« back to table of contents

Spoiler Management During Algeria's Civil War
Explaining the Peace
By Sofiane Khatib1

In December 1991, the Algerian army cancelled a parliamentary election allegedly won by the FIS (Front Islamique du Salut), an Islamic party and the main opposition movement against the Algerian government. Indeed, the FIS capitalized on growing public opposition to a 30-year old regime in the context of strong economic recession and weak governance. A High Committee of State, the collegiate institution closely controlled by the army, then assumed government duties.

Shortly after, guerrilla groups launched an armed struggle to obtain through arms what they could not reach by political means: namely, the creation of an alternative political system based on Islamic principles. During the next ten years, Algeria was subject to waves of violence involving government security forces on one side and a mosaic of Islamic rebel groups on the other.2 Thus far, the terrible harvest of the Algerian turmoil guerilla warfare, mass-killings, targeted assassinations of intellectuals has claimed the lives of 60,000 to 100,000 citizens, mostly civilians, tens of thousands have been wounded, hundreds of thousands of peasants have been displaced, and more than 3,000 have disappeared.3

Throughout these violent years there were several negotiation attempts. The first two peace efforts took place in 1995. In January, the main political opposition parties, including the FIS, offered a framework for peace negotiation through the Sant’Egidio platform.4 In February, the government issued the Rahma Law, which effectively offered a unilateral amnesty for most rebel fighters. Both initiatives failed to cease hostilities. The third case, in 1997, involved direct negotiation between the army and the AIS (Armée Islamique du Salut), the main guerilla group and fighting arm of the FIS. The talks succeeded in securing a cease-fire, and in 1999 the Civil Concord law gave a legal framework for the implementation of this peace agreement.

In this paper I will demonstrate that an effective handling of both internal and external spoilers groups aiming to derail peace processes by the Algerian regime was a key factor behind the success of the 1997 negotiation. I first will review briefly the spoiler management framework developed by Stanford professor Stephen John Stedman. Second, I will describe the two peace processes of 1995 and, applying Stedman’s framework, I will show how those negotiation outcomes helped set the stage for the peace talks of 1997. Thirdly, I will review the 1997 peace talks, detailing how internal and external spoilers were managed. Finally, I will analyze the current prospects of ending political violence in Algeria.

Setting the Framework: Spoiler Definition and Typology

Stedman defines spoilers as agents aiming to derail a peace negotiation process when “at least two warring parties have committed themselves publicly to a pact or have signed a comprehensive peace agreement.”5 According to Stedman, spoilers can be either insiders or outsiders. An insider spoiler signs an agreement, yet fails to implement it. For instance, during the Angolan peace process, the UNITA (Union for the Total Independence of Angola) acted as an internal spoiler. Despite having signed the Bicesse Accords in 1991, it returned to war shortly after losing the elections. Internal spoilers tend to depend on secrecy, keeping their true intentions hidden and staying in the peace process as long as it is advantageous to do so.

External spoilers are parties excluded from the peace agreement who try to derail progress through violence. During the Palestinian-Israeli peace process, Hamas consistently acted as an external spoiler, strategically concentrating its violence around major milestones of the process, such as the 1993 Olso peace accords, the 1994 Cairo Agreements, or the 1996 Palestinian and Israeli elections.6 External spoiler tactics include the assassination of moderates who could be potential negotiators, massacres that coincide with any progress in the negotiation process, and the creation of alliances within each camp to sabotage any agreement.

The number of spoilers also matters, as actions taken to marginalize one spoiler may inadvertently strengthen another. For instance, in Rwanda, the pressure exerted by the United Nations on the government to fulfill the terms of the Arusha Accord strengthened the CDR, an external spoiler, and enabled this group to sabotage the peace process.
Stedman also emphasizes the importance of recognizing that parties in a civil war differ in their goals and commitments. Spoilers vary by type: they can be limited, greedy, or total. Limited spoilers have narrowly defined goals, and thus can be persuaded to take part in negotiations if such talks address their goals effectively. On the other extreme, total spoilers have fixed goals, are insensible to the costs they may incur in achieving these goals, often rely on radical ideologies, and lack the necessary pragmatism to negotiate a deal in realistic terms. Greedy spoilers fit between these two extremes. Their demands expand when faced with low risks and costs and contract when faced with high risks and costs. A spoiler’s evolution (for example from greedy to limited) partly depends on whether the spoiler’s behavior is due to its leadership or its followers that is, its “locus” in Stedman’s terms. For instance, if the spoiler’s locus is due to its leaders’ positions, then a change in its leadership may help transform a total spoiler into a limited one.

Stedman defines three strategies to manage spoilers: inducement, socialization, and coercion. Inducement involves the fulfillment of a spoiler’s demands. This strategy is an especially appropriate way to deal with limited and greedy spoilers that is, agents with whom it is possible to talk in pragmatic terms. Socialization is a broader term that encompasses the necessity for a spoiler to accept a set of norms (e.g., respect of democratic rules), and, like the inducement option, it is applicable to limited and greedy spoilers. Finally, coercion involves punishing the spoiler’s behavior and reducing its capacity to obstruct the peace process. This strategy specifically aims to reduce the demands of limited or greedy spoilers and to marginalize total spoilers.

Bruce Jones, Deputy Director at the Center on International Cooperation at New York University, gives a more general definition of spoilers7 by distinguishing between two basic types: i) “losers,” which can be induced into a peace process and ii) spoilers, which can only be marginalized or eliminated.

This framework is well suited to describe the roles, actions, and evolution of spoilers during the Algerian turmoil. It is also a good framework to assess how the Algerian regime managed internal and external spoilers in order to curb violence. In my analysis I will consider internal spoilers as factions within the Algerian regime and civil society opposed to negotiations. This heterogeneous group was formed by apparatchiks, advocates of laicism, and security forces personnel, among others; the Algerian media referred to this group with the illustrative name of “éradicateurs.”8 The external spoilers, meanwhile, were diverse guerrilla groups opposed to any form of negotiation with the government.

To develop my analysis, I will broaden Stedman’s previously-stated definition of spoilers to include actors opposing any peace process even before formal talks begin. Indeed, Stedman’s definition is well suited to analyze reactions to the peace agreement signed in 1997, but it is too restrictive to judge the first two peace initiatives, Sant’Egidio and Rahma, since these attempts were turned down by the opposing parties before any negotiation took place. Nevertheless, the framework described above can be used to explain some of the reasons behind the failure of the 1995 peace initiatives: internal spoilers framed the peace proposal in terms not readily acceptable to other parties, while external spoilers worsened any prospects for peace through symbolic and deadly attacks a typical total spoiler strategy.

The Sant'Egidio Process

After nearly four years of escalating violence and no foreseeable solution to the crisis, the main political parties, under the umbrella of the Sant’Egidio community in Italy, offered a framework for peace. The platform included political parties representing both laic and Islamic sensibilities and accounted for more than 80% of the votes in the 1991 election. The three most important actors supporting this platform were the FIS, the projected winner of the election; the FLN (Front de Libération National), the party that ruled Algeria after its independence from France in 1962; and the FFS (Front des Forces Socialistes), a laic party with a strong regional presence in Berber areas.

In their final statement, the parties jointly proclaimed the rejection of violence to achieve political goals, as well as the respect of human rights and democracy. The statement also explicitly asked the Algerian regime to withdraw the army from the political scene, to recognize ethnic rights, to implement détente measures, and to end the practices of torture, retaliation against civilians, extra-judicial execution, and censorship. Similarly, the statement also asked rebel groups to cease attacks against civilians, foreigners, and public assets. Finally, the parties demanded the creation of a national assembly that would prepare and oversee new general elections.9

Three main spoiler tactics have contributed to rendering this initiative fruitless. First, internal spoilers within the Sant’Egidio platform accused the regime of “dirty war” practices by using a “naming and shaming” strategy. This prevented moderate members of the regime from participating in this process, as their involvement in the talks could have been interpreted by hard-liners within the Algerian government as an explicit acceptance of these charges. As a matter of fact, the Sant’Egidio public statement seemed more focused on drawing international attention to the crisis than on bringing the various parties to the negotiating table. Second, no public gesture was sent by the FIS toward its armed wing to signal its control over the insurgents in the battlefield; thus the government could not have been sure who really controlled the insurgency. Third, in a move undertaken by total spoilers, a suicide car bombing attack blew up the central police station in Algiers just two weeks after the Sant’Egidio statement was issued causing 43 deaths. This action, and the fact that it was not publicly condemned by the FIS, made it even less likely that the government would participate in peace talks. In February, one month later, a general rebellion in a high security prison caused more than 100 deaths, most of them former members of guerrilla groups, further worsening any prospect for peace.10 The Sant’Egidio process provided two lessons for the Algerian government. First, the FIS political leadership would likely adopt a hard stance in future negotiations, and second, any future peace process would likely be met with brutal acts by total spoilers. Nevertheless, few were prepared for the levels of violence reached in the fall of 1997.

The Rahma Law, Stalled Peace, and Gradual Progress

Nearly one month after the Sant’Egidio initiative, the Algerian government issued the Rahma Law, which reduced penalties against rebel group members. Citizens who turned in their arms and explosives and who were not involved in blood crimes, permanent injuries, or public asset destruction avoided prosecution. Insurgents who were involved in blood crimes could have their death penalty commuted to 15-20 years of imprisonment, and all other sentences were reduced by half.11

This initiative was not a negotiation; on the contrary, the law was a call to insurgents for desertion. Yet, it was heavily criticized by hard-liners as a concession.12 For some, the law was too indulgent as it gave special treatment to murderers of civilians and security forces personnel. At first, this initiative had little effect on the ground, most likely because at this point the guerrilla combatants still considered their goals plausibly achievable. The Rahma initiative showed how difficult it was for the government to frame a peace process while, at the same time, managing the most laic elements of its constituency.

The two previous processes showed that in 1995 neither party was committed to negotiations in realistic or pragmatic terms because the prospects of advancing their positions still seemed plausible. Using William Zartman’s terms,13 the conflict’s “hurting stalemate” was not yet reached. Yet, the reactions within each camp toward these peace attempts allowed for the identification of future would-be spoilers and potential drivers of a future negotiation process.

Quite illustratively, the regime thereafter focused its efforts on establishing direct talks with guerrilla commanders in the field, bypassing FIS representatives. Government officials perceived, through their previous experience, that FIS leaders were going to be hard negotiators who would impede any progress. As a consequence, the regime preferred to reduce the leverage of the FIS by first reaching an agreement with the AIS, the FIS’s armed wing, and then, thanks to this new “rapport de force,” negotiate an easier and more advantageous pact with the FIS.

The AIS Ceasefire and the Civil Concord Process (1997-1999)

This spoiler management strategy, combined with increased military pressure, helped unlock the situation in 1997. After direct talks between the army and the AIS, a cease-fire was reached in September and, as expected, was backed afterwards by the FIS leadership. This agreement included amnesty for AIS combatants and freedom for a select number of FIS representatives, as well as its main leader Abassi Madani.14 This cease-fire was a key milestone in the Algerian crisis because it publicly demonstrated the willingness of the FIS political and military leadership, the expected winner of the 1991 elections, to abandon armed struggle.

By successfully managing its own spoilers, the regime was able to frame the agreement in terms much more acceptable to the insurgents than those posed in 1995. A key aspect of this strategy was the army’s authority to handle the talks. The army was seen by laic portions of the society as the sole institution able to prevent the advancement of an Islamic-based political system in 1991. Furthermore, the army had paid a heavy toll in its fight against the guerrillas. Thus, this institution was not regarded by the public as prone to advance the Islamic agenda, so its “ownership” of the negotiations shielded the peace process against internal spoiler pressures. It would have been much more difficult for the government, which had gained some independence from the army after elections in 1995 and 1997,15 to drive the peace talks unilaterally as the government would have then faced pressures from two angles: i) appeasing measures would have been interpreted by its laic constituency as conceding toward Islamists in return for increased national and international legitimacy, and ii) the army, not managing the process, would have been much less willing to accept the terms of the deal.

Yet, external spoilers pursued an extremely violent campaign against the Army-AIS peace talks, the bloodiest actor being the GIA (Groupe Islamique Armée), a spin-off formed in 1993 by extremist elements of the AIS. The GIA was organized as a loose cell-type structure under regional chiefs or “emirs,” and each of its “battalion” commanders executed the instructions of the organization’s leader. Unlike the AIS, which focused its attacks primarily upon representatives of the government (army servicemen, police personnel, government officials) and upon the national infrastructure, the GIA primarily targeted entire segments of society for their declared ideas, professional activities, or mere presence. GIA squads assassinated intellectuals, journalists, members of the Catholic religious community, and foreigners. By 1994, an internal war had started between the GIA and the AIS over leadership of the armed struggle and strategies to gain power. In 1996 AIS elements killed GIA leader Jamel Zitouni. The best illustration of GIA viewpoints and methods the GIA used was a statement issued by one its leaders, Rachid Abou Tourab, on the possibility of peace talks: “neither truce, nor dialogue, nor reconciliation, nor security, but blood, blood, destruction, destruction.”16

The prospects of peace between the AIS and the army moved the GIA to employ further extreme methods to derail the negotiation process. Using a logic of ex-communion implying that all persons not explicitly supporting the GIA were enemies, the group started targeting Islamic moderates and civilians. The most infamous of their “Khmer Rouge” type of acts were the massacre of 700 civilians, including women and children in the outskirts of Algiers, during three different night assaults that took place between August and September 1997, right when peace talks between the AIS and the army were near conclusion. These attacks had two main objectives: i) to fuel the claims of internal spoilers within the government camp, forcing the army to pull back from the negotiation process and ii) to destabilize the Algerian government by showing how incapable it was in protecting its own population, even in its own capital. Some media even claimed that the massacres were the result of a criminal connivance between internal and external total spoilers.17

Despite GIA actions, an armistice was signed in September, but the GIA continued its campaign during the following holy month of Ramadan, causing more than one thousand deaths. Although the GIA still exists, its capacity to harm has gradually diminished in response to further military pressure and isolation from the population. Another insurgent guerilla, the GSPC (Groupe Salafiste pour la Prédication et le Combat), still opposes the peace process as well. This group was created by former elements of the GIA and the AIS who disagreed both with the GIA’s extreme methods and the peace process undertaken by the AIS. The GSPC replaced the AIS, the main perpetrator of attacks against the army, and is reported to be affiliated with the Al-Qaida network.18

In 1999, after the election of new President Abdelaziz Bouteflika the government promulgated the Civil Concord Law, which effectively provided the legal framework to implement the 1997 peace agreement. This law declared that citizens not involved in massive killings, bombing of public places, or sexual crimes (all of which were clear references to the GIA’s total violence strategy) would be placed under probation for a period ranging from 3 to 5 years, and could even participate in the fight against the remaining active terrorist groups. The imprisonment sentences were further decreased from those under the Rahma Law; for instance, death penalty and perpetual imprisonment were commuted to a maximum of eight years imprisonment for individuals under probation.

Because committees formed by lawyers and state officials were appointed to set the applicability and length of the probation period for each individual case,19 this system allowed a flexible interpretation of the law’s requirements and thus made possible the implementation of the 1997 agreement which called for an overall amnesty for AIS fighters.20

The Prospects for Peace

By 1999, several thousand guerilla members from the AIS, as well as other groups, had taken advantage of the amnesty offered under the Civil Concord Law. As a result the AIS dissolved itself in January 2000.21 However, despite having greatly diminished the death toll has declined from many thousands persons per year during the peak years of violence to less than a thousand per year in 200322 political violence has not disappeared yet in Algeria. To solve the conflict conclusively, the Algerian regime is pursuing a double strategy of military pressure and negotiations. Using Bruce Jones’s terminology, the Algerian government has identified the GSPC as a “loser” that could potentially be induced in the peace process,23 while it has identified the GIA as a total spoiler that could only be dealt with militarily.24

The army approached the GSPC from a different angle than the AIS. The GSPC’s leadership has strongly opposed any type of talks, acting either as a greedy or total spoiler; but military pressure and the loose organization of the GSPC in independent units permitted the army to weaken the command and control capacity of its leadership. Because the locus of the spoiler was more the result of the GSPC leadership than of its followers, the army was able to negotiate successfully with each rebel unit directly, and has thus driven the progressive demobilization of the GSPC. This course of action has proven to be quite successful. During the months of April and May, several GSPC units were reported to have abandoned the struggle.25

Concluding Remarks

The Algerian case exemplifies the importance of having an effective spoiler management strategy in order to enhance the possibility of reaching a peace agreement. The 1995 peace proposals highlighted the potential internal and external spoilers within each camp. The FIS political leadership acted, in Stedman’s terms, as a greedy spoiler, while internal spoilers within the government sphere demonstrated their capacity to block agreements. In 1997, the handling of the negotiation by the army shielded the process against internal spoilers within the regime, while the FIS’s “greediness” was reduced by first reaching a truce with its military wing, and then negotiating properly with the political leadership. Similarly, the GSPC, a spoiler of the 1997 agreement, is currenty being dismantled thanks to direct negotiation between army officials and each guerilla unit. The GSPC leadership has acted and is still acting as a total spoiler.

Spoiler management, however, is by no means the only factor behind the decline of political violence in Algeria. From a security standpoint, through the use of tactics better adapted to guerrilla warfare and the mobilization of civilian self-defense groups in isolated rural areas, military pressure greatly weakened rebel groups. From a political standpoint, an increased legitimization of the government thanks to several elections (in 1995, 1997, and 1999) and the inclusion of moderate Islamic parties willing to play the political game greatly diminished public support of the guerrillas. Nevertheless, spoiler management, as demonstrated, was a key factor in ending major violence in Algeria.


1 I would like to thank the “La Caixa” fellowship for supporting my research.
2 “Freedom in the World 2003: The Annual Survey of Political Rights and Civil Liberties Algeria,” paragraphs
3 “More than one million internally displaced people ignored by the international community,” paragraphs 1-3
4 Sant’Egidio is an Italian religious community whose peace building efforts have contributed to conflict resolution
in several countries, among them Mozambique, Albania, and Kosovo.
5 Stedman, 7
6 Kydd and Walter, 280-281
7 Jones
8 Dridi, paragraph 7
9 “La Plate-forme de Rome,” sections A-C
10 Loos, article 12, paragraph 2
11 “Ordonnance 95-12 du 25 février 1995 portant mesures de clémence”
12 Goutali, paragraph 2
13 Zartman, 11
14 Sidhoum, section “Mercredi 16 juillet 1997”
15 United Nations Eminent Panel, part four, section 1
16 “Head of Algeria’s most feared Islamic rebel army arrested: papers,” paragraph 11
17 “Algeria denies blame for massacres,” paragraphs 23-28
18“Algerian fighters to surrender”
19 “Loi no. 99-08 du 13 juillet 1999 relative au rétabl-
issement de la concorde civile”
20 “Truth and justice obscured by the shadow
of impunity,” paragraphs 20-27
21 “Algeria grants full amnesty to Islamic rebels,” paragraph 3
22 “Fighting persists in rural areas between Islamist extremist groups and security forces (2003-February 2004),” paragraph 3
23 Zerrouky, “Terrorisme : le pouvoir négocie la reddition
du GSPC La réconciliation à quel prix?”
24 Zerrouky, “Les GIA ne veulent pas se
25 “Algerian fighters to surrender”


“Algeria denies blame for massacres.” BBC, 7 Jan 1998. Nov 2004. <>.
“Algeria grants full amnesty to Islamic rebels.” CNN, 11 Jan 2000. Nov 2004. <>.
“Algerian fighters to surrender.” Al Jazeera, 23 April 2004. Nov 2004. <>.
Dridi, Daïkha. “De ce côté-ci des massacres.” Le Quotidien d’Oran, 4 Jan 2001. Nov 2004. <>.
“Fighting persists in rural areas between Islamist extremist groups and security forces (2003-February
2004).” Global IDP Project, Norwegian Refugee Council. Nov 2004. <
Winter 2005 61
“Freedom in the World 2003: The Annual Survey of Political Rights and Civil Liberties Algeria.” Freedom House. Nov 2004. <>.
Goutali, Amine. “Autres temps, autres moeurs.” L’Expression, 27 April 2004. Nov 2004. <>.
“Head of Algeria’s most feared Islamic rebel army arrested: papers.” Agence France-Presse, November 20, 2003. Nov 2004. <>.
Jones, Bruce. “Failed States: The UN’s Role.” Lecture.
15 April 2004.
Kydd, Andrew, and Barbara F. Walter. “Sabotaging the Peace: The Politics of Extremist Violence.” International
Organization 56.2 (2002): 263-296.
“La Plate-forme de Rome.” 13 Jan 1995. Nov 2004. <>.
“Loi no. 99-08 du 13 juillet 1999 relative au rétablissement
de la concorde civile.” Journal Officiel de la République Algérienne, 13 July 1995. Nov 2004. <>.
Loos, Baudouin. “Algérie: le dialogue conserve-t-il une chance malgré les bains de sang?” Algeria Watch, February 27, 1995. Nov 2004. <>.
“More than one million internally displaced people ignored by the international community.” Global IDP Project, Norwegian Refugee Council, 2004. Nov 2004. <>.
“Ordonnance 95-12 du 25 février 1995 portant mesures de clémence.” Journal Officiel de la République
Algérienne, 1 March 1995. Nov 2004. <>.
United Nations Eminent Panel. “Report of the Panel Appointed By the Secretary-General of the United Nations
to Gather Information on the Situation in Algeria in Order to Provide the International Community with Greater Clarity on That Situation.” United Nations, 10 Sept 1998. Nov 2004. <>.
Sidhoum, Salah-Eddine. “Chronologie d’une tragédie cachée (11 janvier 1992 11 janvier 2002 ).” Algeria Watch, 11 Jan 2002. Nov 2004. <>.
Stedman, Stephen John. “Spoiler Problems in Peace Processes.” International Security 22.2 (1997): 5-53.
“Truth and justice obscured by the shadow of impunity.”
Amnesty International, 8 Nov 2002. Nov 2004. <>.
Zartman, William. Elusive Peace: Negotiating an End to Civil War. The Brookings Institution: Washington D.C., 1995.
Zerrouky, Hassane. “Les GIA ne veulent pas se render.” Le Matin, 23 April 2004. Nov 2004. <>.
Zerrouky, Hassane. “Terrorisme : le pouvoir négocie
la reddition du GSPC La réconciliation àquel prix?” Le Matin, 23 April 2004. Nov 2004. <>.