Policy Brief - Belfer Center for Science and International Affairs, Harvard Kennedy School Quarterly Journal: International Security

Rethinking Rebellions: A New Approach to Ending Civil War

April 2010

This policy brief is based on an article published in the Spring 2010 issue of International Security.


  • Negotiated Settlement to Civil War May Cause More Harm than Good. Civil wars ending in negotiated settlements are (1) more likely to recur; (2) no more likely to lead to democracy than other types of settlements; and (3) do not deliver increased prospects for economic prosperity. 
  • Rebel Victories Lead to Greater Stability and Democratization. Victories that allow combatants to fight to a decisive political outcome tend to result in ended wars that stay ended. Rebel victories in particular, lead to greater stability and democratization as compared to the status quo ante.
  • Giving War and Peace a Chance. Rather than simply shift to a willingness to support rebels in achieving victory, third parties should pay greater attention to security-sector reform during negotiations. Such settlements, which could credibly guarantee both benefits from cooperation and harm from defection, are likely to hold out better prospects for enduring peace, liberty, and prosperity following a civil war.

Since the end of World War II, policymakers have shown a marked preference for settling civil wars through negotiated settlements. The core recommendation of this policy is to employ third-party resources—primarily in the form of economic incentives and good offices—to halt the violence and preserve the combatants. Scholars of civil wars, for their part, have devoted the bulk of their analyses to exploring how best to achieve negotiated settlements. In recent years, however, other scholars have introduced a counterargument. Supporters of this "give war a chance" option advocate allowing belligerents to continue fighting until one side achieves a military victory.

The logic of both arguments is compelling, but flawed. The negotiated settlements camp takes it as axiomatic that the sooner the violence is halted, the greater the collective benefit. In contrast, the "give war a chance" camp assumes axiomatically that violence unfettered is the best path to lasting peace.


This study's data set of all civil wars fought from 1940–2007 (137 wars in total) shows that victories occurred more than four times as often as negotiated settlements and seven times as often as cease-fires/stalemates. The trend, however, was clearly toward a negotiated settlement. Through the 1980s, victory by rebels or governments was the dominant termination type, ending between 75 and 100 percent of wars per decade. In the 1990s, negotiated solutions ended 41 percent of all civil wars.

Does the trend toward negotiations correlate with improved outcomes? The data suggest that it does not.

Civil War Recurrence. Wars ended through negotiated settlement are twice as likely to reignite as those ending in victory. These renewed conflicts are more likely to last longer than wars ended by other means. Further, recurring civil wars following negotiated settlements were roughly 50 percent more deadly.

Not only does it matter that the war ended with victory but also who achieved that victory. The data show that rebel victories were more stable than government victories. Whereas 17 percent of wars ending in government victory recurred, only 6 percent of wars won by rebels did so.

Post–Civil War Politics. Negotiated settlements are associated with higher levels of authoritarianism over time. Incumbent governments faced with the likelihood of renewed war seem to sink precipitously into authoritarianism as they attempt to avert another round of fighting. Cease-fires/stalemates do not appear to have an impact on the level of autocracy or democracy. Although in general victory does not have much impact on regime type, the data suggest that when governments win, repression remains, whereas levels of autocracy decreased after rebel victories.

Post–Civil War Prosperity. Economic growth or decline is unrelated to the type of civil war settlement. Most of the states that suffered civil wars followed the same trajectory, with little divergence. The highest degree of divergence occurred among states whose civil wars ended with a rebel victory. These states suffered a decline in gross domestic product immediately following the war. Within ten years, however, they recovered, displaying the same level of economic performance as states whose civil wars ended in something other than a rebel victory.


At their ideal extremes, neither negotiated settlements (as typically designed and implemented) nor victories are apt to resolve civil wars in a way that (1) spares lives, property, and cultural treasures; (2) endures; (3) creates space for greater political liberty; and (4) establishes the conditions necessary for economic reconstruction, recovery, and development.

Policymakers thinking about future conflict management strategies should ask: What is it about negotiated settlements that has produced failure, and what is it about victories that has led to success?

A key strength of negotiated settlements is their credible promise to provide mutual benefit as a reward for continued compliance. The collective goods of (1) no more violence and (2) the opportunity to participate in an electoral process and a government that promises to represent and protect the interests of the citizens in some form are important benefits. Where negotiated settlements fall short, however, is in the credibility of their promise to inflict harm should one or both sides in a settlement fail to comply with its terms.

In contrast, victory means that the government or rebel military is left intact. Rebel victories are more stable than government ones because the rebels are not only in a position to harm (or threaten harm) their populations but also to benefit them. In winning, a rebel military organization remains capable of containing moves against its government. But because it is a rebel organization, it must to appeal to a portion of its domestic audience for approval as well as also to an international community not predisposed to the overthrow of national governments. This is also consistent with the move toward the greatest level of democratization following rebel victories. Rebels need to buttress the legitimacy of their win: allowing greater liberalization of the political system is an effective means to doing so.

Therefore, to achieve success, negotiated settlements need to offer a mix of benefits and threats. Securing the peace demands (perhaps paradoxically) making these threats as credible as the benefits, and then balancing them on a case-by-case basis in negotiated settlements or, failing that, in support of victory by one side or the other.

Rather than simply shift to a willingness to support rebels in achieving victory, third parties should pay greater attention to security-sector reform during negotiations, thereby increasing the possibility of achieving both short-term and long-term gains in better designed, implemented, and supported settlements. Such settlements, which could credibly guarantee both benefits from cooperation and harm from defection, are likely to hold out better prospects for enduring peace, liberty, and prosperity following a civil war.

The alternative is continuing to make promises of intervention that are progressively less credible to both domestic and other target audiences and accepting civil wars that will last decades. This, in itself, is something the developed world—where civil wars are rare—can no longer afford to ignore as a matter of interest, if not moral principle.

Statements and views expressed in this policy brief are solely those of the author and do not imply endorsement by Harvard University, the John F. Kennedy School of Government, or the Belfer Center for Science and International Affairs.


Toft, Monica Duffy. "Ending Civil Wars: A Case for Rebel Victory?" International Security 34, no. 4 (Spring 2010): 7–36. http://belfercenter.org/publication/20031

Toft, Monica Duffy. Securing the Peace: The Durable Settlement of Civil Wars. Princeton, N.J.: Princeton University Press, January 2010. http://belfercenter.org/publication/19670

Sambanis, Nicholas and Jonah Schulhofer-Wohl. "What’s in a Line? Is Partition a Solution to Civil War?" International Security 34, no. 2 (Fall 2009): 82–118. http://belfercenter.org/publication/19624

Willard-Foster, Melissa. "Planning the Peace and Enforcing the Surrender: Deterrence in the Allied Occupations of Germany and Japan." Journal of Interdisciplinary History 40, no. 1 (Summer 2009): 33–56. http://belfercenter.org/publication/19548

Toft, Monica Duffy. "Nasty, Brutish and Long." Prospect, April 2009. http://belfercenter.org/publication/18930

Greenhill, Kelly M. and Solomon Major. "The Perils of Profiling: Civil War Spoilers and the Collapse of Intrastate Peace Accords." International Security 31, no. 3 (Winter 2006/07): 7–40. http://belfercenter.org/publication/830


For more information on this publication: Please contact International Security
For Academic Citation: Toft, Monica Duffy. “Rethinking Rebellions: A New Approach to Ending Civil War.” Policy Brief, Quarterly Journal: International Security, Belfer Center for Science and International Affairs, Harvard Kennedy School, April 2010.