Abstract
Abstract
Studies of the impact of foreign military intervention on the duration of civil wars most often fail to distinguish conflicts in which a single external state intervenes from those in which several outside states intervene. One influential quantitative analysis that does explore this distinction (Cunningham, Journal of Peace Research 47(2), 115–127, 2010) focuses primarily on whether or not the interests and preferences of the intervening state(s) coincide with those of prominent local actors. By revising this study’s dataset to clarify the distinction between single-state interventions and multiple-state interventions, it can be demonstrated that the latter are associated with lengthier wars than the former. Both types of foreign military interventions are correlated with civil wars that last longer than average.
Foreign military intervention affects the duration of civil wars (Elbadawi and Sambanis 2000; Linebarger and Enterline 2016; Paquin and Saideman 2017), but precisely how it does so remains an open question. Balch-Lindsay and Enterline (2000) claim that external intervention shortens civil wars whenever the interveners favor one side or the other, but prolongs the fighting if outside states back rival combatants (see also Balch-Lindsay et al. 2008). Regan (2002) likewise argues that foreign intervention in support of one side or the other accompanies shorter wars, yet he also shows that external involvement lengthens the conflict whenever it favors none of the warring parties. Regan and Aydin (2006), by contrast, find that foreign military intervention is associated with longer civil wars, whereas diplomatic initiatives undertaken by external actors effectively reduce the duration of internal warfare.
Cunningham (2010) makes a notable contribution to this research program by proposing that civil wars persist whenever outsiders inject extraneous goals into an ongoing conflict. It is often the case that ‘external states intervene to pursue independent objectives in the war outside of the goals of the domestic combatants. These states, then, fight to advance those (largely self-interested) objectives, not necessarily to help one side win or to help resolve the conflict’ at hand (Ibid.: 116). The introduction of such independent objectives makes it ‘more difficult [to bring the fighting to an end,] because there is [now] an additional actor who has to be either defeated militarily or consent to an agreement to end the war’ (Ibid.: 117). External intervention can therefore be expected to accompany lengthier civil wars.
Cunningham’s (2010: 122–124) sophisticated statistical analysis confirms the positive correlation between foreign military intervention and prolonged civil wars, but the proposed explanation for this finding conflates two analytically distinct lines of argument. The first hypothesis—and the one that Cunningham (Ibid.: 118–119) emphasizes—is that any single intervening state is likely to bring to the fighting a ‘separate agenda’ that is out of step with the preferences or interests of local combatants, thereby making a negotiated settlement harder to achieve. The second is that whenever multiple states intervene, their respective priorities and ambitions will tend to clash with one another as well as with those of local actors, making a mutually acceptable resolution more elusive (Ibid.: 117). In order to tease these divergent explanatory accounts apart, it would be useful to know to what extent single-state interventions differ from multiple-state interventions.
Single-state Versus Multiple-state Intervention
Unfortunately, no readily accessible dataset exists that clearly distinguishes between single-state military interventions and multiple-state interventions. The distinction between these two types of interventions is mentioned, but not analyzed, by Elbadawi and Sambanis (2000) and Regan (2002). Lounsbery et al. (2011) give greater weight to this distinction, but end up equating multiple-state interventions with joint operations carried out under the auspices of multilateral organizations (see also Regan 2000: 101–121). Talentino (2006) draws a distinction between unilateral and multilateral interventions, but includes in the former category instances in which several states intervene in a single conflict on their own initiative rather than under the auspices of a multilateral organization. Only Regan (2000: 31–32) directly addresses the difference between single-state and multiple-state interventions, reporting that civil wars in which one, two or three external states intervene exhibit approximately equal duration, but wars in which more than three outside states intervene last substantially longer.
Consequently, it makes good sense to explore the difference between single-state and multiple-state interventions using the dataset compiled by Cunningham (2010: 126–127). These data consist of 36 civil wars that elicited foreign military intervention, all of which broke out between January 1946 and May 2003. The cases are coded according to whether or not each intervening state involved itself in the conflict in pursuit of objectives that were congruent with those of the local combatant(s) it claimed to support. This procedure makes it possible for a given civil war to be counted not only as an instance in which one intervening state had goals that were in line with those of its local allies but also as an instance in which another intervening state harbored goals that stood at odds with those of its local allies. The 1975–2002 war in Angola, the 1993–2003 war in the Republic of Congo, and the 1996–2002 war in the Democratic Republic of Congo (DRC), for example, are accorded this mixed coding (Ibid.: 126–127). 1
The difficulties inherent in determining whether or not an intervening state’s objectives actually line up with the goals of local actors are apparent in Cunningham’s (2010: 121) illustrative discussion of the 1975–2002 war in Angola.
If the original data are rearranged according to the number of external states that intervened in each particular case, then the 36 civil wars separate into 21 instances of single-state intervention and 15 instances of multiple-state intervention. Five of these cases, however, appear to be characterized in problematic ways: (a) Egypt is designated as the sole intervening state in the 1962–1970 civil war in Yemen, ignoring the role of Saudi Arabia, the UK, and Iran in the conflict (Ferris 2013; Gause 1990; Orkaby 2017); (b) Mauritania is designated as the lone intervening state in the 1975–1991 war in Morocco over the Western Sahara, neglecting Algeria’s active involvement (Jensen 2011; Zunes and Munday 2010); (c) the Soviet Union’s 1946 intervention in northern Iran is counted twice; and (d) Chad and Angola are listed as having intervened in the Republic of Congo from November 1993 to March 2003, even though no outside states were involved in the initial round of warfare that took place from November 1993 to December 1994 and the second round of fighting that broke out in June 1997 came to an end that October (Clark 1998). 2
Cunningham (2010: 122) stipulates that in the case of recurring conflicts, ‘when[ever] there is a two-year break in fighting a new conflict is [to be] coded’.
A revised listing of pertinent foreign military interventions is offered in Appendix A. This dataset includes 20 single-state interventions and 18 multiple-state interventions. Three of the civil wars that attracted a single intervener lasted no more than 1 month each (Cuba in 1961, Gabon in 1964, and Gambia in 1981), another instance lasted just 2 months (Oman in 1957), and one more only 3 months (Nepal in 1950–1951). The average duration of the civil wars that attracted one intervening state is 77 months. By contrast, only one of the civil wars that pulled in multiple outside states lasted 1 month (Lesotho in 1998), and the next shortest instances lasted 2 months (Sri Lanka in 1971) and 4 months (Republic of Congo in 1997). The average duration of the wars that involved more than one external state is 124 months.
Types of Foreign Military Intervention and Duration of Civil Wars
Implications for Future Research
Why multiple-state military interventions accompany lengthier civil wars constitutes an important puzzle for scholars and policymakers alike. It may well be that single states intervene in conflicts that are simply easier to resolve, whereas multiple-state interventions occur in situations that are harder to bring to an end. Alternatively, Cunningham’s (2010: 117) argument that external states inject extraneous objectives into an ongoing civil war implies that whenever a greater number of outside actors becomes involved, the likelihood of reaching a bargain that is acceptable to all parties will plummet (see also Cunningham 2006, 2011). Yet the problems inherent in ascertaining the true objectives of intervening states (Cunningham 2010: 120), and the fact that it proved impossible to make a definitive classification of the goals of one-third (19 out of 60) of the interveners in the original dataset (Ibid.: 127), leave the door open to other kinds of explanations.
Cunningham’s (2010: 115) remarks concerning the impact of foreign military intervention on the 1996–2002 civil war in the DRC point the way toward a more compelling account. ‘Within a year’, he observes, ‘Rwanda and Uganda were again at war with the Congolese government, this time fighting alongside different domestic insurgents’. Furthermore,
throughout the 1998–2002 phase of the war, the rebel groups with which Rwanda and Uganda allied changed, those two countries battled each other, and each country proved willing to support whatever group it saw as advancing its goals and opposed any government or rebel organization it saw as standing in the way. (Ibid.: 117)
From this perspective, the initial objectives of the intervening states matter considerably less than do the actions that interveners undertake once they become engaged in the conflict. Contention among intervening states, along with shifting alignments between interveners and local forces, evidently played a crucial role in prolonging the fighting in the DRC. It can therefore be expected that civil wars with multiple-state interventions, in which there is both a substantial degree of rivalry among interveners and a high level of fluidity and uncertainty in relations between interveners and local actors, will take longer to resolve than wars with single-state interventions.
Both single-state and multiple-state interventions are associated with civil wars that exceed the average length of all post-1945 civil wars, which stands at 65 months (Cunningham 2011). Yet the average duration of single-intervener wars hovers close to that general average. It is multiple-intervener civil wars whose duration is exceptional. Whether larger numbers of intervening states bring to a conflict more numerous dyadic divergences of goals, or perhaps more pronounced divergences among the objectives of the actors involved, might be worth investigating. A more promising basis for explaining the increased duration of civil wars that attract multiple intervening states, however, is likely to be found in the dynamics of adversary–ally interaction (Snyder 1997). This conceptual apparatus, which was designed to explain relations among security-maximizing states in an anarchic arena, fits uneasily with the negotiation-based theories that dominate current scholarship on civil wars. But it has the advantage of relying less heavily on efforts to figure out the preferences and motivations of canny political leaders entangled in brutal struggles for power in socially and culturally diverse parts of the world.
Appendix A
**End date recorded as ‘ongoing’ in Cunningham (2011).
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.