Ending Civil Wars: The Implementation of Peace Agreements
Most experts agree that international involvement is necessary to help end civil wars, but few have looked at how, when, and where the United Nations and outside governments should do so. Here a team of scholars examines the varied experiences of the last 20 years to offer practical advice. They find that the local conditions surrounding settlements vary widely. In relatively benign environments, such as Guatemala and Namibia, traditional peacekeeping approaches involving confidence building and dialogue are possible. In more demanding environments, such as Bosnia and Sierra Leone, outside states need to compel and deter local factions to ensure they comply with the peace agreement. But the most important source of failures is the presence of "spoilers" -- factions that continue to resist peace -- and unhelpful neighboring states that oppose a settlement. Another interesting finding is that no civil war accord has ever been completely successful in places where valuable, exportable commodities such as gems or timber exist. The authors conclude that the only chance for peace is the direct, long-term enforcement presence of a major state that sees the resolution of civil conflict to be in its own strategic interest. A model of rigorous analysis that yields usable -- if sobering -- knowledge.
Related
Some threats to international security are so potentially damaging that preventing them in advance is preferable to remedying their effects. In such cases, states should judge preventive actions by a standard of legitimacy, not strict legality.
U.N. Secretary-General Boutros Boutros-Ghali's clearest statement yet of his bid for reelection conveniently glosses over the greatest stain on his record: his failure to seize the moral initiative in Bosnia.
A new, hybrid form of peacekeeping is on the rise: regional interventions backed by the U.N. This solution may not be pretty, but unlike U.N. missions, it works.

Sign-up for free weekly updates from ForeignAffairs.com.