How and When Amnesty during Conflict Affects Conflict Termination

• Date: 01 October 2020
• DOI: 10.1177/0022002720909884
• Published date: 01 October 2020
• Author: Lesley-Ann Daniels
• Subject Matter: Articles

Article

How and When Amnesty

during Conflict Affects

Conflict Termination

Lesley-Ann Daniels

1

Abstract

In the difficult process of ending civil wars, gran ting amnesty duri ng conflict is see n

as a useful option, withan underpinning assumption that trading justice for peace is

effective. However, is the case? This article tries to bring some clarity to when and

how amnesty given during conflict has an impact. Amnesty should have different

effects on diverse conflict endings: negotiated settlement, rebel victory, govern-

ment victory, or conflict reduction. The article also disaggregates amnesties to test

direct impacts as an incentive or through reducing the commitment problem, and

indirect effects that give military advantage to the government. Using a cross-

national data set of amnesties in dyadic conflicts from 1975 to 2011, the

research finds that amnesty’s strongest effect is, surprisingly, not as an incentive

but rather to reduce commitment problems. It can lead to negotiated settlements

but also to government military advantage. The results have implications for

negotiations and conflict resolution.

Keywords

civil war, amnesty, conflict termination, bargaining, transitional justice, commitment

problems

1

Institut Barcelona d’Estudis Internacionals, Spain

Corresponding Author:

Lesley-Ann Daniels, Institut Barcelona d’Estudis Internacionals, Campus de la Ciutadella, Ramon Trias

Fargas 25–27, 08005 Barcelona, Spain.

Email: ldaniels@ibei.org

Journal of Conflict Resolution

2020, Vol. 64(9) 1612-1637

ªThe Author(s) 2020

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0022002720909884

journals.sagepub.com/home/jcr

The Algerian civil war was known as the “dirty war” for its brutality against civi-

lians, journalists, foreigners, and other factions as well as the state. The Islamic

Salvation Front fought a violent civil war after the government cancelled results of

elections in 1991 under pressure from the military. The rebels splintered into the

Islamic Salvation Army (AIS) and the more hard-line Armed Islamic Group (GIA),

and both committed atrocities. Between 44,000 and 150,000 people died (Hagelstein

2008; Schulhofer-Wohl 2007).

During the course of the conflict, the government repeatedly used amnesties. For

example, an amnesty law in 1999 was included as part of the negotiations with the

more moderate AIS. Press reports suggest about 1,500 fighters took up this amnesty

(Human Rights Watch 2001). When the six-month limit to this law ended, President

Bouteflika issued a further amnesty by presidential decree. This second provision

was more generous and allowed amnesty for all crimes, even the most heinous. The

leaders of AIS, along with a breakaway group from GIA, decided to end their

conflict, and between 2,000 and 4,000 more fighters took advantage of this amnesty.

The rump of the GIA was slowly overwhelmed by military operations and by 2002

had effectively disappeared (Mallinder 2009; Tlemcani 2008).

The so-called peace versus justice debate hinges on the idea that the unpalatable

sacrifice of justice is necessary for peace, and there has been a large literature written

on the normative considerations involved (e.g., Ro ht-Arriaza 2006; Skaar 2012;

Sriram 2004; Vinjamuri and Snyder 2004). Although there has been a strong inter-

national turn against amnesties (Orentlicher 1991, 1996; United Nations [UN]

Secretary-General 1999), the use of amnesty has been remarkably constant as shown

in Figure 1. Indeed, recently there has been a reassertion of the advantages of

amnesties (Putnam 2002; Snyder and Vinjamuri 2003; Mallinder 2008; Freeman

2009; Mendez 2012; Patterson 2012). Amnesties have even been advocated as

acceptable under the responsibility to protect with the argument that the first priority

must be to save lives and that ending the conflict swiftly will bring a lower overall

death toll (Anonymous 1996). Underpinning these arguments is an assumption that

amnesties are effective in helping bring conflicts to a close. Given the stakes and the

moral hazard involved, surely we want to know whether the granting of an amnesty

actually helps to make peace. However, “there is much less empirical certainty about

amnesty’s benefits than many realists are ready to concede” (Freeman 2009, 7).

To address this problem, researchers are increasingly generating cross-national

analyses of justice mechanisms and civil war (Binningsbø et al. 2012; Reiter, Olsen,

and Payne 2012; Olsen, Payne, and Reiter 2010; Dancy 2018). Much of this research

focuses on the impact of amnesty after civil war, but there is increasing recognition

that the majority of amnesties occur during civil wars (Olsen, Payne, and Reiter

2010; Loyle and Binningsbø 2018), and I add to this literature. During the period

under study, 1975 to 2011, I register 269 amnesties during conflict, while a data set

of postconflict amnesties for the same period registers only sixty-eight amnesties

(Binningsbø et al. 2012). In most years, amnesty is given in more than 10 percent of

conflicts. Such amnesties are important in terms of determining the conflict ending

Daniels 1613