I am grateful to Kasper Thede Anderskov, Jurgen Brauer, Anke Hoeffler, Guy Lamb, Bjørn Lomborg, Andrew Mack, Ron Smith, and an anonymous referee for comments and suggestions. The usual disclaimer applies.

Introduction: the problem

In each year of the 1980s and 1990s there were between thirty and forty major armed conflicts in progress, though over the 2000s active major armed conflicts have declined. In 2007, there were fourteen major armed conflicts active in thirteen locations around the world, nearly all of which were internal or interstate disputes over government or territory – civil wars. There was a wide variation in the intensity of these conflicts, from “low intensity” guerrilla–government conflicts to conflicts between relatively large and well-equipped armies. While there are less of them they tend to last longer and range from ideological struggles (Mozambique, Eritrea, Nicaragua) to more fragmented decentralized conflicts (Somalia and Rwanda), with many a mixture of both at any particular time. The nature of war has changed, with an increasing role for less formal armies, lack of battlefield engagement, and increased involvement of civilians as victims. There are now more states and more disputes, but still relatively few lead to war.