Article Review 98 on “How Civil Wars End: The International System, Norms, and the Role of External Actors.”11 min read

By Tim1965 / Tim from Washington, D.C., United States of America – Own work / Civil War Unknowns Memorial – looking E – Arlington National Cemetery – 2011, CC BY-SA 3.0, https://commons.wikimedia.org/w/index.php?curid=16636644

This is a significant article which attempts to document something that many political scientists have suspected[1]—that negotiated settlements of civil wars increased greatly after the Cold War and have declined since 9/11. This finding is significant because it implies that 1990-2001 may be an atypical period in conflict resolution and that generalizations based on behavior during that time may be misleading as guides to future actions, an unsettling suggestion to scholars and practioners alike. The authors argue that this change in behavior is the result of a change in international norms, driven by changes in the international environment. I think this is interesting, but rather to my surprise I am not persuaded by the article’s empirical foundation.

H-Diplo | ISSF Article Review 98
issforum.org

H-Diplo/ISSF Editors: Seth Offenbach and Diane Labrosse
H-Diplo/ISSF Web and Production Editor:  George Fujii

Lise Morjé Howard and Alexandra Stark.  “How Civil Wars End:  The International System, Norms, and the Role of External Actors.”  International Security 42:3 (Winter 2017/18):  127-171.  DOI:  https://doi.org/10.1162/ISEC_a_00305.

Review by Roy Licklider, Rutgers University and Columbia University
Published by ISSF on 31 May 2018

http://tiny.cc/ISSF-AR98
https://issforum.org/articlereviews/98-civil-wars
https://issforum.org/ISSF/PDF/ISSF-AR98.pdf

This is a significant article which attempts to document something that many political scientists have suspected[1]—that negotiated settlements of civil wars increased greatly after the Cold War and have declined since 9/11. This finding is significant because it implies that 1990-2001 may be an atypical period in conflict resolution and that generalizations based on behavior during that time may be misleading as guides to future actions, an unsettling suggestion to scholars and practioners alike. The authors argue that this change in behavior is the result of a change in international norms, driven by changes in the international environment. I think this is interesting, but rather to my surprise I am not persuaded by the article’s empirical foundation.

A very nice literature summary identifies three types of explanations for why civil wars end in different ways: domestic-structural characteristics of the states involved, bargaining dynamics including commitment problems, and third-party interventions. Lise Morjé Howard and Alexandra Stark argue that the first two were not changed by the end of the Cold War or by 9/11. The puzzle then is why the nature of third party interventions (widely defined) changed twice, once toward supporting negotiated settlements after the Cold War and then back to “stabilization” which became another term for military victory against fundamentalist Islamic groups and their allies after 9/11.

Howard and Stark argue that during the Cold War the United States and the Soviet Union pursued victories in their proxy wars because they saw the stakes as important. The unipolar system after the Cold War, dominated by the United States with low levels of international threat, was conducive to the norm of negotiated settlements and support of democratic governments for humanitarian and practical reasons, but 9/11 and the threat of fundamentalist Islamic movements, along with failures of democratization, resulted in a norm shift away from compromise with such groups.

Howard and Stark demonstrate the changes by using the Uppsala Conflict Data to compare how civil wars and armed conflicts have ended during three different periods.[2]

Table 1: Civil Wars and Armed Conflict Terminations, 1946-2013
1946-1989 1990-2001 2002-2013
Military victories 55% (82) 18% (23) 22% (16)
Negotiated settlements 11% (17) 38% (49) 32% (23)
Low Activity 33% (49) 44% (57) 47% (34)
TOTALS 148 129 74

 

I read this table as confirming that the second period (1990-2001) saw a major increase in the percentage of negotiated settlements and a corresponding decline in military victories. I do not, however, see much of a change between the second period and the third (2002-2013). As they note, negotiated settlements remain dominant, and the chi square statistics do not approach .05. The bulk of the article is thus dedicated to explaining a difference which does not seem to exist, and may have been more convincing had the authors been asked to focus on the former, since their argument that international norms influence policy might have been more plausible in explaining the shift from the Cold War to the post-Cold War, period 1 to 2.

In the first table in the article (Table 2 here) they also sort out those conflicts after 9/11 involving actors designated as Foreign Terrorist Organizations by the U.S. State Department and argue plausibly that these are less likely to be resolved by negotiation than others.

Table 2: (Table 1 in the Article): Terrorist Organizations and “Civil War” Termination
(Quotation marks added; a more accurate title is “Civil Wars and Armed Conflicts”)
DESIGNATED TERRORIST ORGANIZATIONS VICTORY OR ONGOING AS OF 2013 SETTLEMENT OR LOW ACTIVITY TOTAL
NO 58.70% (27) 87.10% (54) 75% (81)
YES 41.30% (19) 12.90% (8) 25% (27)
TOTAL 100% (46) 100% (62) 100% (108)

 

This conclusion is not certain, since Howard and Stark lump victories with those that were ongoing and put negotiated settlements and low activity together as well. Nonetheless, these differences are so striking that it suggests that the post-9/11 period norm is two-tracked, with negotiated settlements being inappropriate for those involving organizations regarded as terrorist while the remaining conflicts still fall under the norm of the second period. Presumably, if those involving terrorist organizations were omitted from the original table, there would be even less difference between the second and third periods. It would also be interesting to know if any conflicts in the first or second period involved organizations seen as terrorist. If so, this would have been an important finding.

I have another concern about the empirical work, but this is much broader than this particular article. I think it is a mistake to conflate civil wars and armed conflicts. If a militant group of Americans attacks a U.S. military installation, kills 25 people, and suffers 5 deaths, this is not a war. The original Uppsala data collection process made this distinction quite clear;[3] indeed the formal name of the database is Conflict Data, not wars. This distinction has increasingly been ignored, presumably in search of the larger number of cases which make some statistical processes more appropriate. Intuitively the distinction seems likely to make a difference. For example, I would guess that armed conflicts usually lead to government victories, are less likely to get extensive international intervention, and are less likely to result in substantial political changes than civil wars. I have no data on this, but it would be easy to run analyses of armed conflicts and civil wars separately; if they yield the same results, it makes the conclusions more persuasive and if not you have some interesting findings. It might also be interesting to use the battle death statistic as a ratio or logarithmic control variable since the 1,000 figure is obviously arbitrary. To their credit the authors do use the dichotomous 1,000 battle death measure as a control variable in their logistic analysis of shifts in mediation attempts and find that it is not significant.

Tracing international norms is enormously difficult. Howard and Stark do a very impressive job of tracking the change in global norms which they believe has driven the behavior change; indeed I believe that this is a model of such analysis. Using data from the Bercovitch Data Center’s Civil War Mediation Dataset, they show that mediation efforts increased substantially after 1989 and then declined after 2001.[4] A content analysis of the annual reports of the United Nations Security Council to the General Assembly shows that words with the stem “democ” (democracy, democratization) and “negotiate” declined after 2000 while “terror” and “stabilization” increased substantially after that time (163-168). I find this persuasive, but these changes would explain either the large-scale change in termination patterns suggested by the authors or my more limited one.

They also note that this norm change is unusual since the literature usually sees norms as the product of norm entrepreneurs, individuals who are able to persuade policymakers to adopt them. But they contend that in the post-9/11 period the norms were driven by national governments, the United States and its “great power allies.” In practice I think we are talking about the United States. The “great powers” are presumably China and Russia, neither of whom has shown any great enthusiasm for negotiated settlements in their own behavior either before or after 9/11. The authors assert that: “Norms were not an epiphenomenal, but rather a necessary and causal factor in this process” (130). I would argue the opposite, that after the Cold War, and especially after 9/11 the United States was both the sole creator and enforcer of these ideas, making these norms essentially an instrument of American foreign policy rather than a set of behaviors developed and integrated by other governments. The dominant American role in the United Nations and peacekeeping generally seems an adequate explanation to me for the changes in levels of mediation and rhetoric in the Security Council.

In the best recent research tradition, the authors turn to case studies to study causation. They argue that the cases of El Salvador and Bosnia show that the United States spent a lot of money and effort to create negotiated settlements after the Cold War when it could have allowed its ally in each case (the El Salvador government and the Bosniak/Bosnian Croatian forces) to win its wars and that this behavior was driven by the global norm change. I am not persuaded. Allowing the El Salvador government to win arguably would not have produced a stable country but one which would have been much more troubled than the current version in a sensitive region for the U.S. with the constant risk of spillover to neighboring countries. In Bosnia, Russian or Serbian intervention if the Bosnian Serbs were really threatened by military defeat was always a possibility, making a negotiated settlement a requirement rather than an option. It is also worth noting that it took a long time for the U.S. to decide to get involved despite the high civilian death toll, that its initial military involvement was to covertly support the development of Bosniak and Bosnian Croatian forces in violation of the United Nations arms embargo (talk about violating a norm), and that it was unwilling to commit American ground troops until an agreement had been worked out. The analysis also overlooks cases like Somalia and Rwanda where the U.S. was unwilling to pay the price involved in supporting this supposed norm.

I think that everything in this article can be explained simply as a result of American foreign policy. Perhaps this is merely a semantic issue. A norm, after all, is presumably an agreement on certain kinds of behavior. Does it matter whether it is defined and enforced by a single actor or not? It seems to me that norm also implies some sort of collective action in formation and enforcement, and I do not see that in this case. Howard and Stark, citing Peter Katzenstein,[5] argue “in Waltzian fashion” that “the international political environment—characterized by both material and ideational factors such as polarity, perceptions of first-order threats, and great power goals—gives rise to clusters of ideas of appropriate behavior known as norms” (129). This seems plausible, but I am not convinced that this is such a case. I think it can all be explained at Waltz’s second level, as the product of a particular (and somewhat peculiar) state’s foreign policy. I am persuaded in part by the counterfactual, my belief that another state in the same international situation would have produced a very different set of ‘appropriate behaviors.’

 

Roy Licklider is Professor of Political Science at Rutgers University and Adjunct Senior Research Associate at the Saltzman Institute for War and Peace Studies at Columbia University. His research focuses on civil war termination and postwar statebuilding. His books include Stopping the Killing: How Civil Wars End (New York: New York University Press, 1993) and New Armies from Old: Merging Competing Militaries after Civil Wars (Washington: Georgetown University Press, 2014). He is currently studying the impact of different kinds of collective memories on civil war recurrence.

©2018 The Authors | Creative Commons Attribution-NonCommercial-NoDerivs 3.0 United States License

 

Notes

[1] Cf. Mimmi Söderberg Kovacs and Isak Svensson, “The Return of Victories? The Growing Trend of Militancy in Ending Armed Conflict,” paper prepared for the 7th General Conference of the European Consortium for Political Research, September 2013; Adam Knight, “After the End of History, Before Everything Changed: Civil Wars, Democratization, and the End of the Cold War,” Political Science Department, Rutgers University, New Brunswick, NJ, 2017.

[2] My thanks to Professor Howard for providing me with the online appendix on which these figures are based.

[3] Peter Wallensteen and Karin Axell, “Armed Conflict at the End of the Cold War: 1989-92. Journal of Peace Research 30:3 (August 1993): 331-346.

[4] Karl DeRouen Jr. and Jacob Bercovitch, “Trends in Civil War Mediation,” in J. Joseph Hewitt, Jonathan Wilkenfeld, and Ted Robert Gurr, eds., Peace and Conflict (New York: Routledge, 2012): 59-70; Karl DeRouen Jr., Jacob Bercovitch, and Paulina Pospieszna, “Introducing the Civil Wars Mediation Dataset,” Journal of Peace Research 48:5 (2011): 663-672, DOI:  https://doi.org/10.1177/0022343311406157.

[5] Peter J. Katzenstein, “Introduction: Alternative Perspectives on National Security,” in Katzenstein, ed., The Culture of National Security: Norms and Identity in World Politics (New York: Columbia University Press, 1996), 5.