chapter  4
11 Pages

The effects of mediators on peace institutions: statistical analyses

In the first part of this book, the basic logic of biased mediation has been laid out – the mechanisms through which biased mediators help to bring about peace institutions. Now we will examine the general empirical patterns. In this chapter I will therefore make an empirical examination of the effects of biased and unbiased mediators on the provisions in peace agreements reached in internal armed conflicts. I will examine the basic proposition that biased mediators outperform unbiased peacemakers in negotiating agreements on institutional peace arrangements through a statistical analysis of contemporary armed conflicts, in which the general pattern between biased mediation and institutional arrangements is examined. In particular, distinguishing between unbiased, governmentbiased, and rebel-biased mediators, the study examines the effect of these types of mediators on the provisions in peace agreements reached in internal armed conflicts. The study explores a range of different peace-agreement stipulations, constituting what we have called peace institutions: political, military, and territorial power-sharing; third-party security guarantees; and justice arrangements, involving amnesties (retributive justice) as well as the repatriation of civilians (restorative justice). This chapter consists of two sections: first the design of the research is presented, and then the results of the empirical analysis are reported.1

Research design This study uses unique data on the provisions stipulated in peace agreements, taken from the Terms of Peace Agreements Data (TOPAD) project (Nilsson, Svensson and Sundberg 2006).2 The study covers the 1989-2004 period and builds on data from the Uppsala Conflict Data Program (UCDP). This data set consists of 124 peace agreements.3 The unit of analysis is mediation-dyad-year, which implies that I focus primarily on the 320 conflict-dyad-years that have experienced mediation.4 In circumstances in which there were two or more agreements in one particular conflict-dyad-year, the “highest” value that was reached is used as the basis for the coding. For instance, in Guatemala, the “Agreement on the Resettlement of Population Groups Uprooted by the Armed Conflict” concerns the resettlement of refugees and displaced persons, and the content of this peace agreement is used to code the year 1994, although other