DP10109 Predicting Winners in Civil Wars

Author(s): Stephen H Haber, Kris James Mitchener, Kim Oosterlinck, Marc Weidenmier
Publication Date: August 2014
Keyword(s): asset prices, civil wars, conflict, predictions markets
JEL(s): F3, G1, N2, O1
Programme Areas: Development Economics, Economic History
Link to this Page: cepr.org/active/publications/discussion_papers/dp.php?dpno=10109

We develop a method to estimate which side will win a civil war. The key insight we deliver is that, for typical sovereign debt contracts, the probability of debt repayment will equal the probability of victory in a civil war. We test our predictor for standard outcomes in civil wars, including when the incumbent government loses (the Chinese Nationalists), when a new government is installed by a foreign power and decides to repudiate debt (the restoration of Ferdinand VII of Spain), and when there is a secession (the U.S. Confederacy). For China, markets were predicting a Communist victory three years before it happened. For the U.S., markets never gave the South much more than a 40 percent chance of maintaining the Confederacy. For Spain, markets considered the restoration of Ferdinand VII as likely (probabilities above 50%) as soon as France declared its intention to send military forces to the area.