War and Reconciliation: Reason and Emotion in Conflict Resolution (review)

In War and Reconciliation, William Long and Peter Brecke seek to understand whether reconciliation between former belligerents contributes to conflict resolution. As such, they have gone far toward making their analysis an important contribution to the growing literature on this topic. War and Reconciliation is one of the handful of books that attempt to develop a theory or theories of reconciliation and conflict resolution. Although the authors attempt to integrate theory and evidence in a way that generates useful policy recommendations, a number of problems—some related to the issue under investigation, others to the authors' presentation and claims—undermine its ultimate contribution to the literature.

The book has five chapters. The first provides an overview of the data on reconciliation following civil and interstate wars and a discussion of two models of reconciliation. Chapters 2 and 3 present synoptic case studies of civil and interstate wars respectively and assess the fit of the authors' models with the cases. Chapter 4 develops the models further and explains how emotions can be incorporated into rationalist models. The last chapter offers recommendations based on the authors' findings.

Although Long and Brecke provide a clear definition of a reconciliation event—an event that "includes the following elements: direct physical contact or proximity between opponents, usually senior representatives of respective factions; a public ceremony accompanied by substantial publicity or media attention that relays the event to the wider national society; and ritualistic or symbolic behavior that indicates the parties consider the dispute resolved and that more amicable relations are expected to follow" (p. 6)—what they are trying to explain is less clear. Their dependent variable is defined in multiple ways: as subsequent relations between former belligerents; effective conflict resolution; order restoration; successful reconciliation; or post-reconciliation relations between former combatants. A better or better-specified dependent variable might have been "the recurrence of war between the former combatants."

The authors' apparent difficulty in choosing what variation in reconciliation explains may be an artifact of the limited data available. The broader the definition, the larger the set of potential cases. Long and Brecke offer two sets of data on civil and interstate conflicts and reconciliation events. What becomes apparent is that reconciliation events as Long and Brecke define and select them are rare: Of 430 violent conflicts at the intrastate level, only eleven experienced a reconciliation event (or about 3 percent of all cases). Of these, 64 percent (seven cases) did not experience a return to violence, whereas 36 percent (four cases) did. Long and Brecke argue that this is striking because war recurred in only nine percent (or about 41 cases) of the civil wars that did not have a reconciliation event after hostilities ended. The problem is that not only are reconciliation events rare, but there are so few cases that the comparative statistics are extremely tenuous.

The data on international wars fare no better. Of fifty-three cases of interstate war, twenty-one involved reconciliation events, but only eight of these provided data capable of testing whether reconciliation events contributed to conflict resolution between the former belligerents. Worse still, even these eight provided only limited information: Five appeared to display improvement of relations, whereas the remaining three were ambiguous. Although the authors admit the limitations of their method and data, the inclusion of these cases and the discussion in the volume make it appear that they hold more validity than they do. For example, Long and Brecke go so far as to state that their findings show that reconciliation events do indeed influence the likelihood of a return to violence (p. 8). This may actually be true, but unfortunately the cases are too few to support such a claim.

Despite the limitations of the data, Long and Brecke present some theoretical insights into how reconciliation events might contribute to conflict resolution. They present two models, a signaling model based in rational choice and a forgiveness model based in evolutionary...