Abstract
How does conflict recurrence impact the likelihood of mediation onset in an intrastate conflict? In this thesis, I argue that a history of armed conflict will affect the conflict actors’ decision to accept mediation. To explore this proposition, I use dyadic data on recurrence and mediation in intrastate conflicts from 1975 to 2013. First, I explore the hypothesis that mediation onset is less likely in recurring intrastate conflict episodes than in initial conflict episodes. Mediation onset appears to be slightly less common in recurring conflict episodes. However, results from the regression analysis do not support the claim that this difference is caused by recurrence per se. Second, I focus exclusively on recurring intrastate conflict episodes. I analyze how the outcome of the previous conflict episode affects the likelihood of mediation onset. The results indicate that mediation onset is more likely in conflict episodes recurring after the previous episode ended with a negotiated agreement than in conflict episodes recurring after other previous outcomes. This is evidence of the need to consider historical ties between conflict actors when understanding decisions to accept mediation in intrastate conflicts.