d’état that got out of hand, and it barely qualified (by our coding criteria) as a civil war’ (see Fearon & Laitin, 2012). The violence that followed the 1991 coup was overwhelmingly one-directional state repression, not reciprocal, as needed to qualify as a civil war.

4. This analysis is intended to provide a probabilistic, not deterministic, theory of the risk of ethnic civil war onset. It is also limited to onset, treating questions of escalation, intervention, the ways in which violence occurs during the war (as in Kalyvas), termination, and recurrence as analytically distinct. While state involvement is an accepted component of the definition of civil war, as noted above (thus distinguishing civil wars from communal violence such as riots and pogroms), the primary unit of analysis herein is major ethnic cleavages.

5. The categorization of Northern Ireland varies across different data sets, as its low casualty rate makes it sensitive to different coding standards for civil war, but it is often coded as a civil war. The relatively low level of violence in this case represents a question of escalation or the lack thereof, which, as noted above, is treated as analytically distinct from the question of onset.

6. Political scientist Christophe Jaffrelot, quoted in ‘India's scourge’, Time Magazine, 24 October 2010. Article by Jyoti Thottam.

7. See BBC News, ‘World Bank rediscovers inequality’, 20 September 2005, as well as the World Bank's World Development Report 2006: Equity and Development.